GAS PIPELINE HYDRAULICS E. Shashi Menon
Boca Raton London New York Singapore
A CRC title, part of the Taylor & Francis...

Author:
E. Shashi Menon

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

GAS PIPELINE HYDRAULICS E. Shashi Menon

Boca Raton London New York Singapore

A CRC title, part of the Taylor & Francis imprint, a member of the Taylor & Francis Group, the academic division of T&F Informa plc.

Copyright 2005 by Taylor & Francis Group, LLC

2785_Discl Page 1 Friday, March 11, 2005 10:28 AM

Published in 2005 by CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2005 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number-10: 0-8493-2785-7 (Hardcover) International Standard Book Number-13: 978-0-8493-2785-8 (Hardcover) Library of Congress Card Number 2004062949 This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data Menon, E. Shashi. Gas pipeline hydraulics / E. Shashi Menon. p. cm. Includes bibliographical references and index. ISBN 0-8493-2785-7 1. Natural gas pipeline--Design and construction. 2. Pipe--Hydrodynamics. I. Title. TN880.5.M455 2005 665.7'44--dc22

2004062949

Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com Taylor & Francis Group is the Academic Division of T&F Informa plc.

Copyright 2005 by Taylor & Francis Group, LLC

and the CRC Press Web site at http://www.crcpress.com

2785_C000.fm Page i Wednesday, April 13, 2005 10:38 AM

Preface Gas Pipeline Hydraulics is a practical handbook for engineers, technicians, and others involved in the design and operation of pipelines transporting natural gas and other compressible fluids. It is based on the author’s 30-year experience in the oil and gas industry. This book will help readers better understand and apply the principles of fluid mechanics to their daily work in the gas pipeline transmission and distribution industry. The book is divided into 10 chapters with several example problems solved fully, as well as additional problems provided as exercises. Chapter 1 introduces the basic properties of natural gas and other compressible fluids that are important in understanding how gas behaves under various conditions of pressure and temperature as it flows through a pipeline. The properties of hydrocarbon gas mixtures, such as gravity, viscosity, and compressibility, are reviewed, and both analytical and graphical methods are explained with illustrative examples. In Chapter 2, the methods of calculating the pressure drop in a gas pipeline are discussed. The General Flow equation is introduced as the basic equation, and the various correlations for friction factor and transmission factors, such as Colebrook and AGA, are explained. Other flow equations, such as Panhandle and Weymouth, are also covered using examples. Chapter 3 extends the concepts of pressure drop calculations further to determine the total pressure required for transporting gas in pipelines under various configurations, such as series and parallel pipelines. The effects of intermediate delivery volumes and injection rates along a distribution pipeline are reviewed. The importance of the contract delivery pressures and the necessity of regulating pressures using a control valve or pressure regulator are also discussed. The effect of gas temperature on the pressure drop in a transmission pipeline is reviewed with example output reports from a commercial gas hydraulics simulation model. Equivalent lengths in series piping and equivalent diameters in parallel piping are covered, as well as pipe looping to increase gas pipeline flow rate. The quantity of gas contained in a section of a pipeline and the calculation of line pack are also reviewed. Chapter 4 discusses compressor stations required to transport gas in a pipeline and how to calculate their numbers and optimum locations. Centrifugal and positive displacement compressors are explained, and their performances are compared. Typical performance characteristics of a centrifugal compressor are analyzed. Isothermal, adiabatic, and polytropic compression processes and horsepower required are discussed with sample calculations. The discharge temperature of the compressed gas and its impact on pipeline throughput, along with the necessity of gas cooling, are explained. In Chapter 5, installing pipe loops to increase the throughput in a gas pipeline is explored. Looping is compared to the option of building intermediate compressor stations. The advantages and disadvantages of looping a pipeline versus installing compressor stations are discussed. Chapter 6 covers the mechanical strength of a pipeline. The effects of pipe diameter, wall thickness, material of construction, and specific safety requirements

Copyright 2005 by Taylor & Francis Group, LLC

2785_C000.fm Page ii Wednesday, April 13, 2005 10:38 AM

dictated by design codes and state and federal regulations are reviewed. Hydrostatic testing requirements and classification of pipelines based upon their proximity to human dwellings and industrial establishments and population density are also covered. Chapter 7 introduces readers to thermal hydraulic analysis. For simplicity, long distance gas pipelines can be treated as isothermal flows. With compressor stations, the higher discharge temperature causes heat transfer between the pipeline gas and its surroundings. The effects of soil thermal conductivity, burial depth of pipeline, and the soil temperature are analyzed in determining the temperature and pressure profile of a gas pipeline. The Joule-Thompson cooling effect is reviewed, and the results from a commercial hydraulic simulation model are discussed. Chapter 8 introduces transient pressure analysis. This is an area that is quite complex, and a complete discussion of the transient hydraulic analysis of gas pipelines requires a separate book. Readers are referred to some excellent references on this subject. Chapter 9 covers valves and flow measurement. The different types of valves used in a gas pipeline and their characteristics are explained. The importance of accurate flow measurement in gas pipeline transactions is stressed. The codes and standards used to ensure proper design, construction, and operation of orifice flow meters are reviewed. Chapter 10 explores economic aspects of gas pipeline systems. Determining the optimum pipe size for a particular gas flow rate, taking into account the initial capital cost and annual operating and maintenance cost, is explained. For a typical gas pipeline system, the various capital cost components are reviewed, along with the recurring annual costs such as operation and maintenance, fuel, and administrative costs. Also, the calculation methodology for determining transportation cost or tariff is covered. At the end of each chapter, additional problems are provided as exercises. A list of references for the material covered in each chapter is included as well. Appendices at the end of the book include tables of conversion factors from USCS units to SI units and vice versa, along with tables of properties of natural gas. Also included are tables showing commonly used pipe sizes, listing allowable internal pressures and hydrotest pressures. A section containing a summary of all hydraulic formulas used in the book is provided as a handy reference. I enjoyed working with the fine staff at CRC Press. In particular, I want to thank Cindy Carelli, acquisitions editor; Theresa Del Forn, project coordinator; and Marsha Hecht, project editor, for their meticulous and prompt attention to all matters concerning the production of this book. They are indeed a very professional group and one of the best I have worked with over the years. I am indebted to my wife Pramila for the enormous amount of typing she did preparing the manuscript and for helping me proofread the final document and check for mathematical accuracy. I also appreciate all the good comments and suggestions that I received from practicing professional engineers such as Ken Zipp, Dan Bhavsar, and Charles Tateosian. In addition, I am thankful to Charles Peterson, Ken Zipp, and Ron Wood for agreeing to review

Copyright 2005 by Taylor & Francis Group, LLC

2785_C000.fm Page iii Wednesday, April 13, 2005 10:38 AM

the manuscript, which resulted in some valuable comments that enhanced the quality of this book. Finally, I would like to dedicate this book to my parents, who encouraged me in all my endeavors throughout my childhood and adult life. E. Shashi Menon

Copyright 2005 by Taylor & Francis Group, LLC

2785_C000.fm Page v Wednesday, April 13, 2005 10:38 AM

Author E. Shashi Menon is an engineering and computer consultant at SYSTEK Technologies, Inc., Lake Havasu City, Arizona. He is a registered professional engineer in California and the author of numerous engineering software programs and professional publications. He is an associate fellow of the Royal Aeronautical Society, London, and a member of the American Society of Mechanical Engineers. Menon earned a bachelor’s degree in mechanical engineering from Bhopal University, India, and an M.S. in mechanical engineering from California State University, Long Beach. He has worked for more than 30 years in manufacturing and the oil and gas industry. He has held the positions of design engineer, project engineer, project manager, chief engineer, and engineering manager for gas pipeline and liquid pipeline companies. Since 1998, Menon has taught pipeline hydraulics courses to engineers in North and South America and has published numerous hydraulics and pump software programs currently used by engineers in the oil and gas industry. He has authored books for Marcel Dekker and McGraw-Hill.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C000.fm Page vii Wednesday, April 13, 2005 10:38 AM

Contents Chapter 1 Gas Properties 1.1 Mass and Weight..............................................................................................1 1.2 Volume..............................................................................................................2 1.3 Density, Specific Weight, and Specific Volume ..............................................3 1.4 Specific Gravity................................................................................................3 1.5 Viscosity ...........................................................................................................4 1.6 Ideal Gases .......................................................................................................9 1.7 Real Gases......................................................................................................14 1.8 Natural Gas Mixtures.....................................................................................16 1.9 Pseudo-Critical Properties from Gas Gravity................................................19 1.10 Impact of Sour Gas and Non-Hydrocarbon Components.............................20 1.11 Compressibility Factor ...................................................................................21 1.11.1 Standing-Katz Method .....................................................................22 1.11.2 Dranchuk, Purvis, and Robinson Method .......................................23 1.11.3 American Gas Association (AGA) Method.....................................23 1.11.4 California Natural Gas Association (CNGA) Method ....................24 1.12 Heating Value .................................................................................................27 1.13 Summary ........................................................................................................27 Problems ..................................................................................................................28 References................................................................................................................29 Chapter 2 Pressure Drop Due to Friction 2.1 Bernoulli’s Equation ......................................................................................31 2.2 Flow Equations...............................................................................................32 2.3 General Flow Equation ..................................................................................33 2.4 Effect of Pipe Elevations ...............................................................................35 2.5 Average Pipe Segment Pressure ....................................................................37 2.6 Velocity of Gas in a Pipeline.........................................................................37 2.7 Erosional Velocity ..........................................................................................40 2.8 Reynolds Number of Flow.............................................................................43 2.9 Friction Factor................................................................................................45 2.10 Colebrook-White Equation ............................................................................47 2.11 Transmission Factor .......................................................................................50 2.12 Modified Colebrook-White Equation ............................................................54 2.13 American Gas Association (AGA) Equation.................................................57 2.14 Weymouth Equation.......................................................................................61 2.15 Panhandle A Equation....................................................................................64 2.16 Panhandle B Equation....................................................................................68 2.17 Institute of Gas Technology (IGT) Equation ................................................70 2.18 Spitzglass Equation ........................................................................................74 2.19 Mueller Equation............................................................................................76 2.20 Fritzsche Equation..........................................................................................77 2.21 Effect of Pipe Roughness ..............................................................................78

Copyright 2005 by Taylor & Francis Group, LLC

2785_C000.fm Page viii Wednesday, April 13, 2005 10:38 AM

2.22 Comparison of Flow Equations .....................................................................80 2.23 Summary ........................................................................................................81 Problems ..................................................................................................................82 References................................................................................................................83 Chapter 3 Pressure Required to Transport 3.1 Total Pressure Drop Required........................................................................85 3.2 Frictional Effect .............................................................................................86 3.3 Effect of Pipeline Elevation ...........................................................................86 3.4 Effect of Changing Pipe Delivery Pressure...................................................90 3.5 Pipeline with Intermediate Injections and Deliveries ...................................93 3.6 Series Piping ................................................................................................104 3.7 Parallel Piping ..............................................................................................111 3.8 Locating Pipe Loop......................................................................................121 3.9 Hydraulic Pressure Gradient........................................................................123 3.10 Pressure Regulators and Relief Valves ........................................................126 3.11 Temperature Variation and Gas Pipeline Modeling ....................................129 3.12 Line Pack......................................................................................................132 3.13 Summary ......................................................................................................135 Problems ................................................................................................................136 References..............................................................................................................137 Chapter 4 Compressor Stations 4.1 Compressor Station Locations .....................................................................139 4.2 Hydraulic Balance........................................................................................146 4.3 Isothermal Compression ..............................................................................146 4.4 Adiabatic Compression ................................................................................148 4.5 Polytropic Compression...............................................................................151 4.6 Discharge Temperature of Compressed Gas ...............................................152 4.7 Horsepower Required...................................................................................153 4.8 Optimum Compressor Locations .................................................................157 4.9 Compressors in Series and Parallel .............................................................163 4.10 Types of Compressors—Centrifugal and Positive Displacement ..............166 4.11 Compressor Performance Curves ................................................................168 4.12 Compressor Station Piping Losses ..............................................................171 4.13 Compressor Station Schematic ....................................................................172 4.14 Summary ......................................................................................................173 Problems ................................................................................................................174 References..............................................................................................................175 Chapter 5 Pipe Loops versus Compression 5.1 Purpose of a Pipe Loop ...............................................................................177 5.2 Purpose of Compression ..............................................................................178 5.3 Increasing Pipeline Capacity .......................................................................179 5.4 Reducing Power Requirements....................................................................189

Copyright 2005 by Taylor & Francis Group, LLC

2785_C000.fm Page ix Wednesday, April 13, 2005 10:38 AM

5.5 Looping in Distribution Piping....................................................................192 5.6 Summary ......................................................................................................198 Problems ................................................................................................................198 References..............................................................................................................199 Chapter 6 Pipe Analysis 6.1 Pipe Wall Thickness.....................................................................................201 6.2 Barlow’s Equation........................................................................................202 6.3 Thick-Walled Pipes ......................................................................................203 6.4 Derivation of Barlow’s Equation .................................................................205 6.5 Pipe Material and Grade ..............................................................................207 6.6 Internal Design Pressure Equation ..............................................................207 6.7 Class Location..............................................................................................209 6.8 Mainline Valves............................................................................................210 6.9 Hydrostatic Test Pressure ............................................................................211 6.10 Blowdown Calculations ...............................................................................242 6.11 Determining Pipe Tonnage ..........................................................................243 6.12 Summary ......................................................................................................246 Problems ................................................................................................................246 References..............................................................................................................247 Chapter 7 Thermal Hydraulics 7.1 Isothermal versus Thermal Hydraulics........................................................249 7.2 Temperature Variation and Gas Pipeline Modeling ....................................251 7.3 Review of Simulation Model Reports .........................................................253 7.4 Summary ......................................................................................................273 Problems ................................................................................................................274 References..............................................................................................................274 Chapter 8 Transient Analysis and Case Studies 8.1 Unsteady Flow .............................................................................................275 8.1.1 Transient Due to Mainline Valve Closure .......................................276 8.1.2 Transient Due to Compressor Shutdown.........................................277 8.2 Case Studies .................................................................................................279 8.2.1 Offshore Pipeline Case ....................................................................279 8.3 Summary ......................................................................................................296 Problems ................................................................................................................296 References..............................................................................................................297 Chapter 9 Valves and Flow Measurements 9.1 Purpose of Valves.........................................................................................299 9.2 Types of Valves ............................................................................................300 9.3 Material of Construction..............................................................................302 9.4 Codes for Design and Construction ............................................................302 9.5 Gate Valve ....................................................................................................303

Copyright 2005 by Taylor & Francis Group, LLC

9.6 9.7 9.8 9.9 9.10 9.11 9.12 9.13 9.14 9.15

Ball Valve .....................................................................................................305 Plug Valve ....................................................................................................305 Butterfly Valve .............................................................................................305 Globe Valve ..................................................................................................306 Check Valve..................................................................................................307 Pressure Control Valve.................................................................................308 Pressure Regulator .......................................................................................309 Pressure Relief Valve ...................................................................................309 Flow Measurement.......................................................................................310 Flow Meters .................................................................................................310 9.15.1 Orifice Meter ..................................................................................310 9.15.1.1 Meter Tube ....................................................................313 9.15.1.2 Expansion Factor ..........................................................314 9.16 Venturi Meter ...............................................................................................321 9.17 Flow Nozzle .................................................................................................323 9.18 Summary ......................................................................................................325 Problems ................................................................................................................325 References..............................................................................................................325 Chapter 10 Pipeline Economics 10.1 Components of Cost.....................................................................................328 10.2 Capital Costs ................................................................................................330 10.2.1 Pipeline...........................................................................................330 10.2.2 Compressor Stations.......................................................................332 10.2.3 Mainline Valve Stations .................................................................333 10.2.4 Meter Stations and Regulators.......................................................333 10.2.5 SCADA and Telecommunication System......................................333 10.2.6 Environmental and Permitting .......................................................334 10.2.7 Right of Way Acquisitions.............................................................334 10.2.8 Engineering and Construction Management .................................335 10.2.9 Other Project Costs ........................................................................335 10.3 Operating Costs............................................................................................336 10.4 Determining Economic Pipe Size................................................................339 10.5 Summary ......................................................................................................353 Problems ................................................................................................................353 References..............................................................................................................354 Appendix A Appendix B Appendix C Appendix D Appendix E

Units and Conversions ..................................................................355 Physical Properties of Various Gases ...........................................359 Pipe Properties—U.S. Customary System of Units.....................363 GASMOD Output Report .............................................................373 Summary of Formulas...................................................................379

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 1 Friday, April 1, 2005 3:46 PM

CHAPTER

1

Gas Properties In this chapter we will discuss the properties of gases that influence gas flow through a pipeline. We will explore the relationship of pressure, volume, and temperature of a gas and how the gas properties such as density, viscosity, and compressibility change with the temperature and pressure. Starting with the ideal or perfect gases that obey the ideal gas equation, we will examine how real gases differ from ideal gases. The concept of compressibility factor, or gas deviation factor, will be introduced and methods of calculating the compressibility factor using some popular graphical correlation and calculation methods explained. The properties of a mixture of gases will be discussed, and how these are calculated will be covered. Understanding the gas properties is an important first step toward analysis of gas pipeline hydraulics. A fluid can be a liquid or a gas. Liquids are generally considered almost incompressible. A gas is classified as a homogenous fluid with low density and viscosity. It expands to fill the vessel that contains the gas. The molecules that constitute the gas are spaced farther apart in comparison with a liquid and, therefore, a slight change in pressure affects the density of gas more than that of a liquid. Gases, therefore, have higher compressibility than liquids. This implies that gas properties such as density, viscosity, and compressibility factor change with pressure.

1.1 MASS AND WEIGHT Mass is the quantity of matter in a substance. It is sometimes used interchangeably with weight. Strictly speaking, mass is a scalar quantity, whereas weight is a force and, therefore, a vector quantity. Mass is independent of the geographic location, whereas weight depends upon the acceleration due to gravity and, therefore, varies slightly with geographic location. Mass is measured in slugs in the U.S. Customary System (USCS) of units and kilograms (kg) in Systeme International (SI) units.

1

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 2 Friday, April 1, 2005 3:46 PM

2

GAS PIPELINE HYDRAULICS

However, for most purposes, we say that a 10-lb mass has a weight of 10 lb. The pound (lb) is a more convenient unit for mass, and to distinguish between mass and weight, the terms pound mass (lbm) and pound force (lbf) are sometimes used. A slug is equal to approximately 32.2 lb. If some gas is contained in a certain volume and the temperature and pressure change, the mass will remain constant unless some gas is taken out or added to the container. This is known as the principle of conservation of mass. Weight is measured in pounds (lb) in USCS units and in Newton (N) in SI units. Sometimes we talk about mass flow rate through a pipeline rather than volume flow rate. Mass flow rate is measured in lb/hr in USCS units or kg/hr in SI units.

1.2 VOLUME Volume of a gas is the space a given mass of gas occupies at a particular temperature and pressure. Since gas is compressible, it will expand to fill the available space. Therefore, the gas volume will vary with temperature and pressure. Hence, a certain volume of a given mass of gas at some temperature and pressure will decrease in volume as the pressure is increased and vice versa. Suppose a quantity of gas is contained in a volume of 100 ft3 at a temperature of 80°F and a pressure of 200 psi. If the temperature is increased to 100°F, keeping the volume constant, the pressure will also increase. Similarly, if the temperature is reduced, gas pressure will also reduce, provided volume remains constant. Charles’s law states that for constant volume, the pressure of a fixed mass of gas will vary directly with the temperature. Thus, if temperature increases by 20%, the pressure will also rise by 20%. Similarly, if pressure is maintained constant, the volume will increase in direct proportion with temperature. Charles’s law, Boyle’s law, and other gas laws will be discussed in detail later in this chapter. Volume of gas is measured in ft3 in USCS units and m3 in SI units. Other units for volume include thousand ft3 (Mft3) and million ft3 (MMft3) in USCS units and thousand m3 (km3) and million m3 (Mm3) in SI units. When referred to standard conditions (also called base conditions) of temperature and pressure (60°F and 14.7 psia in USCS units), the volume is stated as standard volume and, therefore, measured in standard ft3 (SCF) or million standard ft3 (MMSCF). It must be noted that in the USCS units, the practice has been to use M to represent a thousand, and therefore MM refers to a million. This goes back to the Roman days of numerals, when M represented a thousand. In SI units, a more logical step is followed. For thousand, the letter k (for kilo) is used and the letter M (for Mega) is used for a million. Therefore, 500 MSCFD in USCS units refers to 500 thousand standard cubic feet per day (500,000 ft3/day), whereas 15 Mm3/day means 15 million cubic meters per day in SI units. This distinction in the use of the letter M to denote a thousand in USCS units and M for a million in SI units must be clearly understood. Volume flow rate of gas is measured per unit time and can be expressed as ft3/min, ft3/h, ft3/day, SCFD, MMSCFD, etc. in USCS units. In SI units, gas flow rate is expressed in m3/h or Mm3/day.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 3 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

3

1.3 DENSITY, SPECIFIC WEIGHT, AND SPECIFIC VOLUME Density represents the amount of gas that can be packed in a given volume. Therefore, it is measured in terms of mass per unit volume. If 5 lb of a gas is contained in 100 ft3 of volume at some temperature and pressure, we say that the gas density is 5/100 = 0.05 lb/ft3. Strictly speaking, in USCS units density must be expressed as slug/ft3 since mass is customarily referred to in slug. Thus,

ρ=

m V

(1.1)

where r = density of gas m = mass of gas V = volume of gas Density is expressed in slug/ft3 or lb/ft3 in USCS units and kg/m3 in SI units. A companion term called specific weight is also used when referring to the density of gas. Specific weight, represented by the symbol g, is the weight of gas per unit volume measured in lb/ft3 in USCS units, and is, therefore, contrasted with density, which is measured in slug/ft3. In SI units, the specific weight is expressed in Newton per m3 (N/m3). Quite often, density is also referred to in lb/ft3 in USCS units. The reciprocal of the specific weight is known as the specific volume. By definition, therefore, specific volume represents the volume occupied by a unit weight of gas. It is measured in ft3/lb in USCS units and m3/N in SI units. If the specific weight of a particular gas is 0.06 lb/ft3 at some temperature and pressure, its specific volume is 0.106 or 16.67 ft3/lb.

1.4 SPECIFIC GRAVITY Specific gravity of a gas, sometimes called gravity, is a measure of how heavy the gas is compared to air at a particular temperature. It might also be called relative density, expressed as the ratio of the gas density to the density of air. Because specific gravity is a ratio, it is a dimensionless quantity. G= where G = gas gravity, dimensionless rg = density of gas rair = density of air

Copyright 2005 by Taylor & Francis Group, LLC

ρg ρair

(1.2)

2785_C001.fm Page 4 Friday, April 1, 2005 3:46 PM

4

GAS PIPELINE HYDRAULICS

Both densities in Equation 1.2 must be in the same units and measured at the same temperature. For example, natural gas has a specific gravity of 0.60 (air = 1.00) at 60°F. This means that the gas is 60% as heavy as air. If we know the molecular weight of a particular gas, we can calculate its gravity by dividing the molecular weight by the molecular weight of air, as follows. G=

Mg Mg = M air 28.9625

(1.3)

Mg 29

(1.4)

or G=

Rounding off the molecular weight of air to 29 where G = specific gravity of gas Mg = molecular weight of gas Mair = molecular weight of air = 28.9625 Since natural gas consists of a mixture of several gases (methane, ethane, etc.), the molecular weight Mg in Equation 1.4 is referred to as the apparent molecular weight of the gas mixture. When the molecular weight and the percentage or mole fractions of the individual components of a natural gas mixture are known, we can calculate the molecular weight of the gas mixture by using a weighted average method. Thus, a natural gas mixture consisting of 90% methane, 8% ethane, and 2% propane will have a specific gravity of G=

(0.9 × M1) + (0.08 × M 2) + (0.02 × M 3) 29

where M1, M2, and M3 are the molecular weights of methane, ethane, and propane, respectively, and 29 represents the molecular weight of air. Table 1.1 lists the molecular weights and other properties of several hydrocarbon gases.

1.5 VISCOSITY The viscosity of a fluid represents its resistance to flow. The higher the viscosity, the more difficult it is to flow. Lower-viscosity fluids flow easily in pipes and cause less pressure drop. Liquids have much larger values of viscosity compared to gases. For example, water has a viscosity of 1.0 centiPoise (cP), whereas viscosity of natural gas is approximately 0.0008 cP. Even though the gas viscosity is a small number, it

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 5 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

5

has an important function in determining the type of flow in pipelines. The Reynolds number (discussed in Chapter 2) is a dimensionless parameter that is used to classify flow rate in pipelines. It depends on the gas viscosity, flow rate, pipe diameter, temperature, and pressure. The absolute viscosity, also called the dynamic viscosity, is expressed in lb/ft-s in USCS units and Poise (P) in SI units. A related term is the kinematic viscosity. This is simply the absolute viscosity divided by the density. The two viscosities are related as follows:

ν=

µ ρ

(1.5)

where, in USCS units, n = kinematic viscosity, ft2/s m = dynamic viscosity, lb/ft-s r = density, lb/ft3 and in SI units n = kinematic viscosity, St m = dynamic viscosity, P r = density, kg/m3 Kinematic viscosity is expressed in ft2/s in USCS units and Stokes (St) in SI units. Other units of viscosity in SI units include centipoise (cP) for dynamic viscosity and centistokes (cSt) for kinematic viscosity. Appendix A includes conversion factors for converting viscosity from one set of units to another. The viscosity of a gas depends on its temperature and pressure. Unlike liquids, the viscosity of a gas increases with increase in temperature. Since viscosity represents resistance to flow, as the gas temperature increases, the quantity of gas flow through a pipeline will decrease; hence, more throughput is possible in a gas pipeline at lower temperatures. This is in sharp contrast to liquid flow, where the throughput increases with temperature due to decrease in viscosity and vice versa. It must be noted that, unlike liquids, pressure also affects the viscosity of a gas. Like temperature, the gas viscosity increases with pressure. Figure 1.1 shows the variation of viscosity with temperature for a gas. Table 1.2 lists the viscosities of common gases. Since natural gas is a mixture of pure gases such as methane and ethane, the following formula is used to calculate the viscosity from the viscosities of component gases:

µ=

where m mi yi Mi

= = = =

(

Σ µi yi Mi

(

Σ yi Mi

)

)

dynamic viscosity of gas mixture dynamic viscosity of gas component i mole fraction or percent of gas component i molecular weight of gas component i

Copyright 2005 by Taylor & Francis Group, LLC

(1.6)

CH4 C 2H 6 C 3H 8 C4H10 C4H10 C5H12 C5H12 C5H12 C6H14 C6H14 C6H14 C6H14 C6H14 C7H16 C7H16 C7H16 C7H16 C7H16 C7H16 C7H16 C7H16 C8H18 C8H18 C8H18 C9H20 C10H22 C5H10 C6H12 C6H12

16.0430 30.0700 44.0970 58.1230 58.1230 72.1500 72.1500 72.1500 86.1770 86.1770 86.1770 86.1770 86.1770 100.2040 100.2040 100.2040 100.2040 100.2040 100.2040 100.2040 100.2040 114.2310 114.2310 114.2310 128.2580 142.2850 70.1340 84.1610 84.1610

5000 800 188.65 72.581 51.706 20.443 15.575 36.72 4.9596 6.769 6.103 9.859 7.406 1.621 2.273 2.13 2.012 3.494 3.294 2.775 3.376 0.5371 1.1020 1.7090 0.17155 0.06088 9.917 4.491 3.267

Methane Ethane Propane Isobutane n-butane Iso-pentane n-pentane Neo-pentane n-hexane 2-methyl pentane 3-methyl pentane Neo hexane 2,3-dimethylbutane n-Heptane 2-Methylhexane 3-Methylhexane 3-Ethylpentane 2,2-Dimethylpentane 2,4-Dimethylpentane 3,3-Dimethylpentane Triptane n-octane Di Isobutyl Isooctane n-Nonane n-Decane Cyclopentane Methylcyclopentane Cyclohexane

Copyright 2005 by Taylor & Francis Group, LLC

Critical Constants Pressure Temperature Volume °F ft3/lb psia 666.0 707.0 617.0 527.9 548.8 490.4 488.1 464.0 436.9 436.6 452.5 446.7 454.0 396.8 396.0 407.6 419.2 401.8 397.4 427.9 427.9 360.7 361.1 372.7 330.7 304.6 653.8 548.8 590.7

−116.66 90.07 205.93 274.4 305.52 368.96 385.7 321.01 453.8 435.76 448.2 419.92 440.08 512.8 494.44 503.62 513.16 476.98 475.72 505.6 496.24 564.15 530.26 519.28 610.72 652.1 461.1 499.28 536.6

0.0988 0.0783 0.0727 0.0714 0.0703 0.0684 0.0695 0.0673 0.0688 0.0682 0.0682 0.0667 0.0665 0.0682 0.0673 0.0646 0.0665 0.0665 0.0667 0.0662 0.0636 0.0673 0.0676 0.0657 0.0693 0.0702 0.0594 0.0607 0.0586

Ideal Gas 14.696 psia, 60°F Spgr (air=1.00) ft3/lb-gas 0.5539 1.0382 1.5226 2.0068 2.0068 2.4912 2.4912 2.4912 2.9755 2.9755 2.9755 2.9755 2.9755 3.4598 3.4598 3.4598 3.4598 3.4598 3.4598 3.4598 3.4598 3.9441 3.9441 3.9441 4.4284 4.9127 2.4215 2.9059 2.9059

23.654 12.620 8.6059 6.5291 6.5291 5.2596 5.2596 5.2596 4.4035 4.4035 4.4035 4.4035 4.4035 3.7872 3.7872 3.7872 3.7872 3.7872 3.7872 3.7872 3.7872 3.322 3.322 3.322 2.9588 2.6671 5.411 4.509 4.509

Specific Heat, Btu/lb/°F 14.696 psia, 60°F Ideal Gas 0.52676 0.40789 0.38847 0.38669 0.39500 0.38448 0.38831 0.39038 0.38631 0.38526 0.37902 0.38231 0.37762 0.38449 0.38170 0.37882 0.38646 0.38651 0.39627 0.38306 0.37724 0.38334 0.37571 0.38222 0.38248 0.38181 0.27122 0.30027 0.29012

GAS PIPELINE HYDRAULICS

Formula

Molecular Weight

Vapor Pressure psia at 100°F

2785_C001.fm Page 6 Friday, April 1, 2005 3:46 PM

Gas

6

Table 1.1 Properties of Hydrocarbon Gases

1.609 1400 232.8 62.55 45.97 49.88 64.95 19.12 36.53 59.46 16.68 3.225 1.033 0.3716 0.2643 0.3265 0.3424 0.2582 0.188 4.631 2.313

394.59 85.46 211.9

157.3 0.95 906.71

503.4 731.0 676.6 586.4 615.4 574.9 580.2 509.5 656.0 620.3 582.0 890.4 710.4 595.5 523 541.6 512.9 509.2 587.8 465.4 1174 891.7 506.8 1071 1306 1143 1647 546.9 187.5 731.4 493 1157 3200.1 32.99 1205

570.2 48.54 198.31 296.18 324.31 311.8 292.49 376.86 354 306 403 95.29 552.15 605.5 651.22 674.85 650.95 649.47 703 676.2 463.01 465.31 −220.51 87.73 212.4 315.7 270.2 −221.29 −400.3 −181.4 −232.48 290.69 705.1 −450.31 124.75

0.0600 0.0746 0.0717 0.0683 0.0667 0.0679 0.0681 0.0674 0.0700 0.0653 0.0660 0.0693 0.0531 0.0549 0.0564 0.0557 0.0567 0.0572 0.0534 0.0569 0.0590 0.0581 0.0527 0.0342 0.0461 0.0305 0.0681 0.0517 0.5101 0.0367 0.0510 0.0280 0.04975 0.2300 0.0356

3.3902 0.9686 1.4529 1.9373 1.9373 1.9373 1.9373 2.4215 1.8677 1.8677 2.3520 0.8990 2.6971 3.1814 3.6657 3.6657 3.6657 3.6657 3.5961 4.1500 1.1063 1.5906 0.9671 1.5196 1.1768 2.2120 0.5880 1.0000 0.06960 1.1048 0.9672 2.4482 0.62202 0.1382 1.2589

3.8649 13.527 9.0179 6.7636 6.7636 6.7636 6.7636 5.411 7.0156 7.0156 5.571 14.574 4.8581 4.1184 3.5744 3.5744 3.5744 3.5744 3.6435 3.1573 11.843 8.2372 13.548 8.6229 11.134 5.9235 22.283 13.103 188.25 11.859 13.546 5.3519 21.065 94.814 10.408

0.31902 0.35789 0.35683 0.35535 0.33275 0.35574 0.36636 0.35944 0.34347 0.34223 0.35072 0.39754 0.24295 0.26005 0.27768 0.28964 0.27427 0.27470 0.26682 0.30704 0.32429 0.33074 0.24847 0.19909 0.23838 0.14802 0.49678 0.2398 3.4066 0.21897 0.24833 0.11375 0.44469 1.24040 0.19086

2785_C001.fm Page 7 Friday, April 1, 2005 3:46 PM

98.1880 28.0540 42.0810 56.1080 56.1080 56.1080 56.1080 70.1340 54.0920 54.0920 68.1190 26.0380 78.1140 92.1410 106.1670 106.1670 106.1670 106.1670 104.1520 120.1940 32.0420 46.0690 28.0100 44.0100 34.0820 64.0650 17.0305 28.9625 2.0159 31.9988 28.0134 70.9054 18.0153 4.0026 36.4606

7

Copyright 2005 by Taylor & Francis Group, LLC

C7H14 C 2H 4 C 3H 6 C 4H 8 C 4H 8 C 4H 8 C 4H 8 C5H10 C 4H 6 C 4H 6 C 5H 8 C 2H 2 C 6H 6 C 7H 8 C8H10 C8H10 C8H10 C8H10 C 8H 8 C9H12 CH4O C 2H 6O CO CO2 H 2S SO2 NH3 N2+O2 H2 O2 N2 Cl2 H 2O He HCl

GAS PROPERTIES

Methylcyclohexane Ethylene Propylene Butylene Cis-2-butene Trans-2-butene Isobutene 1-Pentene 1,2-Butadene 1,3-Butadene Isoprene Acetylene Benzene Toluene Ethyl-benzene o-Xylene m-Xylene p-Xylene Styrene Isopropylbenzene Methyl alcohol Ethyl alcohol Carbon monoxide Carbon dioxide Hydrogen sulfide Sulfur dioxide Ammonia Air Hydrogen Oxygen Nitrogen Chlorine Water Helium Hydrogen chloride

2785_C001.fm Page 8 Friday, April 1, 2005 3:46 PM

8

GAS PIPELINE HYDRAULICS

0.024

m

Heliu Air

0.022

Viscosity, cP

0.020

e

ioxid

on d

Carb

0.018 0.016

ne Methane Ethyle e Ethan Propannee i-Butatane n-Bu ntane n-Pe

0.014 0.012 0.010 0.008 0.006 0.004 50

100

150

200

250

300

350

Temperature,°F Figure 1.1

Variation of gas viscosity with temperature.

Therefore, a homogeneous mixture that consists of 20% of a gas A (molecular weight = 18) that has a viscosity 6 × 10−6 Poise and 80% of a gas B (molecular weight = 17) that has a viscosity 8 × 10−6 Poise will have a resultant viscosity of

µ=

(0.2 × 6 × 18 ) + (0.8 × 8 × 17 ) (0.2 × 18 ) + (0.8 × 17 )

× 10 −6 = 7.59 × 10−6 Poise

Table 1.2 Viscosities of Common Gases

Copyright 2005 by Taylor & Francis Group, LLC

Gas

Viscosity (cP)

Methane Ethane Propane i-Butane n-Butane i-Pentane n-Pentane Hexane Heptane Octane Nonane Decane Ethylene Carbon Monoxide Carbon Dioxide Hydrogen Sulphide Air Nitrogen Helium

0.0107 0.0089 0.0075 0.0071 0.0073 0.0066 0.0066 0.0063 0.0059 0.0050 0.0048 0.0045 0.0098 0.0184 0.0147 0.0122 0.0178 0.0173 0.0193

2785_C001.fm Page 9 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

9

It must be noted that all viscosities must be measured at the same temperature and pressure. The reader is referred to W. McCain’s book for further discussion on calculation of viscosities of natural gas mixtures. See the Reference section for more details. Example 1 A natural gas mixture consists of four components C1, C2, C3, and nC4. Their mole fractions and viscosities at a particular temperature and pressure are indicated below, along with their molecular weights. Component

Mole Fraction y

Viscosity, cP

Molecular Weight

C1 C2 C3 nC4

0.8200 0.1000 0.0500 0.0300

0.0130 0.0112 0.0098 0.0091

16.04 30.07 44.10 58.12

Total

1.000

Calculate the viscosity of the gas mixture. Solution Using the given data, we prepare a table as follows. M represents the molecular weight of each component and m the viscosity. Component

y

M

M 1/2

yM 1/2

yM 1/2

C1 C2 C3 nC4

0.8200 0.1000 0.0500 0.0300

16.04 30.07 44.10 58.12

4.00 5.48 6.64 7.62

3.2841 0.5484 0.3320 0.2287

0.0130 0.0112 0.0098 0.0091

0.0427 0.0061 0.0033 0.0021

Total

1.000

4.3932

0.0542

From Equation 1.6, the viscosity of the gas mixture is

µ=

0.0542 = 0.0123 cP 4.3932

1.6 IDEAL GASES An ideal gas is defined as a fluid in which the volume of the gas molecules is negligible when compared to the volume occupied by the gas. Also, the attraction or repulsion between the individual gas molecules and the container is negligible. Further, in an ideal gas, the molecules are considered to be perfectly elastic, and there is no internal energy loss resulting from collision between the molecules. Such ideal gases are said

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 10 Friday, April 1, 2005 3:46 PM

10

GAS PIPELINE HYDRAULICS

to obey several gas laws, such as Boyle’s law, Charles’s law, and the ideal gas law or the perfect gas equation. We will first discuss the behavior of ideal gases and then follow it up with the behavior of real gases. If M represents the molecular weight of a gas and the mass of a certain quantity of gas is m, the number of moles is given by n=

m M

(1.7)

where n is the number that represents the number of moles in the given mass. For example, the molecular weight of methane is 16.043. Therefore, 50 lb of methane will contain approximately 3 moles. The ideal gas law, sometimes referred to as the perfect gas equation, simply states that the pressure, volume, and temperature of the gas are related to the number of moles by the following equation: PV = nRT where P = V = n = R = T =

(USCS units)

(1.8)

absolute pressure, pounds per square inch absolute (psia) gas volume, ft3 number of lb moles as defined in Equation 1.7 universal gas constant, psia ft3/lb mole °R absolute temperature of gas, °R (°F + 460)

The universal gas constant R has a value of 10.73 psia ft3/lb mole °R in USCS units. We can combine Equation 1.7 with Equation 1.8 and express the ideal gas equation as follows: PV =

mRT M

(1.9)

All symbols are as defined previously. It should be noted that the constant R is the same for all ideal gases and, hence, it is called the universal gas constant. It has been found that the ideal gas equation is correct only at low pressures close to the atmospheric pressure. Since gas pipelines generally operate at pressures higher than atmospheric pressures, we must modify Equation 1.9 to take into account the effect of compressibility. The latter is accounted for by using a term called the compressibility factor, or gas deviation factor. We will discuss real gases and the compressibility factor later in this chapter. In the ideal gas Equation 1.9, the pressures and temperatures must be in absolute units. Absolute pressure is defined as the gauge pressure (as measured by a gauge) plus the local atmospheric pressure. Therefore, Pabs = Pgauge + Patm

Copyright 2005 by Taylor & Francis Group, LLC

(1.10)

2785_C001.fm Page 11 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

11

Thus, if the gas pressure is 200 psig and the atmospheric pressure is 14.7 psia, we get the absolute pressure of the gas as Pabs = 200 + 14.7 = 214.7 psia Absolute pressure is expressed as psia, whereas the gauge pressure is expressed as psig. The adder to the gauge pressure, which is the local atmospheric pressure, is also called the base pressure. In SI units, 500 kPa gauge pressure is equal to 601 kPa absolute pressure if the base pressure is 101 kPa. Pressure in USCS units is expressed in pounds per square inch, or psi. In SI units, pressure is expressed in kilopascal (kPa), megapascal (MPa), or Bar. Refer to Appendix A for unit conversion charts. The absolute temperature is measured above a certain datum. In USCS units, the absolute scale of temperature is designated as degree Rankin (°R) and is equal to the sum of the temperature in °F and the constant 460. In SI units the absolute temperature scale is referred to as degree Kelvin (K). Absolute temperature in K is equal to °C + 273. Therefore, °R = °F + 460

(1.11)

K = °C + 273

(1.12)

It is customary to drop the degree symbol for absolute temperature in Kelvin. Ideal gases also obey Boyle’s law and Charles’s law. Boyle’s law relates the pressure and volume of a given quantity of gas when the temperature is kept constant. Constant temperature is also called isothermal condition. Boyle’s law is stated as follows: P1 V2 = P2 V1

or P1V1 = P2V2

(1.13)

where P1 and V1 are the pressure and volume at condition 1 and P2 and V2 are the corresponding values at some other condition 2 where the temperature is not changed. Therefore, according to Boyle’s law, a given quantity of gas under isothermal conditions will double in volume if its pressure is halved and vice versa. In other words, the pressure is inversely proportional to the volume, provided the temperature is maintained constant. Since density and volume are inversely related, Boyle’s law also means that the pressure is directly proportional to the density at constant temperature. Thus, a given quantity of gas at a fixed temperature will double its density when the pressure is doubled. Similarly, a 10% reduction in pressure will cause the density to decrease by the same amount. Charles’s law states that for constant pressure, the gas volume is directly proportional to its temperature. Similarly, if volume is kept constant, the pressure varies directly as the temperature. Therefore, we can state the following:

Copyright 2005 by Taylor & Francis Group, LLC

V1 T1 = V2 T2

at constant pressure

(1.14)

P1 T1 = P2 T2

at constant volume

(1.15)

2785_C001.fm Page 12 Friday, April 1, 2005 3:46 PM

12

GAS PIPELINE HYDRAULICS

Therefore, according to Charles’s law, for an ideal gas at constant pressure, the volume will change in the same proportion as its temperature. Thus, a 20% increase in temperature will cause a 20% increase in volume as long as the pressure does not change. Similarly, if volume is kept constant, a 20% increase in temperature will result in the same percentage increase in gas pressure. Constant pressure is also known as isobaric condition. Example 2 A certain mass of gas occupies a volume of 1000 ft3 at 60 psig. If temperature is constant (isothermal) and the pressure increases to 120 psig, what is the final volume of the gas? The atmospheric pressure is 14.7 psi. Solution Boyle’s law can be applied because the temperature is constant. Using Equation 1.13, we can write V2 =

PV 1 1 P2

or V2 =

(60 + 14.7) × 1000 = 554.57 ft3 120 + 14.7

Note that the pressures must be converted to absolute values before being used in Equation 1.13. Example 3 At 75 psig and 70°F, a gas has a volume of 1000 ft3. If the volume is kept constant and the gas temperature increases to 120°F, what is the final pressure of the gas? For constant pressure at 75 psig, if the temperature increases to 120°F, what is the final volume? Use 14.7 psi for base pressure. Solution Since the volume is constant in the first part of the problem, Charles’s law applies. From Equation 1.15 75 + 14.7 70 + 460 = P2 120 + 460 Solving for P2, we get P2 = 98.16 psia or 83.46 psig

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 13 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

13

For the second part, the pressure is constant and Charles’s law can be applied. From Equation 1.14, we get 1000 70 + 460 = 120 + 460 V2 Solving for V2 we get V2 = 1094.34 ft3 Example 4 An ideal gas occupies a tank volume of 250 ft3 at a pressure of 80 psig and temperature of 110°F. (1) What is the gas volume at standard conditions of 14.73 psia and 60°F? Assume atmospheric pressure is 14.6 psia. (2) If the gas is cooled to 90°F, what is the gas pressure? Solution (1) Initial conditions P1 = 80 + 14.6 = 94.6 psia V1 = 250 ft3 T1 = 110 + 460 = 570°R Final conditions P2 = 14.73 psia V2 is to be calculated T2 = 60 + 460 = 520°R Using the ideal gas Equation 1.8, we can state that 94.6 × 250 14.73V2 = 570 520 V2 = 1464.73 ft3 (2) When the gas is cooled to 90°F, the final conditions are T2 = 90 + 460 = 550°R V2 = 250 ft3 P2 is to be calculated

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 14 Friday, April 1, 2005 3:46 PM

14

GAS PIPELINE HYDRAULICS

The initial conditions are P1 = 80 + 14.6 = 94.6 psia V1 = 250 ft3 T1 = 110 + 460 = 570°R It can be seen that the volume of gas is constant (tank volume) and the temperature reduces from 110°F to 90°F. Therefore, using Charles’s law and Equation 1.15, we can calculate the final pressure as follows. 94.6 570 = P2 550 Solving for P2, we get P2 = 91.28 psia = 91.28 – 14.6 = 76.68 psig

1.7 REAL GASES When dealing with real gases, we can apply the ideal gas equation discussed in the preceding sections and get reasonably accurate results only when the pressures are close to the atmospheric pressure. When pressures are higher, the ideal gas equation will not be accurate for most real gases. The error in calculations at high pressures using the ideal gas equation may be as high as 500% in some instances. This compares with errors of 2 to 3% at low pressures. At higher temperatures and pressures, the “equation of state” that relates pressure, volume, and temperature is used to calculate the properties of gases. Many of these correlations require a computer program to get accurate results in a reasonable amount of time. However, we can modify the ideal gas equation and obtain reasonably accurate results fairly quickly using manual calculations. Two terms called critical temperature and critical pressure need to be defined. The critical temperature of a pure gas is defined as the temperature above which a gas cannot be compressed to form a liquid, regardless of the pressure. The critical pressure is defined as the minimum pressure that is required at the critical temperature to compress a gas into a liquid. Real gases can be considered to follow a modified form of the ideal gas law discussed in Section 1.6. The modifying factor is included in the gas property known as the compressibility factor Z. This is also called the gas deviation factor. It can be defined as the ratio of the gas volume at a given temperature and pressure to the volume the gas would occupy if it were an ideal gas at the same temperature and pressure. Z is a dimensionless number less than 1.0 and it varies with temperature, pressure, and composition of the gas. Using the compressibility factor Z, the ideal gas equation is modified for real gas as follows: PV = ZnRT

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(1.16)

2785_C001.fm Page 15 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

where P = V = Z = T = n = R =

15

absolute pressure of gas, psia volume of gas, ft3 gas compressibility factor, dimensionless absolute temperature of gas, °R number of lb moles as defined in Equation 1.7 universal gas constant, 10.73 psia ft3/lb mole °R

The theorem known as corresponding states says that the extent of deviation of a real gas from the ideal gas equation is the same for all real gases when the gases are at the same corresponding state. The corresponding state can be represented by the two parameters called reduced temperature and reduced pressure. The reduced temperature is the ratio of the temperature of the gas to its critical temperature. Similarly, the reduced pressure is the ratio of the gas pressure to its critical pressure as indicated in the following equations:

where P = T = Tr = Pr = Tc = Pc =

Tr =

T Tc

(1.17)

Pr =

P Pc

(1.18)

absolute pressure of gas, psia absolute temperature of gas, °R reduced temperature, dimensionless reduced pressure, dimensionless critical temperature, °R critical pressure, psia

For example, suppose the critical temperature and critical pressure of methane are 343°R and 666 psia, respectively; the reduced temperature and pressure of the gas at 80°F and 1000 psia pressure are as follows: Tr =

80 + 460 = 1.57 343

and Pr =

1000 = 1.50 666

Therefore, according to the theorem of corresponding states, two gases, A and B, may be at different temperatures and pressures; however, if their reduced temperature and reduced pressure are the same, then their gas deviation factors (Z ) will be the same. Therefore, generalized plots showing the variation of Z with reduced temperature and reduced pressure can be used for most gases for calculating the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 16 Friday, April 1, 2005 3:46 PM

16

GAS PIPELINE HYDRAULICS

Pseudo-reduced pressure, Pr

0.8

1.7

1.2 1.6 3 1.

1.3

1.4 1.5 1.5 1.6 1.7 1.81.9 1.4 2.0 2.2

1.25 1.2

0.5 1.15 0.4

1.1

1.3 1.1 1.2 0.95

1.1

1.4 1.35

0.6

8

1.0 1.05

1.5 1.45

0.7 Compressibility factor, Z

7

1.1

2.4 2.6 3.0

Compressibility factor, Z

0.9

6

5

1.0

5

1. 05

3 4 Pseudo reduced temperature 3.0 2.8 2.6 2.4 2.2 2.0 1.9 1.8 1.7 1.6

1.

2

4

1

1.

1.1

0

1.3

1.2

0.3 1.05 0.25

3.0 1.1 2.8 2.6 2.4 2.2 2.0 1.9 1.0 1.8 1.7 1.6 1.4 1.3 0.9 7 8

Figure 1.2

1.1 MW < 40 1.2 1.1 1.05 9

Compressibility of natural gases Jan. 1, 1941

10 11 12 Pseudo reduced pressure, Pr

13

14

1.0

0.9 15

Compressibility factor chart for natural gases. (From Gas Processors Suppliers Association, Eng. Data Book, Vol. II. With permission.)

compressibility factor. Such a plot is shown in Figure 1.2. The calculation of the compressibility factor Z will be discussed in detail in Section 1.11 of this chapter. 1.8 NATURAL GAS MIXTURES As mentioned earlier, the critical temperature of a pure gas is defined as the temperature above which it cannot be liquefied, whatever the pressure of the gas. Similarly, the critical pressure is defined as the pressure above which liquid and gas cannot

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 17 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

17

coexist, regardless of the temperature. When the gas consists of a mixture of different components, the critical temperature and critical pressure are called the pseudocritical temperature and pseudo-critical pressure, respectively. If we know the composition of the gas mixture, we can calculate these pseudo-critical values of the mixture, using the critical pressure and temperature values of the pure components that constitute the gas mixture. The reduced temperature is defined as the ratio of the temperature of the gas to its critical temperature. Similarly, the reduced pressure is the ratio of gas pressure to its critical pressure. Both temperature and pressure are stated in absolute units. Similar to the pseudo-critical temperature and pseudo-critical pressure discussed above, for a gas mixture, we can define the pseudo-reduced temperature and the pseudo-reduced pressure. Thus,

where P T Tpr Ppr Tpc Ppc

= = = = = =

Tpr =

T Tpc

(1.19)

Ppr =

P Ppc

(1.20)

absolute pressure of gas mixture, psia absolute temperature of gas mixture, °R pseudo-reduced temperature, dimensionless pseudo-reduced pressure, dimensionless pseudo-critical temperature, °R pseudo-critical pressure, psia

In hydrocarbon mixtures, frequently we refer to gas components as C1, C2, C3, etc. These are equivalent to CH4 (methane), C2H6 (ethane), C3H8 (propane), and so on. A natural gas mixture that consists of components such as C1, C2, C3, and so forth is said to have an apparent molecular weight as defined by the equation M a = Σyi Mi

(1.21)

where Ma = apparent molecular weight of gas mixture yi = mole fraction of gas component i Mi = molecular weight of gas component i In a similar manner, from the given mole fractions of the gas components, we use Kay’s rule to calculate the average pseudo-critical properties of the gas mixture.

Copyright 2005 by Taylor & Francis Group, LLC

Tpc = ΣyiTc

(1.22)

Ppc = Σyi Pc

(1.23)

2785_C001.fm Page 18 Friday, April 1, 2005 3:46 PM

18

GAS PIPELINE HYDRAULICS

where Tc and Pc are the critical temperature and pressure, respectively, of the pure component (C1, C2, etc.) and yi refers to the mole fraction of the component. Tpc and Ppc are the average pseudo-critical temperature and pseudo-critical pressure, respectively, of the gas mixture. Example 5 Calculate the apparent molecular weight of a natural gas mixture that has 85% methane, 9% ethane, 4% propane, and 2% normal butane as shown below: Component

Percent

Molecular Weight

C1 C2 C3 n-C4

85 9 4 2

16.01 30.10 44.10 58.10

Total

100

Solution Using Equation 1.21, we get M a = (0.85 × 16.01) + (0.09 × 30.1) + (0.04 × 44.1) + (0.02 × 58.1) = 19.24 Therefore, the apparent molecular weight of the gas mixture is 19.24. Example 6 Calculate the pseudo-critical temperature and the pseudo-critical pressure of a natural gas mixture consisting of 83% methane, 12% ethane, and 5% propane. The critical properties of C1, C2, and C3 components are as follows:

Components

Critical Temperature, °R

Critical Pressure, psia

C1 C2 C3

343 550 666

666 707 617

Solution Using the given data, from Equation 1.22 and Equation 1.23, we calculate the pseudocritical properties as follows: Tpc = (0.83 × 343) + (0.12 × 550) + (0.05 × 666) = 383.99°R

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 19 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

19

and Ppc = (0.83 × 666) + (0.12 × 707) + (0.05 × 617) = 668.47 psia Therefore, the pseudo-critical temperature of the gas mixture = 383.99°R and the pseudo-critical pressure of the gas mixture = 668.47 psia. Example 7 If the temperature of the gas in the previous example is 70°F and the average gas pressure is 1200 psig, what are the pseudo-reduced temperature and pseudo-reduced pressure of this gas? Use 14.7 psia for base pressure. Solutions From Equation 1.19 and Equation 1.20, we get

The pseudo-reduced temperature Tpr = The pseudo-reduced pressure Ppr =

70 + 460 = 1.38 383.99

1200 + 14.7 = 1.82 668.47

1.9 PSEUDO-CRITICAL PROPERTIES FROM GAS GRAVITY If the percentages of the various components in the natural gas mixture are not available, we can calculate approximate values of the pseudo-critical properties of the gas mixture if we know the gas gravity. The pseudo-critical properties are calculated, approximately, from the following equations: Tpc = 170.491 + 307.344 G

(1.24)

Ppc = 709.604 – 58.718 G

(1.25)

where G = gas gravity (air = 1.00) Tpc = pseudo-critical temperature, °R Ppc = pseudo-critical pressure, psia Example 8 Calculate the gravity of a natural gas mixture consisting of 83% methane, 12% ethane, and 5% propane. From the gas gravity, calculate the pseudo-critical temperature and pseudo-critical pressure for this natural gas mixture.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 20 Friday, April 1, 2005 3:46 PM

20

GAS PIPELINE HYDRAULICS

Solution Using Kay’s rule for multicomponent mixtures and Equation 1.4 for gas gravity, we get G=

(0.83 × 16.04) + (0.12 × 30.07) + (0.05 × 44.10) = 0.6595 29

Therefore, the gas gravity is 0.6595. From Equation 1.24 and Equation 1.25, we calculate the pseudo-critical properties as follows: Tpc = 170.491 + 307.344 × (0.6595) = 373.18°R Ppc = 709.604 – 58.718 × (0.6595) = 670.88 psia Comparing the above values with the values calculated using the more accurate method in the previous example, we find that the value of Tpc is off by 2.8% and Ppc is off by 0.4%. These differences are small enough for most calculations related to natural gas pipeline hydraulics.

1.10 IMPACT OF SOUR GAS AND NON-HYDROCARBON COMPONENTS The Standing-Katz chart used for determining the compressibility factor (discussed in Section 1.11) of a gas mixture is accurate only if the amount of non-hydrocarbon components is small. Since sour gases contain carbon dioxide and hydrogen sulfide, adjustments must be made to take into account these components in calculations of the pseudo-critical temperature and pseudo-critical pressure. This method is described below. Depending on the amounts of carbon dioxide and hydrogen sulfide present in the sour gas, we calculate an adjustment factor e from e = 120(A0.9 − A1.6) + 15(B 0.5 – B 4.0)

(1.26)

where e = adjustment factor, °R A = sum of the mole fractions of CO2 and H2S B = mole fraction of H2S The pseudo-critical temperature is modified to get the adjusted pseudo-critical temperature T pc ′ from the following equation: T pc ′ = Tpc – e where T′pc = adjusted pseudo-critical temperature, °R

Copyright 2005 by Taylor & Francis Group, LLC

(1.27)

2785_C001.fm Page 21 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

21

Similarly, the pseudo-critical pressure is adjusted as follows: P′pc =

Ppc × Tpc′ Tpc + B(1 − B)ε

(1.28)

where P′pc = adjusted pseudo-critical pressure, psia. Example 9 The pseudo-critical temperature and the pseudo-critical pressure of a natural gas mixture were calculated as 370°R and 670 psia, respectively. If the CO2 content is 10% and H2S is 20%, calculate the adjustment factor e and the adjusted values of the pseudo-critical temperature and pressure. Solution A = 0.10 + 0.20 = 0.30 B = 0.20 The adjustment factor e from Equation 1.26 is e = 120 (0.30

0.9

– 0.301.6) + 15 (0.200.5 – 0.204) = 29.8082°R

Therefore, the adjustment factor e is 29.81°R. The adjusted values of the pseudo-critical properties are found using Equation 1.27 and Equation 1.28 as follows: T′pc = 370 – 29.81 = 340.19°R and P′pc =

670 × 340.19 = 608.18 psia 370 + 0.20(1 − 0.20) × 29.8082

Therefore, the adjusted pseudo-critical temperature = 340.19°R and the adjusted pseudo-critical pressure = 608.18 psia.

1.11 COMPRESSIBILITY FACTOR The compressibility factor, or gas deviation factor, was briefly mentioned in Section 1.6. It is a measure of how close a real gas is to an ideal gas. The compressibility factor is defined as the ratio of the gas volume at a given temperature and pressure to the volume the gas would occupy if it were an ideal gas at the same temperature and pressure. The compressibility factor is a dimensionless number close to 1.00 and is

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 22 Friday, April 1, 2005 3:46 PM

22

GAS PIPELINE HYDRAULICS

a function of the gas gravity, gas temperature, gas pressure, and the critical properties of the gas. As an example, a particular natural gas mixture can have a compressibility factor equal to 0.87 at 1000 psia and 80°F. Charts have been constructed that depict the variation of Z with the reduced temperature and reduced pressure. Another term, the “supercompressibility factor,” Fpv , which is related to the compressibility factor Z, is defined as follows: Fpv =

1 Z

(1.29)

or Z=

1 ( Fpv )2

(1.30)

As an example, if the compressibility factor Z = 0.85, using Equation 1.29, we calculate the supercompressibility factor, Fpv , as follows: Fpv =

1 0.85

= 1.0847

There are several approaches to calculating the compressibility factor for a particular gas temperature T and pressure P. One method uses the critical temperature and critical pressure of the gas mixture. First, the reduced temperature, Tr, and reduced pressure, Pr, are calculated from the given gas temperature and gas pressure and the critical temperature and critical pressure using Equation 1.17 and Equation 1.18. Once we know the values of Tr and Pr, the compressibility factor can be found from charts similar to the Standing-Katz chart. This will be illustrated using an example. The following methods are available to calculate the compressibility factor: a. b. c. d.

Standing-Katz method Dranchuk, Purvis, and Robinson method AGA method CNGA method

Although the Standing-Katz method is the most popular, we will discuss this as well as the AGA and CNGA methods. 1.11.1

Standing-Katz Method

The Standing-Katz method of calculating compressibility factor is based on the use of a graph that has been constructed for binary mixtures and saturated hydrocarbon vapor. This method is used generally for sweet natural gas mixtures containing various hydrocarbon components. When the natural gas mixture contains appreciable amounts of non-hydrocarbons such as nitrogen, hydrogen sulfide, and carbon dioxide, certain corrections must be applied for these components. These adjustments are applied to

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 23 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

23

the critical temperatures and pressures and were discussed in Section 1.10. The Standing-Katz chart for compressibility factors is shown in Figure 1.2. The use of this chart will be explained in an example problem. 1.11.2

Dranchuk, Purvis, and Robinson Method

In this method of calculating the compressibility factor, the Benedict-Webb-Rubin equation of state is used to correlate the Standing-Katz chart. The coefficients A1, A2, etc. are used in a polynomial function of the reduced density ρr as follows: A A A A ρ5 A7 ρr3 A Z = 1 + A1 + 2 + 33 ρr + A4 + 5 ρr2 + 5 6 r + (− A ρ2 ) Tpr Tpr Tpr Tpr Tpr3 1 + A8 ρr2 e 8 r

(

)

(1.31) where

ρr = and A1 A2 A3 A4 A5 A6 A7 A8 Ppr Tpr

= = = = = = = = = =

0.27 Ppr ZTpr

(1.32)

0.31506237 −1.04670990 −0.57832729 0.53530771 −0.61232032 −0.10488813 0.68157001 0.68446549 pseudo-reduced pressure pseudo-reduced temperature

Other symbols have been defined earlier. 1.11.3

American Gas Association (AGA) Method

The AGA method for the compressibility factor uses a complicated mathematical algorithm and, therefore, does not lend itself easily to manual calculations. Generally, a computer program is used to calculate the compressibility factor. Mathematically, the AGA method is represented by the following function: Z = Function (gas properties, pressure, temperature)

(1.33)

where gas properties include the critical temperature, critical pressure, and gas gravity.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 24 Friday, April 1, 2005 3:46 PM

24

GAS PIPELINE HYDRAULICS

The AGA-IGT Report No. 10 describes in detail this method of calculating Z. This approach is valid for gas temperatures in the range of 30 to 120°F and for pressures not exceeding 1380 psig. The compressibility factors calculated using this method are quite accurate and generally within 0.03% of those calculated using the Standing-Katz chart in this range of temperatures and pressures. When temperatures and pressures are higher than these values, the compressibility factor calculated using this method is within 0.07% of the value obtained from the Standing-Katz chart. The reader may also refer to the AGA publication Report No. 8, Second Edition, November 1992, for more information on compressibility factor calculation methods. 1.11.4

California Natural Gas Association (CNGA) Method

This is a fairly simple equation for quickly calculating the compressibility factor when the gas gravity, temperature, and pressure are known. The following equation is used for calculating the compressibility factor Z: Z=

1 Pavg 344 ,400 (10 )1.785G 1 + T f3.825

(1.34)

This formula for the compressibility factor is valid when the average gas pressure, Pavg, is more than 100 psig. For pressures less than 100 psig, Z is approximately equal to 1.00 where Pavg = average gas pressure, psig Tf = average gas temperature, °R G = gas gravity (air = 1.00) Note that the pressure used in Equation 1.34 is the gauge pressure. In a gas pipeline, the pressure varies along the length of the pipeline. The compressibility factor Z also varies and must therefore be calculated for an average pressure at any location on the pipeline. If two points along the pipeline are at pressures P1 P +P and P2, we could use an average pressure of 1 2 2 . However, the following formula is used for a more accurate value of the average pressure: Pavg =

2 P ×P P1 + P2 − 1 2 3 P1 + P2

(1.35)

Another form of the average pressure in a pipe segment is

Pavg =

Copyright 2005 by Taylor & Francis Group, LLC

2 P13 − P23 3 P12 − P22

(1.36)

2785_C001.fm Page 25 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

25

Example 10 Using the Standing-Katz compressibility chart, calculate the compressibility factor for the gas in Example 7 at 70°F and 1200 psig. Use the values of Tpc and Ppc calculated in Example 7. Solutions From previous Example 7, The pseudo-reduced temperature Tpr = 1.38 The pseudo-reduced pressure Ppr = 1.82 Using the Standing-Katz chart in Figure 1.2, we read the value of Z as Z = 0.770 Example 11 A natural gas mixture consists of the following components: Component

Mole Fraction y

C1 C2 C3 N2 CO2 H 2S

0.780 0.005 0.002 0.013 0.016 0.184

(a) Calculate the apparent molecular weight of the gas, gas gravity, and the pseudocritical temperature and pseudo-critical pressure. (b) Calculate the compressibility factor of the gas at 90°F and 1200 psia. Solution Using Table 1.1, we create the following table showing the molecular weight (M), critical temperature (Tc), and critical pressure (Pc) for each of the component gases. The molecular weight of the mixture and the pseudo-critical temperature and pseudocritical pressure are then calculated using Equation 1.22 and Equation 1.23. Component

y

M

yM

TC

PC

yTC

yPC

C1 C2 C3 N2 CO2 H 2S

0.780 0.005 0.002 0.013 0.016 0.184

16.04 30.07 44.10 28.01 44.01 34.08

12.5112 0.1504 0.0882 0.3641 0.7042 6.2707

343.34 550.07 665.93 227.52 547.73 672.40

667.00 707.80 615.00 492.80 1070.00 1300.00

267.81 2.75 1.33 2.96 8.76 123.72

520.26 3.54 1.23 6.41 17.12 239.20

407.33

787.76

Total

Copyright 2005 by Taylor & Francis Group, LLC

1.000

20.0888

2785_C001.fm Page 26 Friday, April 1, 2005 3:46 PM

26

GAS PIPELINE HYDRAULICS

Therefore, the apparent molecular weight of the natural gas is Mw = ΣyM = 20.09 The gas gravity, using Equation 1.4, is G=

20.09 = 0.6928 29

Next, we calculate the pseudo-critical values Pseudo-critical temperature = ΣyTc = 407.33°R Pseudo-critical pressure = ΣyPc = 787.76 psia Since this sour gas contains more than 5% non-hydrocarbons, we will adjust the pseudo-critical properties using Equation 1.26 through Equation 1.28. We first calculate the temperature adjustment factor e, using Equation 1.26, as follows: A = (0.016 + 0.184) = 0.20 and B = 0.184 Therefore, the adjustment factor is e = 120[(0.2)0.9 – (0.2)1.6] + 15[(0.184)0.5 – (0.184)4.0] = 25.47°R The adjusted pseudo-critical temperature from Equation 1.27 is T′pc = 407.33 – 25.47 = 381.86°R and the adjusted pseudo-critical pressure from Equation 1.28 is P′pc =

787.76 × 381.86 = 731.63 psia 407.33 + 0.184 × (1 − 0.184) × 25.47

Next, we calculate the compressibility factor at 90°F and 1200 psia pressure using these values as follows: From Equation 1.19, pseudo-reduced temperature =

90 + 460 = 1.44 381.86

From Equation 1.20, pseudo-reduced pressure =

1200 = 1.64 731.63

Finally, from the Standing-Katz chart, we get the compressibility factor for the reduced temperature and reduced pressure as Z = 0.825

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 27 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

27

Example 12 The gravity of a natural gas mixture is 0.60. Calculate the compressibility factor of this gas at 1200 psig pressure and a temperature of 70°F, using the CNGA method. Solution Gas temperature Tf = 70 + 460 = 530°R From Equation 1.34, the Z factor can be written as 1 1200 × 344, 400 × (10)1.785 x 0.60 =1+ = 1.1849 Z 530 3.825 Therefore, solving for Z, we get Z = 0.8440

1.12 HEATING VALUE The heating value of a gas is defined as the thermal energy per unit volume of the gas. It is expressed in Btu/ft3. For natural gas, it is approximately in the range of 900 to 1200 Btu/ft3. There are two heating values used in the industry. These are the lower heating value (LHV) and higher heating value (HHV). For a gas mixture, the term gross heating value is used. It is calculated based upon the heating values of the component gases and their mole fractions using the following equation: H m = Σ( yi Hi )

(1.37)

where Hm = gross heating value of mixture, Btu/ft3 yi = mole fraction or percent of gas component i Hi = heating value of gas component, Btu/ft3 For example, a natural gas mixture consisting of 80% of gas A (heating value = 900 Btu/ft3) and 20% of gas B (heating value = 1000 Btu/ft3) will have a gross heating value of Hm = (0.8 × 900) + (0.2 × 1000) = 920 Btu/ft3. 1.13 SUMMARY We discussed several gas properties that influence gas pipeline transportation. The ideal gas equation was introduced along with Boyle’s law and Charles’s law, and how they can be applied with modifications to real gases and real gas mixtures was explained. The gas deviation factor, or compressibility factor, which modifies ideal gas behavior, was introduced. Critical properties of hydrocarbon gases and mixtures and the reduced temperature and pressure that determine the state of a gas were

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 28 Friday, April 1, 2005 3:46 PM

28

GAS PIPELINE HYDRAULICS

explained. Variation of the compressibility factor with pressure and temperature was explored, and calculation methodologies using analytical and graphical approaches were covered. The influence of non-hydrocarbon components in a natural gas mixture was also discussed, along with correction factors for CO2 and H2S in sour gas.

PROBLEMS 1. A natural gas mixture consists of three components, C1, C2, and C3. Their mole fractions and viscosities at a particular temperature are indicated below: Component

Mole Fraction y

Viscosity, cP

C1 C2 C3

0.9000 0.0800 0.0200

0.0130 0.0112 0.0098

Total

1.000

Calculate the viscosity of the gas mixture. 2. At 100 psig and 75°F, a gas has a volume of 800 ft3. If the volume is kept constant and the gas temperature increases to 100°F, what is the final pressure of the gas? Keeping the pressure constant at 100 psig, if the temperature increases to 100°F, what is the final volume? Use 14.73 psi for base pressure. 3. Calculate the apparent molecular weight of a natural gas mixture that has 89% methane, 8% ethane, 2% propane, and 1% normal butane as shown below. Component

Percent

Molecular Weight

C1 C2 C3 C4

89 8 2 1

16.01 30.10 44.10 58.10

Total

100

4. Calculate the pseudo-critical temperature and the pseudo-critical pressure of a natural gas mixture consisting of 89% methane, 8% ethane, and 3% propane. The critical properties of C1, C2, and C3 components are as follows: Components

Critical Temperature, °R

Critical Pressure, psia

C1 C2 C3

343 550 666

667 708 615

5. If the temperature of the gas in the previous example is 80°F and the average gas pressure is 1000 psig, what are the pseudo-reduced temperature and pseudoreduced pressure of this gas? Use 14.7 psia for base pressure.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 29 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

29

6. Calculate the gravity of a natural gas mixture consisting of 84% methane, 10% ethane, and 6% propane. From the gas gravity, calculate the pseudo-critical temperature and pseudo-critical pressure for this natural gas mixture. 7. The pseudo-critical temperature and pressure of a natural gas mixture were calculated as 380°R and 675 psia. If the CO2 content is 12% and H2S is 22%, calculate the adjustment factor e and the adjusted values of the pseudo-critical temperature and pseudo-critical pressure. 8. Using the Standing-Katz compressibility chart, calculate the compressibility factor for the gas in Problem 7 at 80°F and 1000 psig. Use the values of Tpc and Ppc calculated in Problem 7. 9. A natural gas mixture consists of the following components: Component

Mole Fraction y

C1 C2 C3 N2 CO2 H 2S

0.850 0.004 0.002 0.014 0.010 0.120

(a) Calculate the apparent molecular weight of the gas, gravity, and the pseudocritical temperature and pseudo-critical pressure. (b) Calculate the compressibility factor of the gas at 100°F and 1400 psia. 10. The gravity of a natural gas mixture is 0.62. Calculate the compressibility factor of this gas at 1400 psig and a temperature of 80°F, using the CNGA method.

REFERENCES 1. McCain, W.D. Jr., The Properties of Petroleum Fluids, Petroleum Publishing Company, Tulsa, OK, 1973. 2. Holman, J.P., Thermodynamics, 2nd ed., McGraw-Hill, New York, 1974. 3. Gas Processors Suppliers Association, Engineering Data Book, 10th ed., Tulsa, OK, 1994. 4. AGA Report No. 10, Steady Flow in Gas Pipelines, AGA, 1965. 5. AGA Report No. 8, Compressibility Factors, AGA, Nov. 1992.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 30 Friday, April 1, 2005 3:46 PM

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 31 Friday, April 1, 2005 3:23 PM

CHAPTER

2

Pressure Drop Due to Friction In this chapter we will discuss the various methods of calculating the pressure drop due to friction in a gas pipeline. The pipeline throughput (ﬂow rate) will depend upon the gas properties, pipe diameter and length, initial gas pressure and temperature, and the pressure drop due to friction. Commonly used formulas will be reviewed and illustrated using examples. The impact of internal conditions of the pipe on the pipe capacity will also be explored. 2.1 BERNOULLI’S EQUATION As gas ﬂows through a pipeline, the total energy of the gas at various points consists of energy due to pressure, energy due to velocity, and energy due to position or elevation above an established datum. Bernoulli’s equation simply connects these components of the energy of the ﬂowing ﬂuid to form an energy conservation equation. Bernoulli’s equation is stated as follows, considering two points, 1 and 2, as shown in Figure 2.1. ZA +

P V2 PA VA 2 + + H p = ZB + B + B + hf γ 2g γ 2g

(2.1)

where Hp is the equivalent head added to the ﬂuid by a compressor at A and hf represents the total frictional pressure loss between points A and B. Starting with the basic energy Equation 2.1, applying gas laws, and after some simpliﬁcation, various formulas were developed over the years to predict the performance of a pipeline transporting gas. These formulas are intended to show the relationship between the gas properties, such as gravity and compressibility factor, with the ﬂow rate, pipe diameter and length, and the pressures along the pipeline. Thus, for a given pipe size and length, we can predict the ﬂow rate possible through a pipeline based upon an inlet pressure and an outlet pressure of a pipe segment. Simpliﬁcations are sometimes introduced, such as uniform gas temperature and no heat transfer between the gas and the surrounding soil in a buried pipeline, in order 31

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 32 Friday, April 1, 2005 3:23 PM

32

GAS PIPELINE HYDRAULICS

Velocity VA Pressure PA

B

Velocity VB Pressure PB

Flow A

Datum for Elevations Figure 2.1

Energy of flow of a fluid.

to adopt these equations for manual calculations. With the advent of microcomputers, we are able to introduce heat transfer effects and, therefore, more accurately model gas pipelines, taking into consideration gas ﬂow temperatures, soil temperatures, and thermal conductivities of pipe material, insulation, and soil. In this chapter we will concentrate on steady-state isothermal ﬂow of gas in pipelines. Appendix D includes an output report from a commercial gas pipeline simulation model that takes into account heat transfer. For most practical purposes, the assumption of isothermal ﬂow is good enough, since in long transmission lines the gas temperature reaches constant values, anyway. 2.2 FLOW EQUATIONS Several equations are available that relate the gas ﬂow rate with gas properties, pipe diameter and length, and upstream and downstream pressures. These equations are listed as follows: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

General Flow equation Colebrook-White equation Modiﬁed Colebrook-White equation AGA equation Weymouth equation Panhandle A equation Panhandle B equation IGT equation Spitzglass equation Mueller equation Fritzsche equation

We will discuss each of these equations, their limitations, and their applicability to compressible ﬂuids, such as natural gas, using illustrated examples. A comparison of these equations will also be discussed using an example pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 33 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

33

2.3 GENERAL FLOW EQUATION The General Flow equation, also called the Fundamental Flow equation, for the steady-state isothermal ﬂow in a gas pipeline is the basic equation for relating the pressure drop with ﬂow rate. The most common form of this equation in the U.S. Customary System (USCS) of units is given in terms of the pipe diameter, gas properties, pressures, temperatures, and ﬂow rate as follows. Refer to Figure 2.2 for an explanation of symbols used. T P 2 − P22 Q = 77.54 b 1 Pb GT f LZf where Q = f = Pb = Tb = P1 = P2 = G = Tf = L = Z = D =

0.5

D 2.5

(USCS units)

(2.2)

gas ﬂow rate, measured at standard conditions, ft3/day (SCFD) friction factor, dimensionless base pressure, psia base temperature, °R(460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂowing temperature, °R (460 + °F) pipe segment length, mi gas compressibility factor at the ﬂowing temperature, dimensionless pipe inside diameter, in.

It must be noted that for the pipe segment from section 1 to section 2, the gas temperature Tf is assumed to be constant (isothermal ﬂow). In SI units, the General Flow equation is stated as follows:

(

)

2 2 T P1 − P2 Q = 1.1494 × 10 −3 b Pb GT f LZf

0.5

D 2.5

(SI units)

Diameter D Temperature Tf

Pressure P1

1 Figure 2.2

Steady flow in gas pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

Pressure P2

Flow Q

Length L

2

(2.3)

2785_C002.fm Page 34 Friday, April 1, 2005 3:23 PM

34

GAS PIPELINE HYDRAULICS

where Q = gas ﬂow rate, measured at standard conditions, m3/day f = friction factor, dimensionless Pb = base pressure, kPa Tb = base temperature, K (273 + °C) P1 = upstream pressure, kPa P2 = downstream pressure, kPa G = gas gravity (air = 1.00) Tf = average gas ﬂowing temperature, K (273 + °C) L = pipe segment length, km Z = gas compressibility factor at the ﬂowing temperature, dimensionless D = pipe inside diameter, mm Due to the nature of Equation 2.3, the pressures can also be in MPa or Bar, as long as the same consistent unit is used. Equation 2.2 relates the capacity (ﬂow rate or throughput) of a pipe segment of length L, based on an upstream pressure of P1 and a downstream pressure of P2 as shown in Figure 2.2. It is assumed that there is no elevation difference between the upstream and downstream points; therefore, the pipe segment is horizontal. Upon examining the General Flow Equation 2.2, we see that for a pipe segment of length L and diameter D, the gas ﬂow rate Q (at standard conditions) depends on several factors. Q depends on gas properties represented by the gravity G and the compressibility factor Z. If the gas gravity is increased (heavier gas), the ﬂow rate will decrease. Similarly, as the compressibility factor Z increases, the ﬂow rate will decrease. Also, as the gas ﬂowing temperature Tf increases, throughput will decrease. Thus, the hotter the gas, the lower the ﬂow rate will be. Therefore, to increase the ﬂow rate, it helps to keep the gas temperature low. The impact of pipe length and inside diameter is also clear. As the pipe segment length increases for given pressure P1 and P2, the ﬂow rate will decrease. On the other hand, the larger the diameter, the larger the ﬂow rate will be. The term P12 – P22 represents the driving force that causes the ﬂow rate from the upstream end to the downstream end. As the downstream pressure P2 is reduced, keeping the upstream pressure P1 constant, the ﬂow rate will increase. It is obvious that when there is no ﬂow rate, P1 is equal to P2. It is due to friction between the gas and pipe walls that the pressure drop (P1 – P2) occurs from the upstream point 1 to downstream point 2. The friction factor f depends on the internal condition of the pipe as well as the type of ﬂow (laminar or turbulent) and will be discussed in detail beginning in Section 2.8. Sometimes the General Flow equation is represented in terms of the transmission factor F instead of the friction factor f. This form of the equation is as follows. T P 2 − P22 Q = 38.77 F b 1 Pb GT f LZ

Copyright 2005 by Taylor & Francis Group, LLC

0.5

D 2.5

(USCS units)

(2.4)

2785_C002.fm Page 35 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

35

where the transmission factor F and friction factor f are related by F=

2

(2.5)

f

and in SI units

(

2 2 Tb P1 − P2 −4 Q = 5.747 × 10 F Pb GT f LZ

)

0.5

D 2.5

(SI units)

(2.6)

We will discuss several aspects of the General Flow equation before moving on to the other formulas for pressure drop calculation.

2.4 EFFECT OF PIPE ELEVATIONS When elevation difference between the ends of a pipe segment is included, the General Flow equation is modiﬁed as follows: T P 2 − e s P22 Q = 38.77 F b 1 Pb GT f Le Z

0.5

D 2.5

(USCS units)

(2.7)

D 2.5

(2.8)

and in SI units

(

s 2 2 T P1 − e P2 Q = 5.747 × 10 F b Pb GT f Le Z −4

)

0.5

(SI units)

where Le =

L (e s − 1) s

(2.9)

The equivalent length, Le, and the term es take into account the elevation difference between the upstream and downstream ends of the pipe segment. The parameter s depends upon the gas gravity, gas compressibility factor, the ﬂowing temperature, and the elevation difference. It is deﬁned as follows in USCS units: H − H1 s = 0.0375G 2 Tf Z

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(2.10)

2785_C002.fm Page 36 Friday, April 1, 2005 3:23 PM

36

where s = H1 = H2 = e =

GAS PIPELINE HYDRAULICS

elevation adjustment parameter, dimensionless upstream elevation, ft downstream elevation, ft base of natural logarithms (e = 2.718…)

Other symbols are as deﬁned earlier. In SI units, the elevation adjustment parameter s is deﬁned as follows: H − H1 s = 0.0684G 2 Tf Z

(SI units)

(2.11)

where H1 = upstream elevation, m H2 = downstream elevation, m Other symbols are as deﬁned earlier. In the calculation of Le in Equation 2.9, we have assumed that there is a single slope between the upstream point 1 and the downstream point 2 in Figure 2.2. If, however, the pipe segment of length L has a series of slopes, then we introduce a parameter j as follows for each individual pipe subsegment that constitutes the pipe length from point 1 to point 2. j=

es − 1 s

(2.12)

The parameter j is calculated for each slope of each pipe subsegment of length L1, L2, etc. that make up the total length L. The equivalent length term Le in Equation 2.7 and Equation 2.8 is calculated by summing the individual slopes as deﬁned below. Le = j1L1 + j2L2es1 + j3L3es2 + …

(2.13)

The terms j1, j2, etc. for each rise or fall in the elevations of individual pipe subsegments are calculated for the parameters s1, s2, etc. for each segment in accordance with Equation 2.12, from the pipeline inlet to the end of each segment. In the subsequent sections of this chapter, we will discuss how the friction factor and transmission factor are calculated using various equations such as ColebrookWhite and AGA. It is important to note that the General Flow equation is the most commonly used equation to calculate the ﬂow rate and pressure in a gas pipeline. In order to apply it correctly, we must use the correct friction factor or transmission factor. The Colebrook equation, AGA equation, and other empirical equations are used to calculate the friction factor to be used in the General Flow equation. Several other equations, such as Panhandle A, Panhandle B, and Weymouth, calculate the ﬂow rate for a given pressure without using a friction factor or transmission factor. However, an equivalent friction factor (or transmission factor) can be calculated using these methods as well.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 37 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

37

2.5 AVERAGE PIPE SEGMENT PRESSURE In the General Flow equation, the compressibility factor Z is used. This must be calculated at the gas ﬂowing temperature and average pressure in the pipe segment. Therefore, it is important to ﬁrst calculate the average pressure in a pipe segment, described in Figure 2.2. Consider a pipe segment with upstream pressure P1 and downstream pressure P2, as in Figure 2.2. An average pressure for this segment must be used to calculate the compressibility factor of gas at the average gas temperature Tf. As a ﬁrst approximation, we may use an arithmetic average of (P1 + P2)/2. However, it has been found that a more accurate value of the average gas pressure in a pipe segment is Pavg =

PP 2 P1 + P2 − 1 2 P1 + P2 3

(2.14)

Another form of the average pressure in a pipe segment is

Pavg =

3 3 2 P1 − P2 3 P12 − P2 2

(2.15)

It must be noted that the pressures used in the General Flow equation are all in absolute units. Therefore, gauge pressure units should be converted to absolute pressure by adding the base pressure. For example, the upstream and downstream pressures are 1000 psia and 900 psia, respectively. From Equation 2.14, the average pressure is Pavg =

1000 × 900 2 1000 + 900 − = 950.88 psia 1900 3

Compare this to the arithmetic average of Pavg =

1 (1000 + 900) = 950 psia 2

2.6 VELOCITY OF GAS IN A PIPELINE The velocity of gas ﬂow in a pipeline represents the speed at which the gas molecules move from one point to another. Unlike a liquid pipeline, due to compressibility, the gas velocity depends upon the pressure and, hence, will vary along the pipeline even if the pipe diameter is constant. The highest velocity will be at the downstream end, where the pressure is the least. Correspondingly, the least velocity will be at the upstream end, where the pressure is higher.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 38 Friday, April 1, 2005 3:23 PM

38

GAS PIPELINE HYDRAULICS

Consider a pipe transporting gas from point A to point B as shown in Figure 2.2. Under steady state ﬂow, at A, the mass ﬂow rate of gas is designated as M and will be the same as the mass ﬂow rate at point B, if between A and B there is no injection or delivery of gas. The mass being the product of volume and density, we can write the following relationship for point A: M = Qr

(2.16)

The volume rate Q can be expressed in terms of the ﬂow velocity u and pipe cross sectional area A as follows: Q=uA

(2.17)

Therefore, combining Equation 2.16 and Equation 2.17 and applying the conservation of mass to points A and B, we get M1 = u1 A1r1 = M2 = u2 A2 r2

(2.18)

where subscripts 1 and 2 refer to points A and B, respectively. If the pipe is of uniform cross section between A and B, then A1 = A2 = A. Therefore, the area term in Equation 2.18 can be dropped, and the velocities at A and B are related by the following equation: u1 r1 = u2 r2

(2.19)

Since the ﬂow of gas in a pipe can result in variation of temperature from point A to point B, the gas density will also vary with temperature and pressure. If the density and velocity at one point are known, the corresponding velocity at the other point can be calculated using Equation 2.19. If inlet conditions are represented by point A and the volume ﬂow rate Q at standard conditions of 60°F and 14.7 psia are known, we can calculate the velocity at any point along the pipeline at which the pressure and temperature of the gas are P and T, respectively. The velocity of gas at section 1 is related to the ﬂow rate Q1 at section 1 and pipe cross-sectional area A as follows from Equation 2.17: Q1 = u1 A The mass ﬂow rate M at section 1 and 2 is the same for steady-state ﬂow. Therefore, M = Q1r1 = Q2 r2 = Qb rb

(2.20)

where Qb is the gas ﬂow rate at standard conditions and rb is the corresponding gas density. Therefore, simplifying Equation 2.20, ρ Q1 = Qb b ρ1

Copyright 2005 by Taylor & Francis Group, LLC

(2.21)

2785_C002.fm Page 39 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

39

Applying the gas law Equation 1.9, we get P1 = Z1 RT1 ρ1 or

ρ1 =

P1 Z1 RT1

(2.22)

where P1 and T1 are the pressure and temperature at pipe section 1. Similarly, at standard conditions,

ρb =

Pb Z b RTb

(2.23)

From Equation 2.21, Equation 2.22, and Equation 2.23, we get P T Z Q1 = Qb b 1 1 Tb P1 Z b

(2.24)

Since Zb = 1.00, approximately, we can simplify this to P T Q1 = Qb b 1 Z1 Tb P1

(2.25)

Therefore, the gas velocity at section 1 is, using Equation 2.17 and Equation 2.25, u1 =

P T Qb Z1 Pb T1 4 × 144 = Qb Z1 b 1 2 A Tb P1 Tb P1 πD

or Q P ZT u1 = 0.002122 b2 b 1 1 D Tb P1 where u1 = Qb = D = Pb = Tb = P1 = T1 = Z1 =

(USCS units)

upstream gas velocity, ft/s gas ﬂow rate, measured at standard conditions, ft3/day (SCFD) pipe inside diameter, in. base pressure, psia base temperature, °R(460 + °F) upstream pressure, psia upstream gas temperature, °R(460 + °F) gas compressibility factor at upstream conditions, dimensionless

Copyright 2005 by Taylor & Francis Group, LLC

(2.26)

2785_C002.fm Page 40 Friday, April 1, 2005 3:23 PM

40

GAS PIPELINE HYDRAULICS

Similarly, the gas velocity at section 2 is given by Q P Z T u2 = 0.002122 b2 b 2 2 D Tb P2

(USCS units)

(2.27)

In general, the gas velocity at any point in a pipeline is given by Q P ZT u = 0.002122 b2 b D Tb P

(2.28)

In SI units, the gas velocity at any point in a gas pipeline is given by Q P ZT u = 14.7349 b2 b D Tb P

(SI units)

(2.29)

where u = gas velocity, m/s Qb = gas ﬂow rate, measured at standard conditions, m3/day D = pipe inside diameter, mm Pb = base pressure, kPa Tb = base temperature, K(273 + °C) P = pressure, kPa T = average gas ﬂowing temperature, K(273 + °C) Z = gas compressibility factor at the ﬂowing temperature, dimensionless Since the right-hand side of Equation 2.29 contains ratios of pressures, any consistent unit can be used, such as kPa, MPa, or Bar.

2.7 EROSIONAL VELOCITY We have seen from the preceding section that the gas velocity is directly related to the ﬂow rate. As ﬂow rate increases, so does the gas velocity. How high can the gas velocity be in a pipeline? As the velocity increases, vibration and noise are evident. In addition, higher velocities will cause erosion of the pipe interior over a long period of time. The upper limit of the gas velocity is usually calculated approximately from the following equation: umax =

100

ρ

where umax = maximum or erosional velocity, ft/s r = gas density at ﬂowing temperature, lb/ft3

Copyright 2005 by Taylor & Francis Group, LLC

(2.30)

2785_C002.fm Page 41 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

41

Since the gas density r may be expressed in terms of pressure and temperature, using the gas law Equation 1.8, the maximum velocity Equation 2.30 can be rewritten as umax = 100 where Z = R = T = G= P =

ZRT 29GP

(USCS units)

(2.31)

compressibility factor of gas, dimensionless gas constant = 10.73 ft3 psia/lb-moleR gas temperature, °R gas gravity (air = 1.00) gas pressure, psia

Usually, an acceptable operational velocity is 50% of the above. Example 1 A gas pipeline, NPS 20 with 0.500 in. wall thickness, transports natural gas (speciﬁc gravity = 0.6) at a ﬂow rate of 250 MMSCFD at an inlet temperature of 60°F. Assuming isothermal ﬂow, calculate the velocity of gas at the inlet and outlet of the pipe if the inlet pressure is 1000 psig and the outlet pressure is 850 psig. The base pressure and base temperature are 14.7 psia and 60°F, respectively. Assume compressibility factor Z = 1.00. What is the erosional velocity for this pipeline based on the above data and a compressibility factor Z = 0.90? Solution If we assume compressibility factor Z = 1.00, then using Equation 2.26, the velocity of gas at the inlet pressure of 1000 psig is 250 × 10 6 14.7 60 + 460 u1 = 0.002122 = 21.29 ft/s 19.0 2 60 + 460 1014.7 and the gas velocity at the outlet is by proportions u2 = 21.29 ×

1014.7 = 24.98 ft/s 864.7

The erosional velocity is found for Z = 0.90, using Equation 2.31, umax = 100

0.9 × 10.73 × 520 = 53.33 ft/s 29 × 0.6 × 1014.7

Example 2 A gas pipeline, DN 500 with 12 mm wall thickness, transports natural gas (speciﬁc gravity = 0.6) at a ﬂow rate of 7.5 Mm3/day at an inlet temperature of 15°C. Assuming

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 42 Friday, April 1, 2005 3:23 PM

42

GAS PIPELINE HYDRAULICS

isothermal ﬂow, calculate the velocity of gas at the inlet and outlet of the pipe if the inlet pressure is 7 MPa and the outlet pressure is 6 MPa. The base pressure and base temperature are 0.1 MPa and 15°C. Assume compressibility factor Z = 0.95. Solution Inside diameter of pipe D = 500 – (2 × 12) = 476 mm. Flow rate at standard conditions Qb = 7.5 × 106 m3/day. Using Equation 2.29, the velocity of gas at the inlet pressure of 7 MPa is 7.5 × 10 6 0.1 0.95 × 288 u1 = 14.7349 = 6.62 m/s 7.0 4762 15 + 273 and the gas velocity at the outlet is by proportions u2 = 6.62 ×

7.0 = 7.72 m/s 6.0

In the preceding Examples 1 and 2, we have assumed the value of compressibility factor Z to the constant. A more accurate solution will be to calculate the value of Z using one of the methods outlined in Chapter 1, such as the CNGA or Standing-Katz method. For example, if we used the CNGA Equation 1.34, the compressibility factor in Example 1 will be 1 1 + 1000×344400×(10 )1.785× 0.6 5203.825 = 0.8578 at an inlet pressure of 1000 psig.

Z1 =

and Z2 =

1

1 + 5203.825 = 0.8765 at an outlet pressure of 850 psig. 850 × 344400 × (10 )1.785× 0.6

The inlet and outlet gas velocities then will be modiﬁed as follows: Inlet velocity u1 = 0.8578 × 21.29 = 18.26 ft/s Outlet velocity u2 = 0.8765 × 24.98 = 21.90 ft/s

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 43 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

43

2.8 REYNOLDS NUMBER OF FLOW An important parameter in ﬂow of ﬂuids in a pipe is the nondimensional term Reynolds number. The Reynolds number is used to characterize the type of ﬂow in a pipe, such as laminar, turbulent, or critical ﬂow. It is also used to calculate the friction factor in pipe ﬂow. We will ﬁrst outline the calculation of the Reynolds number based upon the properties of the gas and pipe diameter and then discuss the range of Reynolds number for the various types of ﬂow and how to calculate the friction factor. The Reynolds number is a function of the gas ﬂow rate, pipe inside diameter, and the gas density and viscosity and is calculated from the following equation: Re = where Re u D r m

= = = = =

uDρ µ

(USCS units)

(2.32)

Reynolds number, dimensionless average velocity of gas in pipe, ft/s inside diameter of pipe, ft gas density, lb/ft3 gas viscosity, lb/ft-s

The above equation for the Reynolds number is in USCS units. The corresponding equation for the Reynolds number in SI units is as follows: Re = where Re u D r m

uDρ µ

(SI units)

(2.33)

= Reynolds number, dimensionless = average velocity of gas in pipe, m/s = inside diameter of pipe, m = gas density, kg/m3 = gas viscosity, kg/m-s

In gas pipeline hydraulics, using customary units, a more suitable equation for the Reynolds number is as follows: P GQ Re = 0.0004778 b Tb µ D where Pb = base pressure, psia Tb = base temperature, °R (460 + °F) G = speciﬁc gravity of gas (air = 1.0) Q = gas ﬂow rate, standard ft3/day (SCFD) D = pipe inside diameter, in. m = viscosity of gas, lb/ft-s

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(2.34)

2785_C002.fm Page 44 Friday, April 1, 2005 3:23 PM

44

GAS PIPELINE HYDRAULICS

In SI units, the Reynolds number is P GQ Re = 0.5134 b Tb µ D

(SI units)

(2.35)

where Pb = base pressure, kPa Tb = base temperature, °K (273 + °C) G = speciﬁc gravity of gas (air = 1.0) Q = gas ﬂow rate, m3/day (standard conditions) D = pipe inside diameter, mm m = viscosity of gas, Poise Laminar ﬂow occurs in a pipeline when the Reynolds number is below a value of approximately 2000. Turbulent ﬂow occurs when the Reynolds number is greater than 4000. For Reynolds numbers between 2000 and 4000, the ﬂow is undeﬁned and is referred to as critical ﬂow. Thus, For laminar ﬂow, Re ≤ 2000 For turbulent ﬂow, Re > 4000 For critical ﬂow, Re > 2000 and Re ≤ 4000

Most natural gas pipelines operate in the turbulent ﬂow region. Therefore, the Reynolds number is greater than 4000. Turbulent ﬂow is further divided into three regions known as smooth pipe ﬂow, fully rough pipe ﬂow, and transition ﬂow. We will discuss these ﬂow regions in more detail in the subsequent sections of this chapter. Example 3 A natural gas pipeline, NPS 20 with 0.500 in. wall thickness, transports 100 MMSCFD. The speciﬁc gravity of gas is 0.6 and viscosity is 0.000008 lb/ft-s. Calculate the value of the Reynolds number of ﬂow. Assume the base temperature and base pressure are 60°F and 14.7 psia, respectively. Solution Pipe inside diameter = 20 – 2 × 0.5 = 19.0 in. The base temperature = 60 + 460 = 520 °R Using Equation 2.34, we get 14.7 0.6 × 100 × 10 6 Re = 0.0004778 = 5,331,726 520 0.000008 × 19 Since Re is greater than 4000, the ﬂow is in the turbulent region.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 45 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

45

Example 4 A natural gas pipeline, DN 500 with 12 mm wall thickness, transports 3 Mm3/day. The speciﬁc gravity of gas is 0.6 and viscosity is 0.00012 Poise. Calculate the value of the Reynolds number. Assume the base temperature and base pressure are 15°C and 101 kPa, respectively. Solution Pipe inside diameter = 500 – 2 × 12 = 476 mm The base temperature = 15 + 273 = 288 K Using Equation 2.35, we get 101 0.6 × 3 × 10 6 Re = 0.5134 = 5,673,735 15 + 273 0.00012 × 476 Since Re is greater than 4000, the ﬂow is in the turbulent region.

2.9 FRICTION FACTOR In order to calculate the pressure drop in a pipeline at a given ﬂow rate, we must ﬁrst understand the concept of friction factor. The term friction factor is a dimensionless parameter that depends upon the Reynolds number of ﬂow. In engineering literature, we ﬁnd two different friction factors mentioned. The Darcy friction factor is more common and will be used throughout this book. Another friction factor known as the Fanning friction factor is preferred by some engineers. The Fanning friction factor is numerically equal to one-fourth the Darcy friction factor as below. ff =

fd 4

(2.36)

where ff = Fanning friction factor fd = Darcy friction factor To avoid confusion, in subsequent discussions, the Darcy friction factor is used and will be represented by the symbol f. For laminar ﬂow, the friction factor is inversely proportional to the Reynolds number, as indicated below. f=

64 Re

(2.37)

For turbulent ﬂow, the friction factor is a function of the Reynolds number, pipe inside diameter, and internal roughness of the pipe. Many empirical relationships for

Copyright 2005 by Taylor & Francis Group, LLC

Laminar Critical Flow Zone Transition Zone

Complete turbulence, rough pipes 0.05 0.04

0.07

4 / Re

0.01 0.008 0.006 0.004

0.025

0.002

0.02

0.001 0.0008 0.0006 0.0004

Sm

0.015

oo

th

0.0002 0.0001

pip

es

103

2 3 4 56 8104 ×103

2 3 4 56 8105 ×104

2 3 4 56 8106 ×105

Reynolds number Re = VD n Moody diagram.

2 3 4 56 8107 ×106

0.000,01 2 3 4 56 8108 e e D = 0. 000D = 0.00 0,0 ,00 05 1

GAS PIPELINE HYDRAULICS

0.000,05

0.01 0.009 0.008

Figure 2.3

Relative roughness e D

f=6

0.03 Friction factor f

0.02 0.015

flow

0.04

0.03

inar

0.05

Lam

0.06

2785_C002.fm Page 46 Friday, April 1, 2005 3:23 PM

46

Copyright 2005 by Taylor & Francis Group, LLC

0.10 0.09 0.08

2785_C002.fm Page 47 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

47

calculating f have been put forth by researchers. The more popular correlations include the Colebrook-White and AGA equations. Before we discuss the equations for calculating the friction factor in turbulent ﬂow, it is appropriate to analyze the turbulent ﬂow regime. Turbulent ﬂow in pipes (Re > 4000) is subdivided into three separate regions as follows: 1. Turbulent ﬂow in smooth pipes 2. Turbulent ﬂow in fully rough pipes 3. Transition ﬂow between smooth pipes and rough pipes

For turbulent ﬂow in smooth pipes, the friction factor f depends only on the Reynolds number. For fully rough pipes, f depends more on the pipe internal roughness and less on the Reynolds number. In the transition zone between smooth pipe ﬂow and ﬂow in fully rough pipes, f depends on the pipe roughness, pipe inside diameter, and the Reynolds number. The various ﬂow regimes are depicted in the Moody diagram, shown in Figure 2.3. The Moody diagram is a graphic plot of the variation of the friction factor with the Reynolds number for various values of relative pipe roughness. The latter term is simply a dimensionless parameter obtained by dividing the absolute (or internal) pipe roughness by the pipe inside diameter as follows: Relative roughness =

e D

(2.38)

where e = absolute or internal roughness of pipe, in. D = pipe inside diameter, in. The terms absolute pipe roughness and internal pipe roughness are equivalent. Generally, the internal pipe roughness is expressed in microinches (one-millionth of an inch). For example, an internal roughness of 0.0006 in. is referred to as 600 microinches or 600 µin. If the pipe inside diameter is 15.5 in., the relative roughness is, in this case, Relative roughness =

0.0006 = 0.0000387 = 3.87 × 10−5 15.5

For example, from the Moody diagram in Figure 2.3, for Re = 10 million and e/D = 0.0001, we ﬁnd that f = 0.012.

2.10 COLEBROOK-WHITE EQUATION The Colebrook-White equation, sometimes referred to simply as the Colebrook equation, is a relationship between the friction factor and the Reynolds number, pipe roughness, and inside diameter of pipe. The following form of the Colebrook

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 48 Friday, April 1, 2005 3:23 PM

48

GAS PIPELINE HYDRAULICS

equation is used to calculate the friction factor in gas pipelines in turbulent ﬂow. e 2.51 = −2 Log10 + f 3.7 D Re f

1

where f D e Re

= = = =

for Re > 4000

(2.39)

friction factor, dimensionless pipe inside diameter, in. absolute pipe roughness, in. Reynolds number of ﬂow, dimensionless

Since Re and f are dimensionless, as long as consistent units are used for both e and D, the Colebrook equation is the same regardless of the units employed. Therefore, in SI units, Equation 2.39 is used with e and D expressed in mm. It can be seen from Equation 2.39 that in order to calculate the friction factor f, we must use a trial-and-error approach. It is an implicit equation in f, since f appears on both sides of the equation. We ﬁrst assume a value of f (such as 0.01) and substitute it in the right-hand side of the equation. This will yield a second approximation for f, which can then be used to calculate a better value of f, and so on. Generally 3 to 4 iterations are sufﬁcient to converge on a reasonably good value of the friction factor. It can be seen from the Colebrook Equation 2.39, for turbulent ﬂow in smooth pipes, the ﬁrst term within the square brackets is negligible compared to the second term, since pipe roughness e is very small. Therefore, for smooth pipe ﬂow, the friction factor equation reduces to 2.51 = −2 Log10 , f Re f

1

for turbulent ﬂow in smooth pipes

(2.40)

Similarly, for turbulent ﬂow in fully rough pipes, with Re being a large number, f depends mostly on the roughness e and, therefore, the friction factor equation reduces to e = −2 Log10 , 3.7 D f

1

for turbulent ﬂow in fully rough pipes

(2.41)

Table 2.1 lists typical values of pipe internal roughness used to calculate the friction factor. As an example, if Re = 100 million or larger and e/D = 0.0002, the friction factor from Equation 2.41 is 0.0002 = −2 Log10 3.7 f

1

or f = 0.0137, which correlates well with the friction factor obtained from the Moody diagram in Figure 2.3.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 49 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

49

Table 2.1 Pipe Internal Roughness Pipe Material Riveted steel Commercial steel/welded steel Cast iron Galvanized iron Asphalted cast iron Wrought iron PVC, drawn tubing, glass Concrete

Roughness, in.

Roughness, mm

0.0354 to 0.354 0.0018 0.0102 0.0059 0.0047 0.0018 0.000059 0.0118 to 0.118

0.9 to 9.0 0.045 0.26 0.15 0.12 0.045 0.0015 0.3 to 3.0

Example 5 A natural gas pipeline, NPS 20 with 0.500 in. wall thickness, transports 200 MMSCFD. The speciﬁc gravity of gas is 0.6 and viscosity is 0.000008 lb/ft-s. Calculate the friction factor using the Colebrook equation. Assume absolute pipe roughness = 600 µ in. The base temperature and base pressure are 60°F and 14.7 psia, respectively. Solution Pipe inside diameter = 20 – 2 × 0.5 = 19.0 in. Absolute pipe roughness = 600 µ in. = 0.0006 in. First, we calculate the Reynolds number using Equation 2.34: 14.7 0.6 × 200 × 10 6 Re = 0.0004778 = 10,663,452 60 + 460 0.000008 × 19 Using Equation 2.39, 0.0006 2.51 = −2 Log10 + f 3.7 × 19 10, 663, 452 f

1

This equation will be solved by successive iteration. Assume f = 0.01 initially; substituting above, we get a better approximation as f = 0.0101. Repeating the iteration, we get the ﬁnal value as f = 0.0101. Therefore, the friction factor is 0.0101. Example 6 A natural gas pipeline, DN 500 with 12 mm wall thickness, transports 6 Mm3/day. The speciﬁc gravity of gas is 0.6 and viscosity is 0.00012 Poise. Calculate the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 50 Friday, April 1, 2005 3:23 PM

50

GAS PIPELINE HYDRAULICS

friction factor using the Colebrook equation. Assume absolute pipe roughness = 0.03 mm and assume the base temperature and base pressure are 15°C and 101 kPa, respectively. Solution Pipe inside diameter = 500 – 2 × 12 = 476 mm First, we calculate the Reynolds number using Equation 2.35: 101 0.6 × 6 × 10 6 Re = 0.5134 = 11,347,470 15 + 273 0.00012 × 476 Using Equation 2.39, the friction factor is 0.030 2.51 = −2 Log10 + . 3 7 476 × 11, 347, 470 f f

1

This equation will be solved by successive iteration. Assume f = 0.01 initially; substituting above, we get a better approximation as f = 0.0112. Repeating the iteration, we get the ﬁnal value as f = 0.0112. Therefore, the friction factor is 0.0112.

2.11 TRANSMISSION FACTOR The transmission factor F is considered the opposite of the friction factor f. Whereas the friction factor indicates how difﬁcult it is to move a certain quantity of gas through a pipeline, the transmission factor is a direct measure of how much gas can be transported through the pipeline. As the friction factor increases, the transmission factor decreases and, therefore, the gas ﬂow rate also decreases. Conversely, the higher the transmission factor, the lower the friction factor and, therefore, the higher the ﬂow rate will be. The transmission factor F is related to the friction factor f as follows: F=

2

f =

4 F2

(2.42)

f

Therefore,

Copyright 2005 by Taylor & Francis Group, LLC

(2.43)

2785_C002.fm Page 51 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

51

where f = friction factor F = transmission factor It must be noted that the friction factor f in the above equation is the Darcy friction factor. Since some engineers prefer to use the Fanning friction factor, the relationship between the transmission factor F and the Fanning friction factor is given below for reference. F=

1

(2.44)

ff

where ff is the Fanning friction factor. For example, if the Darcy friction factor is 0.025, the transmission factor is, using Equation 2.42, F=

2 0.025

= 12.65

The Fanning friction factor in this case will be 0.025 = 0.00625. Therefore, the 4 1 transmission factor using Equation 2.44 is F = 0.00625 = 12.65, which is the same as calculated using the Darcy friction factor. Thus, it must be noted that there is only one transmission factor, whereas there are two different friction factors. Having deﬁned a transmission factor, we can rewrite the Colebrook Equation 2.39 in terms of the transmission factor using Equation 2.42 as follows: e 1.255F F = − 4 Log10 + Re 3.7 D

(2.45)

Since Re and F are dimensionless, as long as consistent units are used for both e and D, the transmission factor equation is the same regardless of the units employed. Therefore, in SI units, Equation 2.45 is used with e and D expressed in mm. Similar to the calculation of the friction factor f from Equation 2.39, to calculate the transmission factor F from Equation 2.45, an iterative approach must be used. This will be illustrated using an example. Example 7 For a gas pipeline, ﬂowing 100 MMSCFD gas of speciﬁc gravity 0.6 and viscosity of 0.000008 lb/ft-s, calculate the friction factor and transmission factor considering an NPS 20 pipeline, 0.500-in. wall thickness, and an internal roughness of 600 microinches. Assume the base temperature and base pressure are 60°F and 14.7 psia, respectively. If the ﬂow rate increases by 50%, what is the impact on the friction factor and transmission factor?

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 52 Friday, April 1, 2005 3:23 PM

52

GAS PIPELINE HYDRAULICS

Solution The base temperature = 60 + 460 = 520°R Pipe inside diameter = 20 – 2 × 0.500 = 19.0 in. Using Equation 2.34, we calculate the Reynolds number as 14.7 0.6 × 100 × 10 6 Re = 0.0004778 = 5,331,726 520 0.000008 × 19 The relative roughness =

600 × 10 −6 = 0.0000316 19

Using Equation 2.39, the friction factor is 0.0000316 2.51 = −2 Log10 + 3.7 5, 331, 726 f f

1

Solving by successive iteration, we get f = 0.0105 Therefore, the transmission factor F is found from Equation 2.42 as follows: F=

2 0.0105

= 19.53

It must be noted that the friction factor calculated above is the Darcy friction factor. The corresponding Fanning friction factor will be one-fourth the calculated value. When ﬂow rate is increased by 50%, the Reynolds number becomes, by proportion, Re = 1.5 × 5,331,726 = 7,997,589 The new friction factor from Equation 2.39 is 0.0000316 2.51 = −2 Log10 + 3.7 7, 997, 589 f f

1

Solving for f by successive iteration, we get f = 0.0103 The corresponding transmission factor is F=

Copyright 2005 by Taylor & Francis Group, LLC

2 0.0103

= 19.74

2785_C002.fm Page 53 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

53

Compared to the previous values of 0.0105 for the friction factor and 19.53 for the transmission factor, we see the following changes:

Decrease in friction factor =

0.0105 − 0.0103 = 0.019 or 1.9% 0.0105

Increase in transmission factor =

19.74 − 19.53 = 0.0108 or 1.08% 19.53

Thus, increasing the ﬂow rate by 50% reduces the friction factor by 1.9% and increases the transmission factor by 1.08%. Example 8 For a gas pipeline, ﬂowing 3 Mm3/day gas of speciﬁc gravity 0.6 and viscosity of 0.000119 Poise, calculate the friction factor and transmission factor considering a DN 400 pipeline, 10 mm wall thickness, and an internal roughness of 0.02 mm. The base temperature and base pressure are 15°C and 101 kPa, respectively. If the ﬂow rate is doubled, what is the impact on the friction factor and transmission factor? Solution The base temperature = 15 + 273 = 288 K Pipe inside diameter = 400 – 2 × 10 = 380 mm Using Equation 2.35, we calculate the Reynolds number as 101 0.6 × 3 × 10 6 Re = 0.5134 = 7,166,823 288 0.000119 × 380 The relative roughness =

0.02 = 0.0000526 380

Using Equation 2.39, the friction factor is 0.0000526 2.51 = −2 Log10 + . 3 7 7,166, 823 f f

1

Solving by iteration, we get f = 0.0111 Therefore, the transmission factor F is found from Equation 2.42 as follows: F=

Copyright 2005 by Taylor & Francis Group, LLC

2 0.0111

= 18.98

2785_C002.fm Page 54 Friday, April 1, 2005 3:23 PM

54

GAS PIPELINE HYDRAULICS

It must be noted that the friction factor calculated above is the Darcy friction factor. The corresponding Fanning friction factor will be one-fourth the calculated value. When the ﬂow rate is doubled, the Reynolds number becomes Re = 2 × 7,166,823 = 14,333,646 The new value of the friction factor from Equation 2.39 is 0.0000526 2.51 = −2 Log10 + 3.7 14, 333, 646 f f

1

Solving for f by successive iteration, we get f = 0.0109 and the transmission factor is F=

2 0.0109

= 19.16

Therefore, doubling the ﬂow rate increases the transmission factor and decreases the friction factor as follows: Decrease in friction factor =

0.0111 − 0.0109 = 0.018 or 1.8% 0.0111

Increase in transmission factor =

19.16 − 18.98 = 0.0095 or 0.95% 18.98

2.12 MODIFIED COLEBROOK-WHITE EQUATION The Colebrook-White equation discussed in the preceding section has been in use for many years in both liquid ﬂow and gas ﬂow. The U.S. Bureau of Mines, in 1956, published a report that introduced a modiﬁed form of the Colebrook-White equation. The modiﬁcation results in a higher friction factor and, hence, a smaller value of the transmission factor. Because of this, a conservative value of ﬂow rate is obtained due to the higher friction and pressure drop. The modiﬁed version of the ColebrookWhite equation for turbulent ﬂow is as follows: e 2.825 = −2 Log10 + f 3.7 D Re f

1

(2.46)

Rewriting Equation 2.46 in terms of the transmission factor, we get the following version of the modiﬁed Colebrook-White equation: e 1.4125F F = −4 Log10 + Re 3.7 D

Copyright 2005 by Taylor & Francis Group, LLC

(USCS and SI units)

(2.47)

2785_C002.fm Page 55 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

55

Since Re, f, and F are dimensionless, as long as consistent units are used for both e and D, the modiﬁed Colebrook equation is the same, regardless of the units employed. Therefore, in SI units, Equation 2.46 and Equation 2.47 are used with e and D expressed in mm. Upon comparing Equation 2.39 with Equation 2.46, it is seen that the difference between the Colebrook equation and the modiﬁed Colebrook equation lies in the second constant term within the square brackets. The constant 2.51 in Equation 2.39 is replaced with the constant 2.825 in Equation 2.46. Similarly, in the transmission factor equations, the modiﬁed equation has 1.4125 instead of 1.255 in the original Colebrook-White equation. Many commercial hydraulic simulation programs list both Colebrook-White equations. Some use only the original Colebrook-White equation. Example 9 For a gas pipeline, ﬂowing 100 MMSCFD gas of speciﬁc gravity 0.6 and viscosity of 0.000008 lb/ft-s, calculate, using the modiﬁed Colebrook-White equation, the friction factor and transmission factor assuming an NPS 20 pipeline, 0.500 in. wall thickness, and an internal roughness of 600 µ in. The base temperature and base pressure are 60°F and 14.7 psia, respectively. How do these numbers compare with those calculated, using the original Colebrook equation? Solution The base temperature = 60 + 460 = 520°R Pipe inside diameter = 20 – 2 × 0.500 = 19.0 in. Using Equation 2.34, we calculate the Reynolds number as 14.7 0.6 × 100 × 10 6 Re = 0.0004778 = 5,331,726 520 0.000008 × 19 The relative roughness is e 600 × 10 6 = = 3.16 × 10 −5 D 19 From Equation 2.46, the friction factor using the modiﬁed Colebrook equation is 0.0000316 2.825 = −2 Log10 + 3.7 5, 331, 726 f f

1

Solving by successive iteration, we get f = 0.0106

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 56 Friday, April 1, 2005 3:23 PM

56

GAS PIPELINE HYDRAULICS

Therefore, the transmission factor F is found from Equation 2.42 as follows: F=

2 0.0106

= 19.43

By comparing these results with the friction factor and the transmission factor calculated in Example 7 using the unmodiﬁed Colebrook equation, it can be seen that the modiﬁed friction factor is approximately 0.95% higher than that calculated using the original Colebrook-White equation, whereas the transmission factor is approximately 0.51% lower than that calculated using the original Colebrook-White equation. Example 10 A gas pipeline, NPS 20 with 0.500 in. wall thickness, ﬂows 200 MMSCFD gas of speciﬁc gravity 0.6 and viscosity of 0.000008 lb/ft-s. Using the modiﬁed ColebrookWhite equation, calculate the pressure drop in a 50 mi segment of pipe, based on an upstream pressure of 1000 psig. Assume an internal pipe roughness of 600 µ in. and the base temperature and base pressure of 60°F and 14.73 psia, respectively. Neglect elevation effects and use 60°F for gas ﬂowing temperature and compressibility factor Z = 0.88. Solution Inside diameter of pipe = 20 – 2 × 0.5 = 19.0 in. The base temperature = 60 + 460 = 520 °R Gas ﬂow temperature = 60 + 460 = 520 °R First, we calculate the Reynolds number using Equation 2.34. 14.73 0.6 × 200 × 10 6 Re = 0.0004778 = 10,685,214 520 0.000008 × 19 The transmission factor F is calculated from Equation 2.47 as follows: 600 × 10 −6 1.4125F + F = −4 Log10 10, 685, 214 3.7 × 19 Solving for F by successive iteration, F = 19.81 Next, using General Flow Equation 2.4, we calculate the downstream pressure P2 as follows: 60 + 460 1014.732 − P2 2 200 × 10 6 = 38.77 × 19.81 14.73 0.6 × 520 × 50 × 0.88

Copyright 2005 by Taylor & Francis Group, LLC

0.5

× 192.5

2785_C002.fm Page 57 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

57

Solving for P2, we get P2 = 853.23 psia = 838.5 psig Therefore, the pressure drop = 1014.73 – 853.23 = 161.5 psi.

2.13 AMERICAN GAS ASSOCIATION (AGA) EQUATION In 1964 and 1965, the American Gas Association (AGA) published a report on how to calculate the transmission factor for gas pipelines to be used in the General Flow equation. This is sometimes referred to as the AGA NB-13 method. Using the method outlined in this report, the transmission factor F is calculated using two different equations. First, F is calculated for the rough pipe law (referred to as the fully turbulent zone). Next, F is calculated based on the smooth pipe law (referred to as the partially turbulent zone). Finally, the smaller of the two values of the transmission factor is used in the General Flow Equation 2.4 for pressure drop calculation. Even though the AGA method uses the transmission factor F instead of the friction factor f, we can still calculate the friction factor using the relationship shown in Equation 2.42. For the fully turbulent zone, AGA recommends using the following formula for F, based on relative roughness e/D and independent of the Reynolds number: 3.7 D F = 4 Log10 e

(2.48)

Equation 2.48 is also known as the Von Karman rough pipe ﬂow equation. For the partially turbulent zone, F is calculated from the following equations using the Reynolds number, a parameter Df known as the pipe drag factor, and the Von Karman smooth pipe transmission factor Ft: Re F = 4 D f Log10 1.4125Ft

(2.49)

Re Ft = 4 Log10 − 0.6 Ft

(2.50)

and

where Ft = Von Karman smooth pipe transmission factor Df = pipe drag factor that depends on the Bend Index (BI) of the pipe The pipe drag factor Df is a parameter that takes into account the number of bends, ﬁttings, etc. Its value ranges from 0.90 to 0.99. The Bend index is the sum of all the angles and bends in the pipe segment, divided by the total length of the pipe section under consideration. BI =

Copyright 2005 by Taylor & Francis Group, LLC

total degrees of all bends in pipe sectionn total length of pipe section

(2.51)

2785_C002.fm Page 58 Friday, April 1, 2005 3:23 PM

58

GAS PIPELINE HYDRAULICS

Table 2.2 Bend Index and Drag Factor

Bare steel Plastic lined Pig burnished Sand blasted

Extremely Low 5° to 10°

Bend Index Average 60° to 80°

Extremely High 200° to 300°

0.975–0.973 0.979–0.976 0.982–0.980 0.985–0.983

0.960–0.956 0.964–0.960 0.968–0.965 0.976–0.970

0.930–0.900 0.936–0.910 0.944–0.920 0.951–0.930

Note: The drag factors above are based on 40-ft joints of pipelines and mainline valves at 10-mile spacing.

The value of Df is generally chosen from Table 2.2. For further discussion on the bend index and drag factor, the reader is referred to Steady Flow in Gas Pipelines listed in the Reference section. Example 11 Using the AGA method, calculate the transmission factor and friction factor for gas ﬂow in an NPS 20 pipeline with 0.500 in. wall thickness. The ﬂow rate is 200 MMSCFD, gas gravity = 0.6, and viscosity = 0.000008 lb/ft-sec. The absolute pipe roughness is 700 µ in. Assume a bend index of 60°, base pressure of 14.73 psia, and base temperature of 60°F. Solution Inside diameter of pipe = 20 – 2 × 0.5 = 19.0 in. The base temperature = 60 + 460 = 520°R We will ﬁrst calculate the Reynolds number using Equation 2.34. Re =

0.0004778 × 200 × 10 6 × 0.6 × 14.73 = 10,685,214 19 × 0.000008 × 520

Next, calculate the two transmission factors. The fully turbulent transmission factor, using Equation 2.48, is 3.7 × 19 F = 4 Log10 = 20.01 0.0007 For the smooth pipe zone, using Equation 2.50, the Von Karman transmission factor is 10, 685, 214 Ft = 4 Log10 − 0.6 Ft

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 59 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

59

Solving this equation by trial and error, we get Ft = 22.13. From Table 2.2, for a bend index of 60°, the drag factor Df is 0.96. Therefore, for the partially turbulent ﬂow zone, using Equation 2.49, the transmission factor is 10, 685, 214 = 21.25 F = 4 × 0.96 Log10 1.4125 × 22.13 From the above two values of F, using the smaller number, we get the AGA transmission factor as F = 20.01 Therefore, the corresponding friction factor f is found from Equation 2.42 as 2 f

= 20.01

or f = 0.0100 Example 12 Using the AGA method, calculate the transmission factor and friction factor for gas ﬂow in a DN 500 pipeline with 12 mm wall thickness. The ﬂow rate is 6 Mm3/day, gas gravity = 0.6, and viscosity = 0.00012 Poise. The absolute pipe roughness is 0.02 mm. Assume a bend index of 60°, base pressure of 101 kPa, and base temperature of 15°C. For a 60 km pipe length, calculate the upstream pressure needed to hold a downstream pressure of 5 MPa (absolute). Assume ﬂow temperature = 20°C and compressibility factor Z = 0.85. Neglect elevation effects. Solution Inside diameter of pipe = 500 – 2 × 12 = 476 mm The base temperature = 15 + 273 = 288 K Gas ﬂowing temperature = 20 + 273 = 293 K We ﬁrst calculate the Reynolds number from Equation 2.35. 101 0.6 × 6 × 10 6 = 11,347,470 Re = 0.5134 288 0.00012 × 476

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 60 Friday, April 1, 2005 3:23 PM

60

GAS PIPELINE HYDRAULICS

Next, calculate the two transmission factors as follows: The fully turbulent transmission factor, using Equation 2.48, is 3.7 × 476 F = 4 Log10 = 19.78 0.02 For the smooth pipe zone, using Equation 2.50, the Von Karman transmission factor is 11, 347, 470 Ft = 4 Log10 − 0.6 Ft Solving by successive iteration, we get Ft = 22.23 From Table 2.2, for a bend index of 60°, the drag factor is 0.96. Therefore, for the partially turbulent ﬂow zone, using Equation 2.49, the transmission factor is 11, 347, 470 = 21.34 F = 4 × 0.96 Log10 1.4125 × 22.23 Using the smaller of the two values of F, the AGA transmission factor is F = 19.78 Therefore, the corresponding friction factor is found from Equation 2.42 as 2 f

= 19.78

or f = 0.0102 Using the General Flow Equation 2.8, we calculate the upstream pressure P1 as follows: 288 P12 − 5000 2 6 × 10 6 = 5.747 × 10 −4 × 19.78 × 101 0.6 × 293 × 60 × 0.85 Solving for P1, we get P1 = 6130 kPa = 6.13 MPa

Copyright 2005 by Taylor & Francis Group, LLC

0.5

× 4762.5

2785_C002.fm Page 61 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

61

2.14 WEYMOUTH EQUATION The Weymouth equation is used for high pressure, high ﬂow rate, and large diameter gas gathering systems. This formula directly calculates the ﬂow rate through a pipeline for given values of gas gravity, compressibility, inlet and outlet pressures, pipe diameter, and length. In USCS units, the Weymouth equation is stated as follows: T P 2 − e s P22 Q = 433.5E b 1 Pb GT f Le Z where Q = E = Pb = Tb = P1 = P2 = G = Tf = Le = Z = D =

0.5

D 2.667

(2.52)

volume ﬂow rate, standard ft3/day (SCFD) pipeline efﬁciency, a decimal value less than or equal to 1.0 base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂow temperature, °R (460 + °F) equivalent length of pipe segment, mi gas compressibility factor, dimensionless pipe inside diameter, in.

where the equivalent length Le and s were deﬁned earlier in Equation 2.9 and Equation 2.10. By comparing the Weymouth equation with the General Flow equation, we can isolate an equivalent transmission factor as follows: The Weymouth transmission factor in USCS units is F = 11.18(D)1/6

(2.53)

In SI units, the Weymouth equation is as follows: T P 2 − e s P22 Q = 3.7435 × 10 E b 1 Pb GT f Le Z −3

where Q = Tb = Pb = Tf = P1 = P2 = Le =

gas ﬂow rate, standard m3/day base temperature, K (273 + °C) base pressure, kPa average gas ﬂow temperature, K (273 + °C) upstream pressure, kPa downstream pressure, kPa equivalent length of pipe segment, km

Other symbols are as deﬁned previously.

Copyright 2005 by Taylor & Francis Group, LLC

0.5

D 2.667

(2.54)

2785_C002.fm Page 62 Friday, April 1, 2005 3:23 PM

62

GAS PIPELINE HYDRAULICS

The Weymouth transmission factor in SI units is F = 6.521(D)1/6

(2.53a)

You will notice that a pipeline efﬁciency factor, E, is used in the Weymouth equation so we can compare the throughput performance of a pipeline using the General Flow equation that does not include an efﬁciency factor. Example 13 Calculate the ﬂow rate using the Weymouth equation in a gas pipeline system, 15 miles long, NPS 12 pipe with 0.250 in. wall thickness, at an efﬁciency of 0.95. The upstream pressure is 1200 psia, and the delivery pressure required at the end of the pipe segment is 750 psia. Use gas gravity = 0.59 and viscosity = 0.000008 lb/ft-sec. The ﬂowing temperature of gas = 75°F, base pressure = 14.7 psia, and base temperature = 60°F. Assume compressibility factor to be 0.94. Neglect elevation difference along the pipe. How does this compare with the ﬂow rate calculated using the General Flow equation with the Colebrook friction factor? Assume a pipe roughness of 700 µ in. Solution Using Equation 2.52, we get the ﬂow rate for the Weymouth equation as follows: 60 + 460 1200 2 − 750 2 Q = 433.5 × 0.95 14.7 0.59 × (75 + 460) × 15 × 0.94

0.5

× 12.252.667

Q = 163,255,858 SCFD or Q = 163.26 MMSCFD Next, we will calculate the Reynolds number using Equation 2.34. Re =

0.0004778 × Q × 0.59 × 14.7 12.25 × 0.000008 × 520

where Q is the ﬂow rate in SCFD. Simplifying, we get Re = 0.0813 Q. Since Q is unknown, we will ﬁrst assume a transmission factor F = 20 and calculate the ﬂow rate from the General Flow Equation 2.4. 520 1200 2 − 750 2 Q = 38.77 × 20 14.7 0.59 × 535 × 15 × 0.94

Copyright 2005 by Taylor & Francis Group, LLC

0.5

× 12.252.5 = 202,284,747 SCFD

2785_C002.fm Page 63 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

63

or Q = 202.28 MMSCFD Next, we will calculate the Reynolds number and the transmission factor based on this ﬂow rate as Re = 0.0813 × 202,284,747 = 16.45 million and, using Equation 2.45, 700 × 10 −6 1.255F + F = − 4 Log10 3 7 × 12 25 . . 16.45 × 10 6 Solving for F, we get F = 19.09 Using this value, the revised ﬂow rate is found by proportion as Q = 202.28 ×

19.09 = 193.08 MMSCFD 20

Repeating the calculation of Re and F, we get Re = 16.45 ×

193.08 = 15.7 million 202.28

and 700 × 10 −6 1.255F + F = − 4 Log10 6 3.7 × 12.25 15.7 × 10 Therefore, F = 19.08. This is fairly close to the previous value of F = 19.09; therefore, we will use this value and calculate the ﬂow rate as Q = 202.28 ×

19.08 = 192.98 MMSCFD 20

Comparing this result using the General Flow equation with that calculated using the Weymouth equation, we see that the latter equation is quite conservative. Example 14 A natural gas transmission line transports 30 million m3/day of gas from a processing plant to a compressor station site 100 km away. The pipeline can be assumed to be along a ﬂat terrain. Calculate the minimum pipe diameter required such that the maximum pipe operating pressure is limited to 8500 kPa. The delivery pressure desired at the end of the pipeline is a minimum of 5500 kPa. Assume a pipeline efﬁciency of 0.95. The gas gravity is 0.65, and the gas temperature is 18°C. Use the Weymouth equation, considering a base temperature = 15°C and base pressure 101 kPa. The gas compressibility factor Z = 0.92.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 64 Friday, April 1, 2005 3:23 PM

64

GAS PIPELINE HYDRAULICS

Solution The base temperature = 15 + 273 = 288 K The gas ﬂowing temperature = 18 + 273 = 291 K We will assume that given pressures are absolute values. Upstream pressure = 8500 kPa (absolute) Downstream pressure = 5500 kPa (absolute) Using the Weymouth Equation 2.52 and substituting given values, we get 288 8500 2 − 5500 2 30 × 10 6 = 3.7435 × 10 −3 × 0.95 × 101 0.65 × 291 × 100 × 0.92

0.5

× D 2.667

Solving for diameter, D, we get D = 826.1 mm Therefore, the minimum diameter required will be DN 850 with 10 mm wall thickness.

2.15 PANHANDLE A EQUATION The Panhandle A Equation was developed for use in natural gas pipelines, incorporating an efﬁciency factor for Reynolds numbers in the range of 5 to 11 million. In this equation, the pipe roughness is not used. The general form of the Panhandle A equation is expressed in USCS units as follows: T Q = 435.87 E b P b

where Q = E = Pb = Tb = P1 = P2 = G = Tf =

1.0788

P 2 − es P 2 1 2 0.8539 T f Le Z G

0.5394

D 2.6182

volume ﬂow rate, standard ft3/day (SCFD) pipeline efﬁciency, a decimal value less than 1.0 base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂow temperature, °R (460 + °F)

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units) (2.55)

2785_C002.fm Page 65 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

65

Le = equivalent length of pipe segment, mi Z = gas compressibility factor, dimensionless D = pipe inside diameter, in. Other symbols are as deﬁned previously. In SI units, the Panhandle A equation is T Q = 4.5965 × 10 E b P −3

b

where Q = E = Tb = Pb = Tf = P1 = P2 = Le =

1.0788

P 2 − es P 2 1 2 0.8539 T f Le Z G

0.5394

D 2.6182

(SI units)

(2.56)

gas ﬂow rate, standard m3/day pipeline efﬁciency, a decimal value less than 1.0 base temperature, K (273 + °C) base pressure, kPa average gas ﬂow temperature, K (273 + °C) upstream pressure, kPa (absolute) downstream pressure, kPa (absolute) equivalent length of pipe segment, km

Other symbols are as deﬁned previously. Due to the exponents involved in this equation, all pressures must be in kPa. By comparing the Panhandle A equation with the General Flow equation, we can calculate an equivalent transmission factor in USCS units as follows: QG F = 7.2111E D

0.07305

(USCS)

(2.57)

(SI)

(2.58)

and in SI units, it is QG F = 11.85E D

0.07305

Sometimes the transmission factor is used to compare the results of calculations using the General Flow equation and the Panhandle A equation. Example 15 Using the Panhandle A equation, calculate the outlet pressure in a natural gas pipeline, NPS 16 with 0.250 in. wall thickness, 15 miles long. The gas ﬂow rate is 100 MMSCFD at an inlet pressure of 1000 psia. The gas gravity = 0.6 and viscosity = 0.000008 lb/ft-sec. The average gas temperature is 80°F. Assume base pressure = 14.73 psia and base temperature = 60°F. For compressibility factor Z, use the CNGA method. Assume pipeline efﬁciency of 0.92.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 66 Friday, April 1, 2005 3:23 PM

66

GAS PIPELINE HYDRAULICS

Solution The average pressure, Pavg, needs to be calculated before the compressibility factor Z can be determined. Since the inlet pressure P1 = 1,000 psia, and the outlet pressure P2 is unknown, we will have to assume a value of P2 (such as 800 psia) and calculate Pavg and then calculate the value of Z. Once Z is known, from the Panhandle A equation we can calculate the outlet pressure P2. Using this value of P2, a better approximation for Z is calculated from a new Pavg. This process is repeated until successive values of P2 are within allowable limits, such as 0.5 psia. Assume P2 = 800 psia. The average pressure from Equation 2.14 is Pavg =

2 1000 × 800 = 903.7 psia 1000 + 800 − 3 1000 + 800

Next, we calculate the compressibility factor Z using the CNGA method. From Equation 1.34, Z=

1 5 1.785× 0.6 1 + (903.7−14.73)×3.444×310.825×(10) (80+ 460)

or Z = 0.8869 From Panhandle A Equation 2.55, substituting given values, neglecting elevations, we get 60 + 460 100 × 10 6 = 435.87 × 0.92 14.73

1.0788

1000 2 − P22 0.8539 (540 × 15 × 0.8869) (0.6)

5394 0.5

(15.5)2.6182

Solving for P2, we get P2 = 968.02 psia Since this is different from the assumed value of P2 = 800, we recalculate the average pressure and Z using P2 = 968.02 psia The revised average pressure is

Pavg =

Copyright 2005 by Taylor & Francis Group, LLC

2 1000 × 968.02 = 984.10 psia 1000 + 968.02 − 3 1000 + 968.02

2785_C002.fm Page 67 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

67

Using this value of Pavg, we recalculate Z as Z=

1 1.785× 0.6

×(10) 1 + (984.10−14.73)×3.444×310 .825 5

(80+ 460)

or Z = 0.8780 Recalculating P2 from the Panhandle A Equation 2.55, we get 60 + 460 100 × 10 = 435.87 × 0.92 14.73

1.0788

6

1000 2 − P22 0.8539 (540 × 15 × 0.8780) (0.6)

0.55394

(15.5)2.6182

Solving for P2, we get P2 = 968.35 psia This is within 0.5 psi of the previously calculated value. Hence, we will not continue the iteration any further. Therefore, the outlet pressure is 968.35 psia.

Example 16 Using the Panhandle A equation, calculate the inlet pressure required in a natural gas pipeline, DN 300 with 6 mm wall thickness, 24 km long, for a gas ﬂow rate of 3.5 Mm3/day. The gas gravity = 0.6 and viscosity = 0.000119 Poise. The average gas temperature is 20°C. The delivery pressure is 6000 kPa (absolute). Assume base pressure = 101 kPa, base temperature = 15°C, and compressibility factor Z = 0.90, with a pipeline efﬁciency of 0.92. Solution Pipe inside diameter D = 300 – 2 × 6 = 288 mm Gas ﬂow temperature = 20 + 273 = 293 K Using Panhandle A Equation 2.56 and neglecting elevation effect, we substitute 15 + 273 3.5 × 10 6 = 4.5965 × 10 −3 × 0.92 101

Copyright 2005 by Taylor & Francis Group, LLC

1.0788

P12 − 6000 2 0.8539 (293 × 24 × 0.9) (0.6)

0.55394

(288)2.6182

2785_C002.fm Page 68 Friday, April 1, 2005 3:23 PM

68

GAS PIPELINE HYDRAULICS

Solving for inlet pressure, we get P 21 – (6000)2 = 19,812,783 or P1 = 7471 kPa (absolute)

2.16 PANHANDLE B EQUATION The Panhandle B equation, also known as the revised Panhandle equation, is used in large diameter, high pressure transmission lines. In fully turbulent ﬂow, it is found to be accurate for values of Reynolds number in the range of 4 to 40 million. This equation in USCS units is as follows: T Q = 737 E b P b

where Q = E = Pb = Tb = P1 = P2 = G = Tf = Le = Z = D =

1.02

P 2 − es P 2 1 2 0.961 G T L f eZ

0.51

D 2.53

(USCS units)

(2.59)

volume ﬂow rate, standard ft3/day (SCFD) pipeline efﬁciency, a decimal value less than 1.0 base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂow temperature, °R (460 + °F) equivalent length of pipe segment, mi gas compressibility factor, dimensionless pipe inside diameter, in.

Other symbols are as deﬁned previously. In SI units, the Panhandle B equation is T Q = 1.002 × 10 E b P −2

b

where Q = E = Tb = Pb = Tf =

1.02

P 2 − es P 2 1 2 0.961 G T f Le Z

0.51

D 2.53

gas ﬂow rate, standard m3/day pipeline efﬁciency, a decimal value less than 1.0 base temperature, K (273 + °C) base pressure, kPa average gas ﬂow temperature, K (273 + °C)

Copyright 2005 by Taylor & Francis Group, LLC

(SI units)

(2.60)

2785_C002.fm Page 69 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

P1 P2 Le Z

= = = =

69

upstream pressure, kPa (absolute) downstream pressure, kPa (absolute) equivalent length of pipe segment, km gas compressibility factor at the ﬂowing temperature, dimensionless

Other symbols are as deﬁned previously. The equivalent transmission factor for the Panhandle B equation in USCS is given by QG F = 16.7 E D

0.01961

(USCS units)

(2.61)

In SI units, it is QG F = 19.08 E D

0.01961

(SI units)

(2.62)

Example 17 Using the Panhandle B equation, calculate the outlet pressure in a natural gas pipeline, NPS 16 with 0.250 in. wall thickness, 15 miles long. The gas ﬂow rate is 100 MMSCFD at 1000 psia inlet pressure. The gas gravity = 0.6 and viscosity = 0.000008 lb/ft-sec. The average gas temperature is 80°F. Assume base pressure = 14.73 psia and base temperature = 60°F. The compressibility factor Z = 0.90 and pipeline efﬁciency is 0.92. Solution Inside diameter of pipe = 16 – 2 × 0.25 = 15.5 in. Gas ﬂow temperature = 80 + 460 = 540°R Using Panhandle B Equation 2.59, substituting the given values, we get 60 + 460 100 × 10 6 = 737 × 0.92 14.73

1.02

1000 2 − P22 0.961 (0.6) (540 × 15 × 0.90)

0.51

15.52.53

Solving for P2, we get 10002 – P22 = 60,778 P2 = 969.13 psia Compare this with the results of Panhandle A equation in Example 15, where the outlet pressure P2 = 968.35 psia. Therefore, the Panhandle B equation gives a slightly lower pressure drop compared to that from the Panhandle A equation. In other words, Panhandle A is more conservative and will give a lower ﬂow rate for the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 70 Friday, April 1, 2005 3:23 PM

70

GAS PIPELINE HYDRAULICS

same pressures compared to Panhandle B. In this example, we use the constant value of Z = 0.9, whereas in example 15, Z was calculated using the CNGA equation as Z = 0.8780. If we factor this in, the result for the outlet pressure in this example will be 969.9 psia, which is not too different from the calculated value of 969.13 psia. Example 18 Using the Panhandle B equation, calculate the inlet pressure in a natural gas pipeline, DN 300 with 6 mm wall thickness, 24 km long. The gas ﬂow rate is 3.5 Mm3/day, gas gravity = 0.6, and viscosity = 0.000119 Poise. The average gas temperature is 20°C, and the delivery pressure is 6,000 kPa (absolute). Assume base pressure = 101 kPa, base temperature = 15°C, and compressibility factor Z = 0.90. The pipeline efﬁciency is 0.92. Solution Inside diameter of pipe = 300 – 2 × 6 = 288 mm Gas ﬂow temperature = 20 + 273 = 293 K Neglecting elevations, using Panhandle B Equation 2.60, we get 15 + 273 3.5 × 10 6 = 1.002 × 10 −2 × 0.92 101

1.02

P12 − 6000 2 0.961 (0.6) (293 × 24 × 0.9)

0.51

2882.53

Solving for the inlet pressure P1, we get P21 – (6000) 2 = 19,945,469 P1 = 7480 kPa (absolute) Compare this with the results of the Panhandle A equation in Example 16, where the inlet pressure P1 = 7471 kPa (absolute). Again, we see that the Panhandle B equation gives a slightly lower pressure drop compared to that obtained from the Panhandle A equation.

2.17 INSTITUTE OF GAS TECHNOLOGY (IGT) EQUATION The IGT equation proposed by the Institute of Gas Technology is also known as the IGT distribution equation and is stated as follows for USCS units: T P 2 − e s P22 Q = 136.9 E b 01.8 Pb G T f Le µ 0.2

0.555

D 2.667

where Q = volume ﬂow rate, standard ft3/day (SCFD) E = pipeline efﬁciency, a decimal value less than 1.0

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(2.63)

2785_C002.fm Page 71 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

Pb Tb P1 P2 G Tf Le Z D m

= = = = = = = = = =

71

base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂow temperature, °R (460 + °F) equivalent length of pipe segment, mi gas compressibility factor, dimensionless pipe inside diameter, in. gas viscosity, lb/ft-s

Other symbols are as deﬁned previously. In SI units, the IGT equation is expressed as follows: T P 2 − e s P22 Q = 1.2822 × 10 E b 01.8 Pb G T f Le µ 0.2 −3

where Q = E = Tb = Pb = Tf = P1 = P2 = Le = m =

0.555

D 2.667

(SI units)

(2.64)

gas ﬂow rate, standard m3/day pipeline efﬁciency, a decimal value less than 1.0 base temperature, K (273 + °C) base pressure, kPa average gas ﬂow temperature, K (273 + °C) upstream pressure, kPa (absolute) downstream pressure, kPa (absolute) equivalent length of pipe segment, km gas viscosity, Poise

Other symbols are as deﬁned previously. Example 19 Using the IGT equation, calculate the ﬂow rate in a natural gas pipeline, NPS 16 with 0.250 in. wall thickness, 15 mi long. The inlet and outlet pressure are 1000 psig and 800 psig, respectively. The gas gravity = 0.6 and viscosity = 0.000008 lb/ft-s. The average gas temperature is 80°F, base pressure = 14.7 psia, and base temperature = 60°F. The compressibility factor Z = 0.90, and the pipeline efﬁciency is 0.95. Solution Inside diameter of pipe = 16 – 2 × 0.25 = 15.5 in. The pressures given are in psig, and they must be converted to absolute pressures.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 72 Friday, April 1, 2005 3:23 PM

72

GAS PIPELINE HYDRAULICS

Therefore, P1 = 1000 + 14.7 = 1014.7 psia P2 = 800 + 14.7 = 814.7 psia Tb = 60 + 460 = 520°R Tf = 80 + 460 = 540°R Substituting in IGT Equation 2.63, we get 520 1014.72 − 814.72 Q = 136.9 × 0.95 0.8 14.7 (0.6) × 540 × 15 × (8 × 10 −6 )0.2

0.555

15.52.667

Q = 263.1 × 106 ft3/day = 263.1 MMSCFD Therefore, the ﬂow rate is 263.1 MMSCFD. Example 20 A natural gas pipeline, DN 400 with 6 mm wall thickness, 24 km long, is used to transport gas at an inlet pressure of 7000 kPa (gauge) and an outlet pressure of 5500 kPa (gauge). The gas gravity = 0.6 and viscosity = 0.000119 Poise. The average gas temperature is 20°C. Assume base pressure = 101 kPa and base temperature = 15°C. The compressibility factor Z = 0.90 and pipeline efﬁciency is 0.95. a) Calculate the ﬂow rate using the IGT equation. b) What are the gas velocities at inlet and outlet? c) If the velocity must be limited to 10 m/s, what should the minimum pipe size be, assuming the ﬂow rate and inlet pressure remain constant? Solution Inside diameter of pipe D = 400 – 2 × 6 = 388 mm All pressures are given in gauge values and must be converted to absolute values. Inlet pressure P1 = 7000 + 101 = 7101 kPa (absolute) Outlet pressure P2 = 5500 + 101 = 5601 kPa (absolute) Base temperature Tb = 15 + 273 = 288 K Flowing temperature Tf = 20 + 273 = 293 K From IGT Equation 2.64, we get the ﬂow rate in m3/day as 288 71012 − 56012 Q = 1.2822 × 10 × 0.95 0.8 101 (0.6) × 293 × 24 × (1.19 × 10 −4 )0.2 −3

Copyright 2005 by Taylor & Francis Group, LLC

0.555

(388)2.667

2785_C002.fm Page 73 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

73

or Q = 7,665,328 m3/day = 7.67 Mm3/day (a) Therefore, the ﬂow rate is 7.67 Mm3/day. (b) Using Equation 2.29, we calculate the average velocity of the gas at the inlet pressure as 7.67 × 10 6 101 0.9 × 293 Inlet velocity u1 = 14.7349 288 7101 = 9.78 m/s 2 388 In the preceding, we assumed a constant compressibility factor, Z = 0.9. Similarly, at the outlet pressure, the average gas velocity is 7.67 × 10 6 101 0.9 × 293 Outlet velocity u2 = 14.7349 288 5601 = 12.4 m/s 2 388 (c) Since the velocity must be limited to 10 m/s, the pipe diameter must be increased. Increasing the pipe diameter will also increase the outlet pressure if we keep both the ﬂow rate and inlet pressure the same as before. The increased outlet pressure will also reduce the gas velocity as can be seen from Equation 2.29. We will try a DN 450 pipe with 10 mm wall thickness. Inside diameter of pipe D = 450 – 2 × 10 = 430 mm Assuming P1 and Q are the same as before, we calculate the new outlet pressure P2 from IGT Equation 2.64 as 288 71012 − P22 7.67 × 10 6 = 1.2822 × 10 −3 × 0.95 −4 0.2 0.8 101 (0.6) × 293 × 24 × (1.19 × 10 )

0.555

(430)2.667

Solving for P2, we get P2 = 6228 kPa The new velocity at the outlet will be 7.67 × 10 6 101 0.9 × 293 u2 = 14.7349 = 9.08 m/s 430 2 288 6228 Since this is less than the 10 m/s speciﬁed, the DN 450 pipe is satisfactory. In the preceding calculations we assumed the same compressibility factor for both inlet and outlet pressures. Actually, a more nearly correct solution would be to calculate Z using the CNGA equation at both inlet and outlet conditions and using these values in the calculation of gas velocities. This is left as an exercise for the reader.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 74 Friday, April 1, 2005 3:23 PM

74

GAS PIPELINE HYDRAULICS

2.18 SPITZGLASS EQUATION The Spitzglass equation has been around for many years and originally was used in fuel gas piping calculations. There are two versions of the Spitzglass equation. One equation is for low pressure (less than or equal to 1 psig) and another is for high pressure (more than 1 psig). These equations have been modiﬁed to include a pipeline efﬁciency and compressibility factor. The low-pressure (less than or equal to 1 psig) version of the Spitzglass equation in USCS units is T P1 − P2 Q = 3.839 × 10 E b Pb GT f Le Z 1 + 3D.6 + 0.03D 3

(

)

0.5

D 2.5

(USCS units) (2.65)

where Q = E = Pb = Tb = P1 = P2 = G = Tf = Le = D = Z =

volume ﬂow rate, standard ft3/day (SCFD) pipeline efﬁciency, a decimal value less than 1.0 base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂow temperature, °R (460 + °F) equivalent length of pipe segment, mi pipe inside diameter, in. gas compressibility factor, dimensionless

Other symbols are as deﬁned previously. The low-pressure (less than 6.9 kPa) version of the Spitzglass equation in SI units is T P1 − P2 Q = 5.69 × 10 E b 91.44 P + + 0.0012 D GT L Z 1 b D f e −2

(

)

0.5

D 2.5

(SI units) (2.66)

where Q = E = Tb = Pb = P1 = P2 = G = Tf =

gas ﬂow rate, standard m3/day pipeline efﬁciency, a decimal value less than 1.0 base temperature, K (273 + °C) base pressure, kPa upstream pressure, kPa (absolute) downstream pressure, kPa (absolute) gas gravity (air = 1.00) average gas ﬂow temperature, K (273 + °C)

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 75 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

75

Le = equivalent length of pipe segment, km Z = gas compressibility factor, dimensionless Other symbols are as deﬁned previously. The high-pressure (more than 1 psig) version in USCS units is as follows. T P12 − e s P22 Q = 729.6087 E b Pb GT f Le Z 1 + 3D.6 + 0.03D

(

)

0.5

D 2.5

(USCS units) (2.67)

where Q = E = Pb = Tb = P1 = P2 = G = Tf = Le = D = Z =

volume ﬂow rate, standard ft3/day (SCFD) pipeline efﬁciency, a decimal value less than 1.0 base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂow temperature, °R (460 + °F) equivalent length of pipe segment, mi pipe inside diameter, in. gas compressibility factor, dimensionless

Other symbols are as deﬁned previously. In SI units, the high-pressure (more than 6.9 kPa) version of the Spitzglass equation is T P12 − e s P22 Q = 1.0815 × 10 E b 91 . 44 Pb GT f Le Z 1 + D + 0.0012 D −2

(

)

0.5

D 2.5

(SI units) (2.68)

where Q = E = Tb = Pb = P1 = P2 = G = Tf = Le = Z =

gas ﬂow rate, standard m3/day pipeline efﬁciency, a decimal value less than 1.0 base temperature, K (273 + °C) base pressure, kPa upstream pressure, kPa (absolute) downstream pressure, kPa (absolute) gas gravity (air = 1.00) average gas ﬂow temperature, K (273 + °C) equivalent length of pipe segment, km gas compressibility factor, dimensionless

Other symbols are as deﬁned previously.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 76 Friday, April 1, 2005 3:23 PM

76

GAS PIPELINE HYDRAULICS

Example 21 Calculate the fuel gas capacity of an NPS 6 pipe, with an inside diameter of 6.065 in. and a total equivalent length of 180 ft. The ﬂowing temperature of fuel gas is 60°F, and the inlet pressure is 1.0 psig. Consider a pressure drop of 0.7 in the water column and the speciﬁc gravity of gas = 0.6. Assume pipeline efﬁciency E = 1.0 and compressibility factor Z = 1.0. The base pressure and base temperature are 14.7 psia and 60°F, respectively. Solution Base temperature = 60 + 460 = 520°R Gas ﬂowing temperature = 60 + 460 = 520°R Pressure drop (P1 – P2) = 1.0 −

0.7 × 0.433 = 0.9747 psi 12

Since this is low pressure, using Spitzglass Equation 2.65, we get 520 0.9747 Q = 3.839 × 10 × 1.0 × 180 × 1.0 1 + 3.6 + 0.03 × 6.065 14.7 0.6 × 520 × 5280 6.065 3

(

)

0.5

6.0652.5

Q = 2,794,842 SCFD = 2.79 MMSCFD Therefore, the fuel gas capacity is 2.79 MMSCFD.

2.19 MUELLER EQUATION The Mueller equation is another form of the ﬂow rate vs. pressure relationship in gas pipelines. In USCS units, it is expressed as follows: T P12 − e s P22 Q = 85.7368 E b 0.7391 0.2609 Pb G T f Le µ where Q = E = Pb = Tb = P1 = P2 = G =

0.575

D 2.725

volume ﬂow rate, standard ft3/day (SCFD) pipeline efﬁciency, a decimal value less than 1.0 base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00)

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(2.69)

2785_C002.fm Page 77 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

Tf Le D m

= = = =

77

average gas ﬂow temperature, °R (460 + °F) equivalent length of pipe segment, mi pipe inside diameter, in. gas viscosity, lb/ft-s

Other symbols are as deﬁned previously. In SI units, the Mueller equation is as follows: T P12 − e s P22 Q = 3.0398 × 10 −2 E b 0.7391 0.2609 Pb G T f Le µ where Q = E = Tb = Pb = P1 = P2 = G = Tf = Le = m =

0.575

D 2.725

(SI units)

(2.70)

gas ﬂow rate, standard m3/day pipeline efﬁciency, a decimal value less than 1.0 base temperature, K (273 + °C) base pressure, kPa upstream pressure, kPa (absolute) downstream pressure, kPa (absolute) gas gravity (air = 1.00) average gas ﬂow temperature, K (273 + °C) equivalent length of pipe segment, km gas viscosity, cP

Other symbols are as deﬁned previously.

2.20 FRITZSCHE EQUATION The Fritzsche formula, developed in Germany in 1908, has found extensive use in compressed air and gas piping. In USCS units, it is expressed as follows: T P2 − P2 Q = 410.1688 E b 01.8587 2 Pb G T f Le

0.538

D 2.69

(USCS units)

(2.71)

D 2.69

(SI units)

(2.72)

All symbols are as deﬁned before. In SI units, T P 2 − es P 2 Q = 2.827 E b 10.8587 2 Pb G T f Le All symbols are as deﬁned before.

Copyright 2005 by Taylor & Francis Group, LLC

0.538

2785_C002.fm Page 78 Friday, April 1, 2005 3:23 PM

78

GAS PIPELINE HYDRAULICS

2.21 EFFECT OF PIPE ROUGHNESS In the preceding sections, we used the pipe roughness as a parameter in the friction factor and transmission factor calculations. Both the AGA and Colebrook-White equations use the pipe roughness, whereas the Panhandle and Weymouth equations do not use the pipe roughness directly in the calculations. Instead, these equations use a pipeline efﬁciency to compensate for the internal conditions and age of the pipe. Therefore, when comparing the predicted ﬂow rates or pressures using the AGA or Colebrook-White equations with the Panhandle or Weymouth equations, we can adjust the pipeline efﬁciency to correlate with the pipe roughness used in the former equations. Since most gas pipelines operate in the turbulent zone, the laminar ﬂow friction factor, which is independent of pipe roughness, is of little interest to us. Concentrating, therefore, on turbulent ﬂow, we see that Colebrook-White Equation 2.45 is affected by variation in pipe internal roughness. For example, suppose we want to compare an internally coated pipeline with an uncoated pipeline. The internal roughness of the coated pipe might be in the range of 100 to 200 µin., whereas the uncoated pipe might have a roughness of 600 to 800 µin. or more. If the pipe is NPS 20 with a 0.500 in. wall thickness, the relative roughness using the lower roughness value is as follows: For coated pipe, e 100 × 10 −6 = = 5.263 × 10 −6 D 19 and For uncoated pipe, e 600 × 10 −6 = = 3.1579 × 10 −5 D 19 Substituting these values of relative roughness in Equation 2.45 and using a Reynolds number of 10 million, we calculate the following transmission factors: F = 21.54 for coated pipe and F = 19.83 for uncoated pipe Since the ﬂow rate is directly proportional to the transmission factor F, from General Flow Equation 2.4 we see that the coated pipe will be able to transport 21.54 −19.83 = 0.086 = 8.6% more ﬂow rate than the uncoated pipe, if all other parameters 19.83 remain the same. This is true in the fully turbulent zone where the Reynolds number has little effect on the friction factor f and the transmission factor F. However, in the smooth

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 79 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

79

pipe zone, pipe roughness has less effect on the friction factor and the transmission factor. This is evident from the Moody diagram in Figure 2.3. Using a Reynolds number of 106, we ﬁnd from the Moody diagram in Figure 2.3, for coated pipe, that f = 0.0118

and F = 18.41

f = 0.0122

and F = 18.10

and, for the uncoated pipe,

Therefore, the increase in ﬂow rate in this case will be 18.41 − 18.10 = 0.017 = 1.7% 18.10 Thus, the impact of pipe roughness is less in the smooth pipe zone or for a lower Reynolds number. A similar comparison can be made using the AGA equation. Figure 2.4 shows the effect of pipe roughness on the pipeline ﬂow rate considering the AGA and Colebrook-White equations. The graph is based on NPS 20 pipe, 0.500 in. wall thickness, 120 miles long, with 1200 psig upstream pressure and 800 psig downstream pressure. The ﬂowing temperature of gas is 70°F. It can be seen that as the pipe roughness is increased from 200 to 800 µin., the ﬂow rate decreases from 224 MMSCFD to 206 MMSCFD for the Colebrook-White equation and from 220 MMSCFD to 196 MMSCFD for the AGA equation.

240

Flow rate, MMSCFD

230 220 Colebrook-White 210 AGA 200 190 180 200

400

600 Roughness, microinches

Figure 2.4

Effect of pipe roughness.

Copyright 2005 by Taylor & Francis Group, LLC

800

1000

2785_C002.fm Page 80 Friday, April 1, 2005 3:23 PM

80

GAS PIPELINE HYDRAULICS

We can therefore conclude that decreasing the pipe roughness directly results in a throughput increase in a pipeline. However, the cost of internally coating a pipe to reduce the pipe roughness must be weighed against the revenue increase due to enhanced ﬂow rate. We will revisit this issue in Chapter 10, when we discuss pipeline economics.

2.22 COMPARISON OF FLOW EQUATIONS In the preceding sections, we calculated the ﬂow rates and pressures in gas pipelines using the various ﬂow equations. Each equation is slightly different from the other, and some equations consider the pipeline efﬁciency while others use an internal pipe roughness value. How do these equations compare when predicting ﬂow rates through a given pipe size when the upstream or downstream pressure is held constant? Obviously, some equations will predict higher ﬂow rates for the same pressures than others. Similarly, if we start with a ﬁxed upstream pressure in a pipe segment at a given ﬂow rate, these equations will predict different downstream pressures. This indicates that some equations calculate higher pressure drops for the same ﬂow rate than others. Figure 2.5 and Figure 2.6 show some of these comparisons when using the AGA, Colebrook-White, Panhandle, and Weymouth equations. In Figure 2.5, we consider a pipeline 100 mi long, NPS 16 with 0.250 in. wall thickness, operating at a ﬂow rate of 100 MMSCFD. The gas ﬂowing temperature is 80°F. With the upstream pressure ﬁxed at 1400 psig, the downstream pressure was calculated using the different ﬂow equations. By examining Figure 2.5, it is clear that the highest pressure drop is predicted by the Weymouth equation and the lowest pressure drop is predicted by the Panhandle B equation. It must be noted that we

1450

Pressure, psig

1400 1350 1300 Panhandle B Panhandle A Colebrook-White AGA

1250 1200

Weymouth 1150 0.00

100.00 Distance, mi

Figure 2.5

Comparison of flow equations.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 81 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

81

Upstream pressure, psig

1200

Weymouth AGA Colebrook-White Panhandle B Panhandle A

1100

1000

900

800 200

300

400

500

600

Flow rates, MMSCFD Figure 2.6

Upstream pressures for various flow equations.

used a pipe roughness of 700 µin. for both the AGA and Colebrook equations, whereas a pipeline efﬁciency of 0.95 was used in the Panhandle and Weymouth equations. Figure 2.6 shows a comparison of the ﬂow equations from a different perspective. In this case, we calculated the upstream pressure required for an NPS 30 pipeline, 100 miles long, holding the delivery pressure constant at 800 psig. The upstream pressure required for various ﬂow rates, ranging from 200 to 600 MMSCFD, was calculated using the ﬁve ﬂow equations. Again it can be seen that the Weymouth equation predicts the highest upstream pressure at any ﬂow rate, whereas the Panhandle A equation calculates the least pressure. We therefore conclude that the most conservative ﬂow equation that predicts the highest pressure drop is the Weymouth equation and the least conservative ﬂow equation is Panhandle A.

2.23 SUMMARY In this chapter we introduced the various methods of calculating the pressure drop in a pipeline transporting gas and gas mixtures. The more commonly used equations for pressure drop vs. ﬂow rate and pipe size were discussed and illustrated using example problems. The effect of elevation changes was explained, and the concepts of the Reynolds number, friction factor, and transmission factor were introduced. The importance of the Moody diagram and how to calculate the friction factor for laminar and turbulent ﬂow were explained. We compared the more commonly used pressure drop equations, such as AGA, Colebrook-White, Weymouth, and Panhandle equations. The use of a pipeline efﬁciency in comparing various equations was illustrated using an example. The average velocity of gas ﬂow was introduced, and the limiting value of erosional velocity was discussed.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 82 Friday, April 1, 2005 3:23 PM

82

GAS PIPELINE HYDRAULICS

PROBLEMS 1. A gas pipeline, NPS 18 with 0.375 in. wall thickness, transports natural gas (speciﬁc gravity = 0.6) at a ﬂow rate of 160 MMSCFD at an inlet temperature of 60°F. Assuming isothermal ﬂow, calculate the velocity of gas at the inlet and outlet of the pipe if the inlet pressure is 1200 psig and the outlet pressure is 700 psig. The base pressure and base temperature are 14.7 psia and 60°F, respectively. Assume the compressibility factor Z = 0.95. What is the pipe length for these pressures, if elevations are neglected? 2. A natural gas pipeline, DN 400 with 10 mm wall thickness, transports 3.2 Mm3/ day. The speciﬁc gravity of gas is 0.6 and viscosity is 0.00012 Poise. Calculate the value of the Reynolds number. Assume the base temperature and base pressure are 15 C and 101 kPa, respectively. 3. A natural gas pipeline, NPS 20 with 0.500 in. wall thickness, 50 miles long, transports 220 MMSCFD. The speciﬁc gravity of gas is 0.6 and viscosity is 0.000008 lb/ft-s. Calculate the friction factor using the Colebrook equation. Assume absolute pipe roughness = 750 µin. The base temperature and base pressure are 60°F and 14.7 psia, respectively. What is the upstream pressure for an outlet pressure of 800 psig? 4. For a gas pipeline ﬂowing 3.5 Mm3/day gas of speciﬁc gravity 0.6 and viscosity of 0.000119 Poise, calculate the friction factor and transmission factor, assuming a DN 400 pipeline, 10 mm wall thickness, and internal roughness of 0.015 mm. The base temperature and base pressure are 15°C and 101 kPa, respectively. If the ﬂow rate is increased by 50%, what is the impact on the friction factor and transmission factor? If the pipe length is 48 km, what is the outlet pressure for an inlet pressure of 9000 kPa? 5. A gas pipeline ﬂows 110 MMSCFD gas of speciﬁc gravity 0.65 and viscosity of 0.000008 lb/ft-s. Calculate, using the modiﬁed Colebrook-White equation, the friction factor and transmission factor, assuming an NPS 20 pipeline, 0.375 in. wall thickness, and internal roughness of 700 µin. The base temperature and base pressure are 60°F and 14.7 psia, respectively. 6. Using the AGA method, calculate the transmission factor and friction factor for gas ﬂow in an NPS 20 pipeline with 0.375 in. wall thickness. The ﬂow rate is 250 MMSCFD, gas gravity = 0.6, and viscosity = 0.000008 lb/ft-sec. The absolute pipe roughness is 600 µin. Assume a bend index of 60°, base pressure = 14.73 psia, and base temperature = 60°F. If the ﬂow rate is doubled, what pipe size is needed to keep both inlet and outlet pressures the same as that at the original ﬂow rate? 7. A natural gas transmission line transports 4 million m3/day of gas from a processing plant to a compressor station site 100 km away. The pipeline can be assumed to be along a ﬂat terrain. Calculate the minimum pipe diameter required such that the maximum pipe operating pressure is limited to 8500 kPa. The delivery pressure desired at the end of the pipeline is a minimum of 5500 kPa. Assume a pipeline efﬁciency of 0.92. The gas gravity is 0.60, and the gas temperature is 18°C. Use the Weymouth equation, considering a base temperature = 15°C and base pressure = 101 kPa. The gas compressibility factor Z = 0.90.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 83 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

83

8. Using the Panhandle B equation, calculate the outlet pressure in a natural gas pipeline, NPS 16 with 0.250 in. wall thickness, 25 miles long. The gas ﬂow rate is 120 MMSCFD at 1200 psia inlet pressure. The gas gravity = 0.6 and viscosity = 0.000008 lb/ft-sec. The average gas temperature is 80°F. Assume the base pressure = 14.73 psia and base temperature = 60°F. The compressibility factor Z = 0.90 and pipeline efﬁciency is 0.95.

REFERENCES 1. Nayyar, M.L., Piping Handbook, 7th ed., McGraw-Hill, New York, 2000. 2. Baumeister, T., Ed., Standard Handbook for Mechanical Engineers, 7th ed., McGrawHill, New York, 1967. 3. Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003. 4. Westaway, C.R. and Loomis, A.W., Cameron Hydraulic Data, 16th ed., IngersollRand, Montvale, NJ, 1981. 5. Flow of Fluids through Valves, Fittings and Pipe, Crane Company, New York, 1976. 6. McCain, W.D. Jr., The Properties of Petroleum Fluids, Petroleum Publishing Company, Tulsa, OK, 1973. 7. Holman, J.P., Thermodynamics, 2nd ed., McGraw-Hill, New York, 1974. 8. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 9. Engineering Data Book, 10th ed., Gas Processors Suppliers Association, Tulsa, OK, 1994. 10. Pipeline Design for Hydrocarbon Gases and Liquids, American Society of Civil Engineers, New York, 1975. 11. Steady Flow in Gas Pipelines, Contract Report No. 10, July 1965, Pipeline Research Council International, Houston, TX, 1965. 12. Katz, D.L. et al., Handbook of Natural Gas Engineering, McGraw-Hill, New York, 1959.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 84 Friday, April 1, 2005 3:23 PM

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 85 Friday, April 1, 2005 3:09 PM

CHAPTER

3

Pressure Required to Transport In this chapter we will extend the use of the concepts of pressure drop calculations developed in Chapter 2 to determine the total pressure required for transporting gas in a pipeline under various conﬁgurations, such as series and parallel pipelines. We will identify the various components that make up this total pressure and analyze their impact on gas pipeline pressures. The effect of intermediate delivery volumes and injection rates along a gas pipeline, the impact of contract delivery pressures, and the necessity of regulating pressures using a control valve or pressure regulators will also be analyzed. Thermal effects due to heat transfer between the gas and the surrounding soil in a buried pipeline, soil temperatures and thermal conductivities, and the JouleThompson effect will be introduced with reference to commercial hydraulic simulation models. Equivalent lengths in series piping and equivalent diameters in parallel piping will be explained. We will compare different pipe looping scenarios to improve pipeline throughput and review the concept of the hydraulic pressure gradient. Calculation methodology for line pack in a gas pipeline will also be discussed.

3.1 TOTAL PRESSURE DROP REQUIRED In the ﬂow of incompressible ﬂuids such as water, the pressure required to transport a speciﬁed volume of ﬂuid from point A to point B will consist of the following components: 1. Frictional component 2. Elevation component 3. Pipe delivery pressure

In addition, in some cases where the pipeline elevation differences are drastic, we must also take into account the minimum pressure in a pipeline such that vaporization of liquid does not occur. The latter results in two-phase ﬂow in the pipeline, which causes higher pressure drop and, therefore, more pumping power requirement in addition to possible damage to pumping equipment. Thus, single-phase incompressible 85

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 86 Friday, April 1, 2005 3:09 PM

86

GAS PIPELINE HYDRAULICS

ﬂuids must be pumped such that the pressure at any point in the pipeline does not drop below the vapor pressure of the liquid. When pumping gases, which are compressible ﬂuids, the three components listed in the preceding section also contribute to the total pressure required. Even though the relationship between the total pressure required and the pipeline elevation is not straightforward (as in liquid ﬂow), the dependency still exists and will be demonstrated using an example problem. Going back to the case of a liquid pipeline, suppose the total pressure required to pump a given volume is 1000 psig and it is composed of the following components: 1. Frictional component = 600 psig 2. Elevation component = 300 psig 3. Delivery pressure = 100 psig

We will now discuss each of these components that make up the total pressure required by comparing the situation between a liquid pipeline and a gas pipeline.

3.2 FRICTIONAL EFFECT The frictional effect results from the ﬂuid viscosity and pipe roughness. It is similar in liquid and gas ﬂow. The effect of friction was discussed in Chapter 2, where we introduced the internal roughness of pipe and how the friction factor and transmission factor were calculated using the Colebrook-White and AGA equations. We also discussed how the Weymouth and Panhandle equations took into account the internal conditions and age of the pipe by utilizing a pipeline efﬁciency factor rather than a friction factor. The magnitude of the pressure drop due to friction in a gas pipeline is generally held to smaller values in comparison with liquid pipelines. This is because efﬁcient gas pipeline transportation requires keeping the average gas pressure as high as possible. As pressure drops due to expansion of gas, there is loss in efﬁciency. The lower the pressure at the downstream end, the higher will be the compression ratio required (hence, the higher the HP) to boost the pressure for shipment downstream to the next compressor station in a long-distance gas pipeline. In this chapter we will continue to calculate the pressure drop due to friction in various pipe conﬁgurations that include ﬂow injection, deliveries, and series and parallel piping.

3.3 EFFECT OF PIPELINE ELEVATION The elevation component referred to in Section 3.1 is due to the difference in elevation along the pipeline that necessitates additional pressure for raising the ﬂuid in the pipeline from one point to another. Of course, a drop in elevation will have the opposite effect of a rise in elevation. The elevation component of 300 psig in the preceding example depends upon the static elevation difference between the beginning of the pipeline, A, and the delivery point, B, and the liquid speciﬁc gravity. In the case of a gas pipeline, the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 87 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

87

elevation component will depend upon the static elevation differences between A and B, as well as the gas gravity. However, the relationship between these parameters is more complex in a gas pipeline compared to a liquid pipeline. The rise and fall in elevations between the origin A and the terminus B have to be accounted for separately and summed up according to Section 2.4 in Chapter 2. Further, compared to a liquid, the gas gravity is several orders of magnitude lower and, hence, the inﬂuence of elevation is smaller in a pipeline that transports gas. Generally, if we were to break down the total pressure required in a gas pipeline into the three components discussed earlier, we would ﬁnd that the elevation component is very small. Let us illustrate this using an example. Example 1 A gas pipeline, NPS 16 with 0.250 in. wall thickness, 50 mi long, transports natural gas (speciﬁc gravity = 0.6 and viscosity = 0.000008 lb/ft-s) at a ﬂow rate of 100 MMSCFD at an inlet temperature of 60°F. Assuming isothermal ﬂow, calculate the inlet pressure required if the required delivery pressure at the pipeline terminus is 870 psig. The base pressure and base temperature are 14.7 psig and 60°F, respectively. Use the Colebrook equation with pipe roughness of 0.0007 in. Case A—Consider no elevation changes along the pipeline length. Case B—Consider elevation changes as follows: inlet elevation of 100 ft and elevation at delivery point of 450 ft, with elevation at the midpoint of 250 ft. Solution Inside diameter of pipe D = 16 – 2 × 0.250 = 15.5 in. First, we calculate the Reynolds number from Equation 2.34: 14.7 0.6 × 100 × 10 6 R = 0.0004778 = 6,535,664 60 + 460 0.000008 × 15.5 Next, using Colebrook Equation 2.39, we calculate the friction factor as 0.0007 2.51 = −2Log10 + f 3.7 × 15.5 6535664 f

1

Solving by trial and error, we get f = 0.0109 Therefore, the transmission factor is, using Equation 2.42, F=

Copyright 2005 by Taylor & Francis Group, LLC

2 0.0109

= 19.1954

2785_C003.fm Page 88 Friday, April 1, 2005 3:09 PM

88

GAS PIPELINE HYDRAULICS

To calculate the compressibility factor Z , the average pressure is required. Since the inlet pressure is unknown, we will calculate an approximate value of Z using a value of 110% of the delivery pressure for the average pressure. The average pressure is Pavg = 1.1 × (870 + 14.7) = 973.17 psia Using CNGA Equation 1.34, we calculate the value of the compressibility factor as Z=

1 +

(

1 (973.17−14.7)×344400 (10 )1.785× 0.6 5203.8825

)

= 0.8629

Case A Since there is no elevation difference between the beginning of the pipeline and the end of the pipeline, the elevation component in Equation 2.7 can be neglected, and es = 1. The outlet pressure is P2 = 870 + 14.7 = 884.7 psia. From General Flow Equation 2.4, substituting the given values, we get P12 − 884.72 520 100 × 10 = 38.77 × 19.1954 14.7 0.6 × 520 × 50 × 0.8629

0.5

6

(15.5)2.5

Therefore, the upstream pressure is P1 = 999.90 psia = 985.20 psig Using this value of P1, we calculate the new average pressure using Equation 2.14: Pavg =

999.9 × 884.7 2 999.9 + 884.7 − = 943.47 psia 999.9 + 884.7 3

compared to 973.17 we used for calculating Z. Recalculating Z using the new value of Pavg, we get Z=

Copyright 2005 by Taylor & Francis Group, LLC

1 +

(

1 (943.47−14.7)×344400 (10 )1.785× 0.6 5203.8825

)

= 0.8666

2785_C003.fm Page 89 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

89

This compares with 0.8629 we calculated earlier for Z . We will now recalculate the inlet pressure using this value of Z. From General Flow Equation 2.4, we get P12 − 884.72 520 100 × 10 = 38.77 × 19.1954 14.7 0.6 × 520 × 50 × 0.8666 6

0.5

(15.5)2.5

Solving for the upstream pressure, we get P1 = 1000.36 psia = 985.66 psig This is close enough to the previously calculated value 985.20 psig, and no further iteration is needed. Therefore, the pressure required at the beginning of the pipeline in case A is 985.66 psig when the elevation difference is zero. We will now calculate the pressure required, taking into account the given elevations at the beginning, midpoint, and end of the pipeline. Case B We will use Z = 0.8666 throughout, as in case A. Using Equation 2.10, the elevation adjustment factor is ﬁrst calculated for each of the two segments. For the ﬁrst segment, from milepost 0.0 to milepost 25.0, we get 250 − 100 s1 = 0.0375 × 0.6 = 0.0075 520 × 0.8666 Similarly, for the second segment, from milepost 25.0 to milepost 50.0, we get 450 − 100 s2 = 0.0375 × 0.6 = 0.0175 520 × 0.8666 Therefore, the adjustment for elevation is, using Equation 2.12,

j=

e 0.0075 − 1 = 1.0038 for the ﬁrst segment 0.0075

and

j=

Copyright 2005 by Taylor & Francis Group, LLC

e 0.0175 − 1 = 1.0088 for the second segment 0.0175

2785_C003.fm Page 90 Friday, April 1, 2005 3:09 PM

90

GAS PIPELINE HYDRAULICS

For the entire length, 450 − 100 s2 = 0.0375 × 0.6 = 0.0175 520 × 0.8666 The equivalent length from Equation 2.13 is then Le = 1.0038 × 25 + 1.0088 × 25 × e0.0075 = 50.5049 mi. Therefore, we see that the effect of the elevation is taken into account partly by increasing the pipe length from 50 mi to 50.50 mi, approximately. Substituting in Equation 2.7, we get P12 − e 0.0175 884.72 520 100 × 10 = 38.77 × 19.1954 . . . 14 7 0 6 520 50 50 0 8666 × × × . 6

0.5

15.52.5

Solving for the inlet pressure P1, P1 = 1008.34 psia = 993.64 psig Thus, the pressure required at the beginning of the pipeline in case B is 993.64 psig, taking into account elevation difference along the pipeline. Compare this with 985.66 calculated ignoring the elevation differences. For simplicity, we assume the same value of Z in the preceding calculations as in the previous case. To be correct, we should recalculate Z based on the average pressure and repeat calculations until the results are within 0.1 psi. This is left as an exercise for the reader. It can be seen from the preceding calculations that, due to an elevation difference of 350 ft (450 ft − 100 ft) between the delivery point and the beginning of the pipeline, the required pressure is approximately 8 psig (993.64 psig − 985.66 psig) more. In a liquid line, the effect of elevation would have been more. The elevation difference of 350 ft in a water line would result in an increased pressure of 350 × 0.433 = 152 psi, approximately, at the upstream end.

3.4 EFFECT OF CHANGING PIPE DELIVERY PRESSURE The delivery pressure component discussed in Section 3.1 is also similar to that between liquid and gas pipelines. The higher the pressure desired at the delivery end or terminus of the pipeline, the higher will be the total pressure required at the upstream end of the pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 91 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

91

The impact of changing the delivery pressure is not linear in the case of a compressible ﬂuid such as natural gas. For example, in a liquid pipeline, changing the delivery pressure from 100 to 200 psig will simply increase the required pressure at the pipe inlet by the same amount. Thus, suppose 1000 psig was the required inlet pressure in a liquid pipeline, at a certain ﬂow rate and at a delivery pressure of 100 psig. When the delivery pressure required is increased to 200 psig, the inlet pressure will increase to exactly 1100 psig. We will now explore the effect of changing the contract delivery pressure at the end of a gas pipeline. In liquid pipelines, an increase or decrease in the delivery pressure will proportionately increase or decrease the upstream pressure. In a gas pipeline, the increase and decrease in the upstream pressure will not be proportionate due to the nonlinear nature of the gas pressure drop. This will be explained in more detail in Section 3.9 in this chapter. Suppose that in the preceding Example 1 (neglecting elevation change), the delivery pressure required increases from 870 to 950 psig. If pressure variations were linear, as in a liquid pipeline, we would expect the required inlet pressure to increase from 985.66 to 985.66 + (950 − 870) = 1066 psig, approximately. However, this is incorrect because the pressure variation is not linear in gas pipelines. We will now calculate the required inlet pressure when the delivery pressure is increased from 870 to 950 psig. All parameters in case A are the same except for the delivery pressure. The increased delivery pressure will cause the compressibility factor to change slightly due to the change in average pressure. However, for simplicity, we will assume Z = 0.8666, as before. The new delivery pressure is P2 = 950 + 14.7 = 964.7 psia Substituting in General Flow Equation 2.4, we get 520 P12 − 964.72 100 × 10 6 = 38.77 × 19.1954 14.7 0.6 × 520 × 50 × 0.8666

0.5

15.52.5

Therefore, P1 = 1071.77 psia = 1057.07 psig Thus, the pressure required at the beginning of the pipeline is approximately 1057 psig. This compares with a value of 1066 psig we calculated if the pressure variation were linear. In general, for a gas pipeline, if the delivery pressure is increased by ∆P, the inlet pressure will increase by less than ∆P. Similarly, if the delivery pressure is decreased by ∆P, the inlet pressure will decrease by less than ∆P. We will illustrate this using the preceding example. Suppose in case A, the delivery pressure was decreased from 870 to 800 psig. If pressure variation were linear, we would expect the pipe inlet pressure to decrease

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 92 Friday, April 1, 2005 3:09 PM

92

GAS PIPELINE HYDRAULICS

by 70 psig to (985.66 – 70) = 916 psig, approximately. However, as indicated earlier, this is incorrect. We will now calculate the actual inlet pressure using the General Flow equation considering the reduced outlet pressure of 800 psig. All parameters in case A are the same except for the delivery pressure. The decreased delivery pressure will cause the compressibility factor to change slightly due to the change in average pressure. However, for simplicity, we will assume Z = 0.8666, as before. New delivery pressure P2 = 800 + 14.7 = 814.7 psia Substituting in General Flow Equation 2.4, we get 520 P12 − 814.72 100 × 10 6 = 38.77 × 19.1954 14.7 0.6 × 520 × 50 × 0.8666

0.5

15.52.5

Therefore, P1 = 939.03 psia = 924.33 psig Thus, the pressure required at the beginning of the pipeline is 924.33 psig. This compares with a value of 916 psig if the pressure variation were linear. Therefore, by decreasing the delivery pressure by 70 psig, the inlet pressure decreases by less than 70 psig. We can perform a similar analysis by changing the inlet pressure by a ﬁxed amount and calculating the effect on the pipe delivery pressure. As before, considering no elevation changes, an inlet pressure of 985.66 psig results in a delivery pressure of 870 psig. Suppose we decrease the inlet pressure to 900 psig (a reduction of 85.66 psig); if pressures were linear, we would expect the delivery pressure to drop to (870 – 85.66) = 784.34 psig. Actually, we will see that the delivery pressure would drop to a number lower than this. In other words, decreasing the inlet pressure by ∆P reduces the outlet pressure by more than ∆P. Following the previous methodology, we calculate the revised delivery pressure by assuming the same Z = 0.8666 for simplicity. New inlet pressure = 900 + 14.7 = 914.7 psia Substituting in General Flow Equation 2.4, we calculate the outlet pressure as 520 914.72 − P2 2 100 × 10 6 = 38.77 × 19.1954 14.7 0.6 × 520 × 50 × 0.8666 Therefore, P2 = 786.54 psia = 771.84 psig

Copyright 2005 by Taylor & Francis Group, LLC

0.5

15.52.5

2785_C003.fm Page 93 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

93

Thus, the delivery pressure is reduced by (870 – 771.84) = 98.16 psig, whereas the inlet pressure was reduced by only 85.66 psig. In general, if the inlet pressure is decreased by ∆P, the delivery pressure will decrease by more than ∆P. On the other hand, if the inlet pressure is increased by ∆P, the delivery pressure will increase by more than ∆P. Therefore, if the inlet pressure is increased from 985.66 to 1085.66 psig (an increase of 100 psig), the delivery pressure will increase from 870 psig to a number larger than 970 psig.

3.5 PIPELINE WITH INTERMEDIATE INJECTIONS AND DELIVERIES A pipeline in which gas enters at the beginning of the pipeline and the same volume exits at the end of the pipeline is a pipeline with no intermediate injection or deliveries. When portions of the inlet volume are delivered at various points along the pipeline and the remaining volume is delivered at the end of the pipeline, we call this system a pipeline with intermediate delivery points. A more complex case with gas ﬂow into the pipeline (injection) at various points along its length combined with deliveries at other points is shown in Figure 3.1. In such a pipeline system, the pressure required at the beginning point A will be calculated by considering the pipeline broken into segments AB, BC, etc. Another piping system can consist of gas ﬂow at the inlet of the pipeline along with multiple pipe branches making deliveries of gas, as shown in Figure 3.2. In this system, pipe AB has a certain volume, Q1, ﬂowing through it. At point B, another pipeline, CB, brings in additional volumes resulting in a volume of (Q1 + Q2) ﬂowing through section BD. At D, a branch pipe, DE, delivers a volume of Q3 to a customer location, E. The remaining volume (Q1 + Q2 – Q3) ﬂows from D to F through pipe segment DF to a customer location at F. In the subsequent sections, we will analyze pipelines with intermediate ﬂow deliveries, injections, as well as branch pipes, as shown in Figure 3.1 and Figure 3.2. The objectives in all cases will be to calculate the pressures and ﬂow rates through the various pipe sections and to determine pipe sizes required to limit pressure drop in certain pipe segments.

100 MMSCFD A

80 MMSCFD

20 MMSCFD Figure 3.1

C

B

NPS 16

30 MMSCFD

Pipeline with injection and deliveries.

Copyright 2005 by Taylor & Francis Group, LLC

90 MMSCFD

50 MMSCFD E

40 MMSCFD

D

2785_C003.fm Page 94 Friday, April 1, 2005 3:09 PM

94

GAS PIPELINE HYDRAULICS

C E

Q2

Q3

Q1

A

Q1 + Q2

10 in. – 10 mi

B

F

D Q1 + Q2 – Q3

Figure 3.2

Pipeline with branches.

Example 2 A 150 mi long natural gas pipeline consists of several injections and deliveries as shown in Figure 3.3. The pipeline is NPS 20, has 0.500 in. wall thickness, and has an inlet volume of 250 MMSCFD. At points B (milepost 20) and C (milepost 80), 50 MMSCFD and 70 MMSCFD, respectively, are delivered. At D (milepost 100), gas enters the pipeline at 60 MMSCFD. All streams of gas may be assumed to have a speciﬁc gravity of 0.65 and a viscosity of 8.0 × 10−6 lb/ft-s. The pipe is internally coated (to reduce friction), resulting in an absolute roughness of 150 µ in. Assume a constant gas ﬂow temperature of 60°F and base pressure and base temperature of 14.7 psia and 60°F, respectively. Use a constant compressibility factor of 0.85 throughout. Neglect elevation differences along the pipeline. a) Using the AGA equation, calculate the pressures along the pipeline at points A, B, C, and D for a minimum delivery pressure of 300 psig at the terminus E. Assume a drag factor = 0.96. b) What diameter pipe will be required for section DE if the required delivery pressure at E is increased to 500 psig? The inlet pressure at A remains the same as calculated above.

250 MMSCFD A

200 MMSCFD C

B

190 MMSCFD

130 MMSCFD D

NPS 20 0.500 in. wall

50 MMSCFD Figure 3.3

70 MMSCFD

Example pipeline with injection and deliveries.

Copyright 2005 by Taylor & Francis Group, LLC

E 300 psig

60 MMSCFD

2785_C003.fm Page 95 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

95

Solution We will start calculations beginning with the last segment DE. Pipe inside diameter D = 20 – 2 × 0.500 = 19.00 in. The ﬂow rate in pipe DE is 190 MMSCFD. Using Equation 2.34, the Reynolds number is 14.7 0.65 × 190 × 10 6 R = 0.0004778 = 10,974,469 520 8 × 10 −6 × 19 Next, calculate the two transmission factors required per AGA. 1) The fully turbulent transmission factor, using Equation 2.48, is 3.7 × 19 F = 4 Log10 = 22.68 150 × 10 −6 2) The smooth pipe zone Von Karman transmission factor, using Equation 2.50, is 10, 974, 469 Ft = 4 Log10 − 0.6 Ft Solving for Ft by iteration, we get Ft = 22.18 Therefore, for a partly turbulent ﬂow zone, the transmission factor, using Equation 2.49, is 10, 974, 469 F = 4 × 0.96Log10 = 21.29 1.4125 × 22.18 Using the smaller of the two values, the AGA transmission factor is F = 21.29 Next, we use General Flow Equation 2.4 to calculate the upstream pressure P1 at D, based on a given downstream pressure of 300 psig at E. P12 − 314.72 520 190 × 10 = 38.77 × 21.29 14.7 0.65 × 520 × 50 × 0.85 6

Copyright 2005 by Taylor & Francis Group, LLC

0.5

192.5

2785_C003.fm Page 96 Friday, April 1, 2005 3:09 PM

96

GAS PIPELINE HYDRAULICS

Solving for P1, we get the pressure at D as P1 = 587.11 psia = 572.41 psig Next, we consider the pipe segment CD, which has a ﬂow rate of 130 MMSCFD. We calculate the pressure at C using the downstream pressure at D calculated above. To simplify calculation, we will use the same AGA transmission factor we calculated for segment DE. A more nearly correct solution will be to calculate the Reynolds number and the two transmission factors as we did for the segment DE. However, for simplicity, we will use F = 21.29 for all pipe segments. Applying General Flow Equation 2.4, we calculate the pressure P1 at C as follows: P12 − 587.112 520 130 × 10 = 38.77 × 21.29 14.7 0.65 × 520 × 20 × 0.85

0.5

(19.0)2.5

6

Solving for P1, we get the pressure at C as P1 = 625.06 psia = 610.36 psig Similarly, we calculate the pressure at B, considering the pipe segment BC that ﬂows 200 MMSCFD. P12 − 625.062 520 200 × 10 = 38.77 × 21.29 14.7 0.65 × 520 × 60 × 0.85

0.5

(19.0)2.5

6

Solving for P1, we get the pressure at B as P1 = 846.95 psia = 832.25 psig Finally, for pipe segment AB that ﬂows 250 MMSCFD, we calculate the pressure P1 at A as follows: P12 − 846.952 520 250 × 10 = 38.77 × 21.29 14.7 0.65 × 520 × 20 × 0.85 6

0.5

(19.0)2.5

Solving for P1, we get the pressure at A as P1 = 942.04 psia = 927.34 psig If we maintain the same inlet pressure, 927.34 psig, at A and increase the delivery pressure at E to 500 psig, we can determine the pipe diameter required for section DE by considering the same upstream pressure of 572.41 psig at D, as we calculated before.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 97 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

97

Therefore, for segment DE, Upstream pressure P1 = 572.41 + 14.7 = 587.11 psia Downstream pressure P2 = 500 + 14.7 = 514.7 psia Using General Flow Equation 2.4, with the same AGA transmission factor as before, we get 520 587.112 − 514.72 190 × 10 = 38.77 × 21.29 14.7 0.65 × 520 × 50 × 0.85

0.5

6

( D)2.5

Solving for the inside diameter D of pipe DE, we get D = 23.79 in. The nearest standard pipe size is NPS 26 with 0.500 in. wall thickness. This will give an inside diameter of 25 in., which is slightly more than the required minimum of 23.79 in. calculated above. The wall thickness required for this pipe diameter and pressure will be dictated by the pipe material and is the subject of Chapter 6. Example 3 A pipeline 100 mi long transports natural gas from Corona to Beaumont. The gas has a speciﬁc gravity of 0.60 and a viscosity of 8 × 10−6 lb/ft-s. What is the minimum pipe diameter required to ﬂow 100 MMSCFD from Corona to Beaumont for a delivery pressure of 800 psig at Beaumont and inlet pressure of 1400 psig at Corona? The gas can be assumed constant at 60°F, and the base pressure and base temperature are 14.7 psia and 60°F, respectively. Use a constant value of 0.90 for the compressibility factor and a pipe roughness of 700 µ in. Compare results using the AGA, ColebrookWhite, Panhandle B, and Weymouth equations. Use 95% pipeline efﬁciency. Neglect elevation differences along the pipeline. How will the result change if the elevation at Corona is 100 ft and at Beaumont is 500 ft? Solution We will ﬁrst use the AGA equation to determine the pipe diameter. Since the transmission factor F depends on the Reynolds number, which depends on the unknown pipe diameter, we will ﬁrst assume a value of F = 20. From General Flow Equation 2.4, we get 520 1414.72 − 814.772 100 × 10 6 = 38.77 × 20.0 14.7 0.6 × 520 × 100 × 0.9

Copyright 2005 by Taylor & Francis Group, LLC

0.5

× ( D)2.5

2785_C003.fm Page 98 Friday, April 1, 2005 3:09 PM

98

GAS PIPELINE HYDRAULICS

Solving for diameter D, D = 12.28 in. or NPS 12 with a 0.250 in. wall thickness, approximately. Next, we will recalculate the transmission factor using this pipe size. Using NPS 12 with a 0.250 in. wall thickness, Inside pipe diameter D = 12.75 – 2 × 0.250 = 12.25 in. Calculating the Reynolds number from Equation 2.34, we get 14.7 0.6 × 100 × 10 6 = 8,269,615 R = 0.0004778 520 8 × 10 −6 × 12.25 The fully turbulent transmission factor, using Equation 2.48, is 3.7 × 12.25 = 19.25 F = 4 Log10 0.0007 For the smooth pipe zone, using Equation 2.50, the Von Karman transmission factor is 8, 269, 615 Ft = 4 Log10 − 0.6 Ft Solving for Ft by iteration, we get Ft = 21.72 Using a drag factor of 0.96, for partly turbulent ﬂow, the transmission factor is, from Equation 2.49, 8, 269, 615 F = 4 × 0.96 × Log10 = 20.85 1.4125 × 21.72 Using the lower of the two values, the AGA transmission factor is F = 19.25 Using this value of F, we recalculate the minimum pipe diameter from General Flow Equation 2.4 as follows: 520 1414.72 − 814.72 100 × 10 6 = 38.77 × 19.25 14.7 0.6 × 520 × 100 × 0.9

Copyright 2005 by Taylor & Francis Group, LLC

0.5

× ( D)2.5

2785_C003.fm Page 99 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

99

Solving for diameter D, D = 12.47 in. We will not continue iteration any further, since the new diameter will not change the value of F appreciably. Therefore, based on the AGA equation, the pipe inside diameter required is 12.47 in. Next, we calculate the transmission factor based on the Colebrook-White equation, assuming an inside diameter of 12.25 in. and the Reynolds number = 8,269,615, calculated earlier. Using Colebrook-White Equation 2.45, we get 0.0007 1.255F F = −4 Log10 + 3.7 × 12.25 8, 269, 615 Solving for F by successive iteration, we get the Colebrook-White transmission factor as F = 18.95 Using the General Flow equation with this Colebrook-White transmission factor, we calculate the diameter as follows: 520 1414.72 − 814.72 100 × 10 = 38.77 × 18.95 14.7 0.6 × 520 × 100 × 0.9 6

0.5

× ( D)2.5

Solving for diameter D, D = 12.55 in. Recalculating the Reynolds number and transmission factor using the pipe inside diameter of 12.55 in., we get R = 8,071,935 and F = 18.94 Therefore, the new diameter required is by proportions, using the General Flow equation, D 12.55

2.5

18.95 = 18.94

or D = 12.55, approximately. There is no appreciable change in the diameter required. Therefore, based on the Colebrook-White equation, the pipe inside diameter required is D = 12.55 in.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 100 Friday, April 1, 2005 3:09 PM

100

GAS PIPELINE HYDRAULICS

Next, we determine the diameter required using Panhandle B Equation 2.59 and a pipeline efﬁciency of 0.95: 520 100 × 10 6 = 737 × 0.95 14.7

1.02

1414.72 − 814.72 0.961 × 520 × 100 × 0.9 0.6

0.51

D 2.53

Solving for diameter D, we get D = 11.93 in. Therefore, based on the Panhandle B equation, the pipe inside diameter required is 11.93 in. Next, we calculate the diameter required, using Weymouth Equation 2.52 and a pipeline efﬁciency of 0.95: 520 1414.72 − 814.772 100 × 10 6 = 433.5 × 0.95 14.7 0.6 × 520 × 100 × 0.9

0.5

D 2.667

Solving for diameter D, we get D = 13.30 in. Therefore, based on the Weymouth equation, the pipe inside diameter required is 13.30 in. In summary, the minimum pipe inside diameter required based on the various ﬂow equations is as follows: AGA — D = 12.47 in. Colebrook-White — D = 12.55 in. Panhandle B — D = 11.93 in. Weymouth equation — D = 13.30 in. It can be seen that the Weymouth equation is the most conservative equation. The AGA and Colebrook-White equations predict almost the same pipe size, while Panhandle B predicts the smallest pipe size. To further illustrate the comparison of various pressure drop equations, refer to the discussion in Chapter 2 and Figure 2.5, which shows how the delivery pressure varies for a ﬁxed ﬂow rate and inlet pressure. Table 3.1 also summarizes the various pressure drop equations used in the gas pipeline industry. Considering elevation effects, with a single slope from Corona (100 ft) to Beaumont (500 ft), the elevation adjustment parameter is, from Equation 2.10, 500 − 100 s = 0.0375 × 0.6 = 0.0192 520 × 0.9

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 101 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

101

Table 3.1 Summary of Pressure Drop Equations Equation

Application

General Flow

Fundamental flow equation using friction or transmission factor; used with Colebrook-White friction factor or AGA transmission factor

Colebrook-White

Friction factor calculated for pipe roughness and Reynolds number; most popular equation for general gas transmission pipelines

Modified Colebrook-White

Modified equation based on U.S. Bureau of Mines experiments; gives higher pressure drop compared to original Colebrook equation

AGA

Transmission factor calculated for partially turbulent and fully turbulent flow considering roughness, bend index, and Reynolds number

Panhandle A Panhandle B

Panhandle equations do not consider pipe roughness; instead, an efficiency factor is used; less conservative than Colebrook or AGA

Weymouth

Does not consider pipe roughness; uses an efficiency factor used for high-pressure gas gathering systems; most conservative equation that gives highest pressure drop for given flow rate

IGT

Does not consider pipe roughness; uses an efficiency factor used on gas distribution piping

Therefore, the equivalent length from Equation 2.9 is Le = 100 ×

e 0.0192 − 1 = 100.97 mi 0.0192

We will apply the elevation correction factor for the extreme cases (Weymouth and Panhandle B equations) that produce the largest and the smallest diameter, respectively. From Weymouth Equation 2.52, we see that, keeping all other items the same, the diameter and pipe length are related by the following equation: D 2.667 L D 13.3

2.667

= Constant 100.97 = 100

0.5

Solving for the pipe inside diameter D, we get D = 13.32 in. This is not an appreciable change from the previous value of 13.30 in. Similarly, from Panhandle B Equation 2.59, we see that the pipe diameter and length are related by D 2.53 = Constant L0.51 D 11.93

Copyright 2005 by Taylor & Francis Group, LLC

2.53

100.97 = 100

0.51

2785_C003.fm Page 102 Friday, April 1, 2005 3:09 PM

102

GAS PIPELINE HYDRAULICS

Solving for pipe inside diameter D, we get D = 11.95 This is not an appreciable change from the previous value of 11.93 in. Therefore, considering elevation difference between Corona and Beaumont, the minimum pipe sizes required are as follows: Panhandle B — D = 11.95 in. Weymouth equation — D = 13.32 in. We thus see that even with a 400 ft elevation difference, the pipe diameter does not change appreciably. Example 4 A natural gas distribution piping system consists of NPS 12 with 0.250 in. wall thickness, 24 mi long, as shown in Figure 3.4. At Yale, an inlet ﬂow rate of 65 MMSCFD of natural gas enters the pipeline at 60°F. At the Compton terminus, gas must be supplied at a ﬂow rate of 30 MMSCFD at a minimum pressure of 600 psig. There are intermediate deliveries of 15 MMSCFD at milepost 10 and 20 MMSCFD at milepost 18. What is the required inlet pressure at Yale? Use a constant friction factor of 0.01 throughout. The compressibility factor can be assumed to be 0.94. The gas gravity and viscosity are 0.6 and 7 × 10 −6 lb/ft-s, respectively. Assume isothermal ﬂow at 60°F. The base temperature and base pressure are 60°F and 14.7 psia, respectively. If the delivery volume at B is increased to 30 MMSCFD and other deliveries remain the same, what increased pressure is required at Yale to maintain the same ﬂow rate and delivery pressure at Compton? Neglect elevation differences along the pipeline. Solution For each section of piping, such as AB, we must calculate the pressure drop due to friction at the appropriate ﬂow rate and then determine the total pressure drop for the entire pipeline. Inside diameter of pipe = 12.75 – 2 × 0.250 = 12.25 in. Friction factor f = 0.01

65 MMSCFD Yale

NPS 12 , 0.250 in. wall thickness B

m.p. 0.0

C m.p. 18.0

m.p. 10.0

A 20 MMSCFD

Yale to Compton gas distribution pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

600 psig

Compton m.p. 24.0 D

15 MMSCFD Figure 3.4

30 MMSCFD

2785_C003.fm Page 103 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

103

Therefore, the transmission factor, using Equation 2.42, is F=

2 0.01

= 20.00

Using General Flow Equation 2.7, for the last pipe segment from milepost 18 to milepost 24, we get PC2 − 614.72 520 30 × 10 = 38.77 × 20.0 14.7 0.6 × 520 × 6 × 0.94

0.5

6

× (12.25)2.5

Solving for the pressure at C, PC = 620.88 psia Next we will use this pressure PC to calculate the pressure PB for the 8 mi section of pipe segment BC ﬂowing 50 MMSCFD. Using General Flow Equation 2.7, PB2 − 620.882 520 50 × 10 = 38.77 × 20 14.7 0.6 × 520 × 8 × 0.94

0.5

× (12.25)2.5

6

Solving for PB, we get PB = 643.24 psia Finally, we calculate the pressure P1 at Yale by considering the 10 mi pipe segment from Yale to point B that ﬂows 65 MMSCFD. P12 − 643.24 2 520 65 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × 10 × 0.94

0.5

× (12.25)2.5

Solving for the pressure at Yale, we get P1 = 688.09 psia = 673.39 psig Therefore, the required inlet pressure at Yale is 673.39 psig. When the delivery volume at B is increased from 15 to 30 MMSCFD and all other delivery volumes remain the same, the inlet ﬂow rate at Yale will increase to 65 + 15 = 80 MMSCFD. If the delivery pressure at Compton is to remain the same as before, the pressures at B and C will also be the same as calculated before, since the ﬂow rate in BC and CD are the same as before. Therefore, we can recalculate the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 104 Friday, April 1, 2005 3:09 PM

104

GAS PIPELINE HYDRAULICS

inlet pressure for the pipe section from Yale to point B considering a ﬂow rate of 80 MMSCFD that causes a pressure of 643.24 psia at B. Using General Flow Equation 2.7, the pressure P1 at Yale is P12 − 643.24 2 520 80 × 10 6 = 38.77 × 20.0 14.7 0.6 × 520 × 10 × 0.94

0.5

× (12.25)2.5

Solving for the pressure at Yale, P1 = 710.07 psia = 695.37 psig Therefore, increasing the delivery volume at B by 15 MMSCFD causes the pressure at Yale to increase by approximately 22 psig.

3.6 SERIES PIPING In the preceding discussions we assumed the pipeline to have the same diameter throughout its length. There are situations where a gas pipeline can consist of different pipe diameters connected together in a series. This is especially true when the different pipe segments are required to transport different volumes of gas, as shown in Figure 3.5. In Figure 3.5, section AB with a diameter of 16 in. is used to transport a volume of 100 MMSCFD, and after making a delivery of 20 MMSCFD at B, the remainder of 80 MMSCFD ﬂows through the 14 in. diameter pipe BC. At C, a delivery of 30 MMSCFD is made, and the balance volume of 50 MMSCFD is delivered to the terminus D through a 12 in. pipeline CD. It is clear that the pipe section AB ﬂows the largest volume (100 MMSCFD), whereas the pipe segment CD transports the least volume (50 MMSCFD). Therefore, segments AB and CD, for reasons of economy, should be of different pipe diameters, as indicated in Figure 3.5. If we maintained the same pipe diameter of 16 in. from A to D, it would be a waste of pipe material and, therefore, cost. Constant diameter

100 MMSCFD

A

NPS 16

80 MMSCFD

B

20 MMSCFD Figure 3.5

Series piping.

Copyright 2005 by Taylor & Francis Group, LLC

NPS 14

50 MMSCFD

C

30 MMSCFD

NPS 12

D

2785_C003.fm Page 105 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

105

is used only when the same ﬂow that enters the pipeline is also delivered at the end of the pipeline, with no intermediate injections or deliveries. However, in reality, there is no way of determining ahead what the future delivery volumes will be along the pipeline. Hence, it is difﬁcult to determine initially the different pipe sizes for each segment. Therefore, in many cases you will ﬁnd that the same-diameter pipe is used throughout the entire length of the pipeline even though there are intermediate deliveries. Even with the same nominal pipe diameter, different pipe sections can have different wall thicknesses. Therefore, we have different pipe inside diameters for each pipe segment. Such wall thickness changes are made to compensate for varying pressures along the pipeline. The subject of pipe strength and its relation to pipe diameter and wall thickness is discussed in Chapter 6. The pressure required to transport gas in a series pipeline from point A to point D in Figure 3.5 is calculated by considering each pipe segment such as AB and BC and applying the appropriate ﬂow equation, such as the General Flow equation, for each segment, as illustrated in Example 5. Another approach to calculating the pressures in series piping systems is to use the equivalent length concept. This method can be applied when the same uniform ﬂow exists throughout the pipeline, with no intermediate deliveries or injections. We will explain this method of calculation for a series piping system with the same ﬂow rate Q through all pipe segments. Suppose the ﬁrst pipe segment has an inside diameter D1 and length L1, followed by the second segment of inside diameter D2 and length L2 and so on. We calculate the equivalent length of the second pipe segment based on the diameter D1 such that the pressure drop in the equivalent length matches that in the original pipe segment of diameter D2. The pressure drop in diameter D2 and length L2 equals the pressure drop in diameter D1 and equivalent length Le2. Thus, the second segment can be replaced with a piece of pipe of length Le2 and diameter D1. Similarly, the third pipe segment with diameter D3 and length L3 will be replaced with a piece of pipe of Le3 and diameter D1. Thus, we have converted the three segments of pipe in terms of diameter D1 as follows: Segment 1 — diameter D1 and length L1 Segment 2 — diameter D1 and length Le2 Segment 3 — diameter D1 and length Le3

For convenience, we picked the diameter D1 of segment 1 as the base diameter to use, to convert from the other pipe sizes. We now have the series piping system reduced to one constant-diameter (D1) pipe of total equivalent length given by Le = L1 + Le2 + Le3

(3.1)

The pressure required at the inlet of this series piping system can then be calculated based on diameter D1 and length Le. We will now explain how the equivalent length is calculated. Upon examining General Flow Equation 2.7, we see that for the same ﬂow rate and gas properties, neglecting elevation effects, the pressure difference (P12 – P22) is

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 106 Friday, April 1, 2005 3:09 PM

106

GAS PIPELINE HYDRAULICS

inversely proportional to the ﬁfth power of the pipe diameter and directly proportional to the pipe length. Therefore, we can state that, approximately, ∆Psq = where ∆Psq C L D

= = = =

CL D5

(3.2)

difference in the square of pressures (P12 – P22) for the pipe segment a constant pipe length pipe inside diameter

Actually, C depends on the ﬂow rate, gas properties, gas temperature, base pressure, and base temperature. Therefore, C will be the same for all pipe segments in a series pipeline with constant ﬂow rate. Hence, we regard C as a constant for all pipe segments. From Equation 3.2 we conclude that the equivalent length for the same pressure drop is proportional to the ﬁfth power of the diameter. Therefore, in the series piping discussed in the foregoing, the equivalent length of the second pipe segment of diameter D2 and length L2 is CL2 CLe2 = D25 D15

(3.3)

or D Le2 = L2 1 D2

5

(3.4)

Similarly, for the third pipe segment of diameter D3 and length L3 , the equivalent length is D Le3 = L3 1 D3

5

(3.5)

Therefore, the total equivalent length Le for all three pipe segments in terms of diameter D1 is 5

D D Le = L1 + L2 1 + L3 1 D2 D3

5

(3.6)

It can be seen from Equation 3.6 that if D1 = D2 = D3, the total equivalent length reduces to (L1 + L2 + L3), as expected. We can now calculate the pressure drop for the series piping system, considering a single pipe of length Le and uniform diameter D1 ﬂowing a constant volume Q. An example will illustrate the use of the equivalent length method.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 107 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

107

100 MMSCFD 100 MMSCFD

12 mi NPS 16

A

Figure 3.6

8 mi NPS 12

24 mi NPS 14

500 psig B

Example problem—series piping.

Example 5 A series piping system, shown in Figure 3.6, consists of 12 mi of NPS 16, 0.375 in. wall thickness connected to 24 mi of NPS 14, 0.250 in. wall thickness and 8 miles of NPS 12, 0.250 in. wall thickness pipes. Calculate the inlet pressure required at the origin A of this pipeline system for a gas ﬂow rate of 100 MMSCFD. Gas is delivered to the terminus B at a delivery pressure of 500 psig. The gas gravity and viscosity are 0.6 and 0.000008 lb/ft-s, respectively. The gas temperature is assumed constant at 60°F. Use a compressibility factor of 0.90 and the General Flow equation with Darcy friction factor = 0.02. The base temperature and base pressure are 60°F and 14.7 psia, respectively. Compare results using the equivalent length method and with the more detailed method of calculating pressure for each pipe segment separately. Solution Inside diameter of ﬁrst pipe segment = 16 – 2 × 0.375 = 15.25 in. Inside diameter of second pipe segment = 14 – 2 × 0.250 = 13.50 in. Inside diameter of third pipe segment = 12.75 – 2 × 0.250 = 12.25 in. Using Equation 3.6, we calculate the equivalent length of the pipeline, considering NPS 16 as the base diameter: 5

15.25 15.25 Le = 12 + 24 × +8× 13.5 12.25

5

or Le = 12 + 44.15 + 23.92 = 80.07 mi Therefore, we will calculate the inlet pressure P1 considering a single pipe from A to B having a length of 80.07 mi and inside diameter of 15.25 in. Outlet pressure = 500 + 14.7 = 514.7 psia

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 108 Friday, April 1, 2005 3:09 PM

108

GAS PIPELINE HYDRAULICS

Using General Flow Equation 2.2, neglecting elevation effects and substituting given values, we get

(

)

0.5

P12 − 514.72 1 520 2.5 100 × 10 = 77.54 14.7 0.6 × 520 × 80.07 × 0.9 15.25 0.02 6

or P12 – 514.72 = 724,642.99 Solving for the inlet pressure P1, we get P1 = 994.77 psia = 980.07 psig Next, we will compare the preceding result, using the equivalent length method, with the more detailed calculation of treating each pipe segment separately and adding the pressure drops. Consider the 8 mi pipe segment 3 ﬁrst, since we know the outlet pressure at B is 500 psig. Therefore, we can calculate the pressure at the beginning of segment 3 using General Flow Equation 2.2, as follows:

(

)

0.5

2 2 1 520 P1 − 514.7 12.252.5 100 × 10 = 77.54 0.02 14.7 0.6 × 520 × 8 × 0.9 6

Solving for the pressure P1, we get P1 = 693.83 psia = 679.13 psig This is the pressure at the beginning of the pipe segment 3, which is also the end of pipe segment 2. Next, consider pipe segment 2 (24 mi of NPS 14 pipe) and calculate the upstream pressure P1 required for a downstream pressure of 679.13 psig, calculated in the preceding section. Using General Flow Equation 2.2 for pipe segment 2, we get

(

)

0.5

2 2 1 520 P1 − 693.83 2.5 100 × 10 = 77.54 14.7 0.6 × 520 × 24 × 0.9 13.5 0.02 6

Solving for the pressure P1, we get P1 = 938.58 psia = 923.88 psig This is the pressure at the beginning of pipe segment 2, which is also the end of pipe segment 1.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 109 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

109

Next, we calculate the inlet pressure P1 of pipe segment 1 (12 mi of NPS 16 pipe) for an outlet pressure of 923.88 psig, just calculated. Using the General Flow equation for pipe segment 1, we get

(

)

0.5

2 2 1 520 P1 − 938.58 2.5 100 × 10 = 77.54 14.7 0.6 × 520 × 12 × 0.9 15.25 0.02 6

Solving for pressure P1, we get P1 = 994.75 psia = 980.05 psig This compares well with the pressure of 980.07 psig we calculated earlier using the equivalent length method. Example 6 A natural gas pipeline consists of three different pipe segments connected in series, pumping the same uniform ﬂow rate of 3.0 Mm3/day at 20°C. The ﬁrst segment, DN 500 with 12 mm wall thickness, is 20 km long. The second segment is DN 400, 10 mm wall thickness, and 25 km long. The last segment is DN 300, 6 mm wall thickness, and 10 km long. The inlet pressure is 8500 kPa. Assuming ﬂat terrain, calculate the delivery pressure, using the General Flow equation and the Colebrook friction factor of 0.02. The gas gravity = 0.65 and viscosity = 0.000119 Poise. The compressibility factor Z = 0.9. The base temperature = 15°C and base pressure = 101 kPa. Compare results using the equivalent length method as well as the method using individual pipe segment pressure drops. Solution Inside diameter of ﬁrst pipe segment = 500 – 2 × 12 = 476 mm Inside diameter of second pipe segment = 400 – 2 × 10 = 380 mm Inside diameter of last pipe segment = 300 – 2 × 6 = 288 mm Equivalent length method: Using Equation 3.6, we calculate the total equivalent length of the pipeline system based on the ﬁrst segment diameter DN 500 as follows: 5

500 − 2 × 12 500 − 2 × 12 Le = 20 + 25 × + 10 × 400 − 2 × 10 300 − 2 × 6

5

or Le = 20 + 77.10 + 123.33 = 220.43 km Thus, the given pipeline system can be considered equivalent to a single pipe DN 500, 12 mm wall thickness, 220.43 km long.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 110 Friday, April 1, 2005 3:09 PM

110

GAS PIPELINE HYDRAULICS

The outlet pressure P2 is calculated using General Flow Equation 2.3 as follows:

(

)

8500 2 − P22 15 + 273 3 × 10 6 = 1.1494 × 10 −3 101 0.65 × 293 × 0.9 × 0.02 × 220.43

0.5

(476)2.5

Solving for P2, we get 8500 2 – P22 = 25,908,801 or P2 = 6807 kPa (absolute) We have assumed that the given inlet pressure is in absolute value. Therefore, the delivery pressure is 6807 kPa (absolute). Next, we calculate the delivery pressure considering the three pipe segments treated separately. For the ﬁrst pipe segment 20 km long, we calculate the outlet pressure P2 at the end of the ﬁrst segment as follows. Using General Flow Equation 2.3, we get

(

)

8500 2 − P22 15 + 273 3 × 10 = 1.1494 × 10 101 0.65 × 293 × 0.9 × 0.02 × 20 6

−3

0.5

(476)2.5

Solving for P2, we get P2 = 8361 kPa (absolute) Thus, the pressure at the end of the ﬁrst pipe segment or the beginning of the second segment is 8361 kPa (absolute). Next, we repeat the calculation for the second pipe segment DN 400, 25 km long, using P1 = 8361 kPa (absolute), to calculate P2:

(

)

83612 − P22 15 + 273 3 × 10 = 1.1494 × 10 101 0.65 × 293 × 0.9 × 0.02 × 25 6

−3

0.5

(380)2.5

Solving for P2, we get P2 = 7800 kPa (absolute) This is the pressure at the end of the second pipe segment, which is also the inlet pressure for the third pipe segment.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 111 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

111

Finally, we calculate the outlet pressure of the last pipe segment (DN 300, 10 km) using P1 = 7800 kPa (absolute) as follows:

(

)

7800 2 − P22 15 + 273 3 × 10 6 = 1.1494 × 10 −3 101 0.65 × 293 × 0.9 × 0.02 × 10

0.5

(288)2.5

Solving for P2, we get P2 = 6808 kPa (absolute) Therefore, the delivery pressure is 6808 kPa (absolute). This compares favorably with the value of 6807 kPa we calculated earlier using the equivalent length approach.

3.7 PARALLEL PIPING Sometimes two or more pipes are connected such that the gas ﬂow splits among the branch pipes and eventually combines downstream into a single pipe, as illustrated in Figure 3.7. Such a piping system is referred to as parallel pipes. It is also called a looped piping system, where each parallel pipe is known as a loop. The reason for installing parallel pipes or loops is to reduce pressure drop in a certain section of the pipeline due to pipe pressure limitation or for increasing the ﬂow rate in a bottleneck section. By installing a pipe loop from B to E, in Figure 3.7 we are effectively reducing the overall pressure drop in the pipeline from A to F, since between B and E the ﬂow is split through two pipes. In Figure 3.7 we will assume that the entire pipeline system is in the horizontal plane with no changes in pipe elevations. Gas enters the pipeline at A and ﬂows through the pipe segment AB at a ﬂow rate of Q. At the junction B, the gas ﬂow splits into the two parallel pipe branches BCE and BDE at the ﬂow rates of Q1 and Q2, respectively. At E, the gas ﬂows recombine to equal the initial ﬂow rate Q and continue ﬂowing through the single pipe EF. In order to calculate the pressure drop due to friction in the parallel piping system, we follow two main principles of parallel pipes. The ﬁrst principle is that of conservation of ﬂow at any junction point. The second principle is that there is a common pressure across each parallel pipe. C

Q1

Q A

Q B

E D

Figure 3.7

Parallel piping.

Copyright 2005 by Taylor & Francis Group, LLC

Q2

F

2785_C003.fm Page 112 Friday, April 1, 2005 3:09 PM

112

GAS PIPELINE HYDRAULICS

Applying the principle of ﬂow conservation, at junction B, the incoming ﬂow into B must exactly equal the total outﬂow at B through the parallel pipes. Therefore, at junction B, Q = Q1 + Q2

(3.7)

where Q = inlet ﬂow at A Q1 = ﬂow through pipe branch BCE Q2 = ﬂow through pipe branch BDE According to the second principle of parallel pipes, the pressure drop in pipe branch BCE must equal the pressure drop in pipe branch BDE. This is due to the fact that both pipe branches have a common starting point (B) and common ending point (E). Therefore, the pressure drop in the branch pipe BCE and branch pipe BDE are each equal to (PB – PE ), where PB and PE are the pressures at junctions B and E, respectively. Therefore, we can write ∆PBCE = ∆PBDE = PB – PE

(3.8)

∆P represents pressure drop, and ∆PBCE is a function of the diameter and length of branch BCE and the ﬂow rate Q1. Similarly, ∆PBDE is a function of the diameter and length of branch BDE and the ﬂow rate Q2. In order to calculate the pressure drop in parallel pipes, we must ﬁrst determine the ﬂow split at junction B. From Equation 3.7, we know that the sum of the two ﬂow rates Q1 and Q2 must equal the given inlet ﬂow rate Q. If both pipe loops BCE and BDE are equal in length and pipe inside diameter, we can infer that the ﬂow rate will be split equally between the two branches. Thus, for identical pipe loops, Q1 = Q2 =

Q 2

(3.9)

In this case, the pressure drop from B to E can be calculated assuming a ﬂow rate of Q2 ﬂowing through one of the pipe loops. To illustrate this further, suppose we are interested in determining the pressure at A for the given ﬂow rate Q and a speciﬁed delivery pressure (PF) at the pipe terminus F. We start with the last pipe segment EF and calculate the pressure required at E for a ﬂow rate of Q in order to deliver gas at F at a pressure PF . We could use the General Flow equation for this and substitute PE for upstream pressure, P1, and PF for downstream pressure P2. Having calculated PE , we can now consider one of the pipe loops, such as BCE, and calculate the upstream pressure PB required for a ﬂow rate of Q2 through BCE for a downstream pressure of PE. In the General Flow equation, the upstream pressure P1 = PB and the downstream pressure P2 = PE. It must be noted that this is correct only for identical pipe loops. Otherwise, the ﬂow rate Q1 and Q2 through the pipe branches BCE and BDE will be unequal. From the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 113 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

113

calculated value of PB, we can now determine the pressure required at A by applying the General Flow equation to pipe segment AB that has a gas ﬂow rate of Q. The upstream pressure P1 will be calculated for a downstream pressure P2 = PB. Consider now a situation in which the pipe loops are not identical. This means that the pipe branches BCE and BDE can have different lengths and different diameters. In this case, we must determine the ﬂow split between these two branches by equating the pressure drops through each of the branches in accordance with Equation 3.8. Since Q1 and Q2 are two unknowns, we will use the ﬂow conservation principle and the common pressure drop principle to determine the values of Q1 and Q2. From the General Flow equation we can state the following: The pressure drop due to friction in branch BCE can be calculated from

(P

2 B

where K1 = L1 = D1 = Q1 =

)

− PE2 =

K1L1Q12 D15

(3.10)

a parameter that depends on gas properties, gas temperature, etc. length of pipe branch BCE inside diameter of pipe branch BCE ﬂow rate through pipe branch BCE

Other symbols are as deﬁned earlier. K1 is a parameter that depends on the gas properties, gas temperature, base pressure, and base temperature that will be the same for both pipe branches BCE and BDE in a parallel pipeline system. Hence, we regard this as a constant from branch to branch. Similarly, the pressure drop due to friction in branch BDE is calculated from

(P

2 B

where K2 = L2 = D2 = Q2 =

)

− PE2 =

K 2 L2Q22 D25

(3.11)

a constant like K1 length of pipe branch BDE inside diameter of pipe branch BDE ﬂow rate through pipe branch BDE

Other symbols are as deﬁned earlier. In Equation 3.10 and Equation 3.11, the constants K1 and K2 are equal, since they do not depend on the diameter or length of the branch pipes BCE and BDE. Combining both equations, we can state the following for common pressure drop through each branch: L1Q12 L2Q22 = D15 D25

Copyright 2005 by Taylor & Francis Group, LLC

(3.12)

2785_C003.fm Page 114 Friday, April 1, 2005 3:09 PM

114

GAS PIPELINE HYDRAULICS

Simplifying further, we get the following relationship between the two ﬂow rates Q1 and Q2: Q1 L2 = Q2 L1

0.5

D1 D 2

2.5

(3.13)

Combining Equation 3.13 with Equation 3.7, we can solve for the ﬂow rates Q1 and Q2. To illustrate this, consider the inlet ﬂow Q = 100 MMSCFD and the pipe branches as follows: L1 = 10 mi

D1 = 15.5 in. for branch BCE

L2 = 15 mi

D2 = 13.5 in. for branch BDE

From Equation 3.7, for ﬂow conservation, we get Q1 + Q2 = 100 From Equation 3.13, we get the ratio of ﬂow rates as Q1 15 = Q2 10

0.5

15.5 13.5

2.5

= 1.73

Solving these two equations in Q1 and Q2 , we get Q1 = 63.37 MMSCFD Q2 = 36.63 MMSCFD Once we know the values of Q1 and Q2, we can easily calculate the common pressure drop in the branch pipes BCE and BDE. An example problem (Example 7) will be used to illustrate this method. Another method of calculating pressure drops in parallel pipes is using the equivalent diameter. In this method, we replace the pipe loops BCE and BDE with a certain length of an equivalent diameter pipe that has the same pressure drop as one of the branch pipes. The equivalent diameter pipe can be calculated using the General Flow equation, as explained next. The equivalent pipe with the same ∆P that will replace both branches will have a diameter De and a length equal to one of the branch pipes, say L1. Since the pressure drop in the equivalent diameter pipe, which ﬂows the full volume Q, is the same as that in any of the branch pipes, from Equation 3.10, we can state the following:

(P

2 B

)

− PE2 =

Ke LeQ 2 De5

(3.14)

where Q = Q1 + Q2 from Equation 3.7 and Ke represents the constant for the equivalent diameter pipe of length Le ﬂowing the full volume Q.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 115 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

115

Equating the value of (PB2 – PE2) to the corresponding values, considering each branch separately, we get, using Equation 3.10, Equation 3.11, and Equation 3.14: K1L1Q12 K 2 L2Q22 Ke LeQ 2 = = D15 D25 De5

(3.15)

Also, setting K1 = K2 = Ke and Le = L1, we simplify Equation 3.15 as follows: L1Q12 L2Q22 L1Q 2 = = D15 D25 De5

(3.16)

Using Equation 3.16 in conjunction with Equation 3.7, we solve for the equivalent diameter De as 1/ 5

1 + Const1 2 De = D1 Const1

(3.17)

where 5

D L Const1 = 1 2 D2 L1

(3.18)

and the individual ﬂow rates Q1 and Q2 are calculated from Q1 =

QConst1 1 + Const1

(3.19)

Q2 =

Q 1 + Const1

(3.20)

and

To illustrate the equivalent diameter method, consider the inlet ﬂow Q = 100 MMSCFD and the pipe branches as follows: L1 = 10 mi

D1 = 15.5 in. for branch BCE

L2 = 15 mi

D2 = 13.5 in. for branch BDE

From Equation 3.18, 5

15.5 15 Const1 = = 1.73 13.5 10

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 116 Friday, April 1, 2005 3:09 PM

116

GAS PIPELINE HYDRAULICS

Using Equation 3.17, the equivalent diameter is 1/ 5

1 + 1.73 2 De = 15.5 1.73

= 18.60 in.

Thus, the NPS 16 and NPS 14 pipes in parallel can be replaced with an equivalent pipe having an inside diameter of 18.6 in. From Equation 3.19 and Equation 3.20, we get the ﬂow rates in the two branch pipes as follows: Q1 =

100 × 1.73 = 63.37 MMSCFD 1 + 1.73

and Q2 = 36.63 MMSCFD Having calculated an equivalent diameter De, we can now calculate the common pressure drop in the parallel branches by considering the entire ﬂow Q ﬂowing through the equivalent diameter pipe. An example problem will illustrate this method. Example 7 A gas pipeline consists of two parallel pipes, as shown in Figure 3.7. It is designed to operate at a ﬂow rate of 100 MMSCFD. The ﬁrst pipe segment AB is 12 miles long and consists of NPS 16, 0.250 in. wall thickness pipe. The loop BCE is 24 mi long and consists of NPS 14, 0.250 in. wall thickness pipe. The loop BDE is 16 miles long and consists of NPS 12, 0.250 in. wall thickness pipe. The last segment EF is 20 miles long and consists of NPS 16, 0.250 in. wall thickness pipe. Assuming a gas gravity of 0.6, calculate the outlet pressure at F and the pressures at the beginning and the end of the pipe loops and the ﬂow rates through them. The inlet pressure at A = 1200 psig. The gas ﬂowing temperature = 80°F, base temperature = 60°F, and base pressure = 14.73 psia. The compressibility factor Z = 0.92. Use the General Flow equation with Colebrook friction factor f = 0.015. Solution From Equation 3.13, the ratio of the ﬂow rates through the two pipe loops is given by Q1 16 = Q2 24

0.5

14 − 2 × 0.25 12.75 − 2 × 0.25

and from Equation 3.7 Q1 + Q2 = 100

Copyright 2005 by Taylor & Francis Group, LLC

2.5

= 1.041

2785_C003.fm Page 117 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

117

Solving for Q1 and Q2, we get Q1 = 51.0 MMSCFD

and Q2 = 49.0 MMSCFD

Next, considering the ﬁrst pipe segment AB, we will calculate the pressure at B based upon the inlet pressure of 1200 psig at A, using General Flow Equation 2.2, as follows:

(

)

0.5

2 2 1 520 1214.73 − P2 2.5 100 × 10 = 77.54 14.73 0.6 × 540 × 12 × 0.92 15.5 0.015 6

Solving for the pressure at B, we get P2 = 1181.33 psia = 1166.6 psig This is the pressure at the beginning of the looped section at B. Next we calculate the outlet pressure at E of pipe branch BCE, considering a ﬂow rate of 51 MMSCFD through the NPS 14 pipe, starting at a pressure of 1181.33 psia at B. Using the General Flow equation, we get

(

)

0.5

2 2 1 520 1181.333 − P2 2.5 51 × 10 = 77.54 14.73 0.6 × 540 × 24 × 0.92 13.5 0.015 6

Solving for the pressure at E, we get P2 = 1145.63 psia = 1130.9 psig Next, we use this pressure as the inlet pressure for the last pipe segment EF and calculate the outlet pressure at F using the General Flow equation, as follows:

(

)

0.5

2 2 1 520 1145.63 − P2 2.5 100 × 10 6 = 77.54 14.73 0.6 × 540 × 20 × 0.92 15.5 0.015

Solving for the outlet pressure at F, we get P2 = 1085.85 psia = 1071.12 psig In summary, the calculated results are as follows: Pressure at the beginning of pipe loops = 1166.6 psig Pressure at the end of pipe loops = 1130.9 psig Outlet pressure at the end of pipeline = 1071.12 psig Flow rate in NPS 14 loop = 51 MMSCFD Flow rate in NPS 12 loop = 49 MMSCFD

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 118 Friday, April 1, 2005 3:09 PM

118

GAS PIPELINE HYDRAULICS

We will now calculate the pressures using the equivalent diameter method. From Equation 3.18, 5

13.5 16 Const 1 = = 1.041 12.25 24 From Equation 3.17, the equivalent diameter is 1/ 5

1 + 1.041 2 De = 13.5 1.041

= 17.67 in.

Thus, we can replace the two branch pipes between B and E with a single piece of pipe 24 mi long, having an inside diameter of 17.67 in., ﬂowing 100 MMSCFD. The pressure at B was calculated earlier as PB = 1181.33 psia Using this pressure, we can calculate the downstream pressure at E for the equivalent pipe diameter as follows:

(

)

0.5

2 2 1 520 1181.33 − P2 17.672.5 100 × 10 = 77.54 0.015 14.73 0.6 × 540 × 24 × 0.92 6

Solving for the outlet pressure at E, we get P2 = 1145.60 psia, which is almost the same as what we calculated before. The pressure at F will therefore be the same as what we calculated before. Therefore, using the equivalent diameter method, the parallel pipes BCE and BDE can be replaced with a single pipe 24 mi long, having an inside diameter of 17.67 in. Example 8 A natural gas pipeline DN 500 with 12 mm wall thickness is 60 km long. The gas ﬂow rate is 5.0 Mm3/day at 20°C. Calculate the inlet pressure required for a delivery pressure of 4 MPa (absolute), using the General Flow equation with the modiﬁed Colebrook-White friction factor. The pipe roughness = 0.015 mm. In order to increase the ﬂow rate through the pipeline, the entire line is looped with a DN 500 pipeline, 12 mm wall thickness. Assuming the same delivery pressure, calculate the inlet pressure at the new ﬂow rate of 8 Mm3/day. The gas gravity = 0.65 and viscosity = 0.000119 Poise. The compressibility factor Z = 0.88. The base temperature = 15°C, and the base pressure = 101 kPa. If the inlet and outlet pressures are held the same as before, what length of the pipe should be looped to achieve the increased ﬂow?

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 119 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

119

Solution Pipe inside diameter D = 500 − 2 × 12 = 476 mm Flow rate Q = 5.0 × 106 m3/day Base temperature Tb = 15 + 273 = 288 K Gas ﬂow temperature Tf = 20 + 273 = 293 K Delivery pressure P2 = 4 MPa The Reynolds number, using Equation 2.35, is 101 0.65 × 5 × 10 6 R = 0.5134 = 10, 330, 330 288 0.000119 × 476 From the modiﬁed Colebrook-White Equation 2.47, the transmission factor is 0.015 1.4125F F = − 4 Log10 + 3.7 × 476 10, 330, 330 Solving by successive iteration, we get F = 19.80 Using General Flow Equation 2.8, the inlet pressure is calculated next: P12 − 4000 2 273 + 15 5 × 10 6 = 5.747 × 10 −4 × 19.80 101 0.65 × 293 × 60 × 0.88

0.5

× (476)2.5

Solving for the inlet pressure, we get P1 = 5077 kPa (absolute) = 5.08 MPa (absolute) Therefore, the inlet pressure required at 5 Mm3/day ﬂow rate is 5.08 MPa. Next, at 8 Mm3/day ﬂow rate, we calculate the new inlet pressure with the entire 60 km length looped with an identical DN 500 pipe. Since the loop is the same size as the main line, each parallel branch will carry half the total ﬂow rate or 4 Mm3/day. We calculate the Reynolds number for ﬂow through one of the loops using Equation 2.35: 101 0.65 × 4 × 10 6 R = 0.5134 = 8, 264, 264 288 0.000119 × 476

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 120 Friday, April 1, 2005 3:09 PM

120

GAS PIPELINE HYDRAULICS

From the modiﬁed Colebrook-White Equation 2.47, the transmission factor is 0.015 1.4125F F = − 4 Log10 + 3.7 × 476 8, 264, 264 Solving by successive iteration, we get F = 19.70 Keeping the delivery pressure the same as before (4 MPa), using General Flow Equation 2.8, we calculate the inlet pressure required as follows: P12 − 4000 2 273 + 15 4 × 10 = 5.747 × 10 × 19.70 101 . × × × . 0 65 293 60 0 88 6

0.5

−4

× (476)2.5

Solving for the inlet pressure, we get P1 = 4724 kPa (absolute) = 4.72 MPa (absolute) Therefore, for the fully looped pipeline at 8 Mm3/day ﬂow rate, the inlet pressure required is 4.72 MPa. Next, keeping the inlet and outlet pressures the same at 5077 kPa and 4000 kPa, respectively, at the new ﬂow rate of 8 Mm3/day, we assume L km of the pipe from the inlet is looped. We will calculate the value of L by ﬁrst calculating the pressure at the point where the loop ends. Since each parallel pipe carries 4 Mm3/day, we use the Reynolds number and transmission factor calculated earlier: R = 8,264,264

and F = 19.70

Using General Flow Equation 2.8, we calculate the outlet pressure at the end of the loop of length L km as follows: 50772 − P22 273 + 15 4 × 10 6 = 5.747 × 10 −4 × 19.70 101 0.65 × 293 × L × 0.88

0.5

× (476)2.5

Solving for pressure in terms of the loop length L, we get P22 = 50772 − 105, 291.13L

(3.21)

Next, we apply the General Flow equation for the pipe segment of length (60 – L) km that carries the full 8 Mm3/day ﬂow rate. The inlet pressure is P2 and the outlet pressure is 4000 kPa. The Reynolds number at 8 Mm3/day is 101 0.65 × 8 × 10 6 R = 0.5134 = 16, 528, 528 288 0.000119 × 476

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 121 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

121

From the modiﬁed Colebrook-White Equation 2.47, the transmission factor is 0.015 1.4125F F = − 4 Log10 + 3.7 × 476 16, 528, 528 Solving by successive iteration, we get F = 19.96 Using General Flow Equation 2.8, we calculate the inlet pressure for the pipe segment of length (60 − L) km as follows: P22 − 4000 2 273 + 15 8 × 10 6 = 5.747 × 10 −4 × 19.96 101 0.65 × 293 × (60 − L ) × 0.88

0.5

× (476)2.5

Simplifying, we get P22 = 4000 2 + 410, 263.77 (60 − L )

(3.22)

From Equation 3.21 and Equation 3.22, eliminating P2, we solve for L as follows: 50772 − 105, 291.13L = 4000 2 + 410, 263.77 (60 − L ) Therefore, L = 48.66 km Thus, 48.66 km out of the 60 km pipeline length will have to be looped starting at the pipe inlet so that at 8 Mm3/day both inlet and outlet pressures will be the same as before at 5 Mm3/day. What will be the effect if the loop was installed starting at the downstream end of the pipeline and proceeding toward the upstream end? Will the results be the same? In the next section we will explore the best location to install the pipe loop.

3.8 LOCATING PIPE LOOP In the preceding example, we looked at looping an entire pipeline to reduce pressure drop and increase the ﬂow rate. We also explored looping a portion of the pipe, beginning at the upstream end. How do we determine where the loop should be placed for optimum results? Should it be located upstream, downstream, or in a midsection of the pipe? We will analyze this in this section. Three looping scenarios are presented in Figure 3.8. In case (a), a pipeline of length L is shown looped with X miles of pipe, beginning at the upstream end A. In case (b), the same length X of pipe is looped, but it is located on the downstream end B. Case (c) shows the midsection of the pipeline being looped. For most practical purposes, we can say that the cost of all three loops will be the same as long as the loop length is the same.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 122 Friday, April 1, 2005 3:09 PM

122

GAS PIPELINE HYDRAULICS

Flow A

B

C (a) Upstream loop

Flow A

B

C (b) Downstream loop

Flow A

C

D

B

(c) Midsection loop Figure 3.8

Different looping scenarios.

In order to determine which of these cases is optimum, we must analyze how the pressure drop in the pipeline varies with distance from the pipe inlet to outlet. It is found that if the gas temperature is constant throughout, at locations near the upstream end, the pressure drops at a slower rate than at the downstream end. Therefore, there is more pressure drop in the downstream section compared to that in the upstream section. Hence, to reduce the overall pressure drop, the loop must be installed toward the downstream end of the pipe. This argument is valid only if the gas temperature is constant throughout the pipeline. In reality, due to heat transfer between the ﬂowing gas and the surrounding soil (buried pipe) or the outside air (above-ground pipe), the gas temperature will change along the length of the pipeline. If the gas temperature at the pipe inlet is higher than that of the surrounding soil (buried pipe), the gas will lose heat to the soil and the temperature will drop from the pipe inlet to the pipe outlet. If the gas is compressed at the inlet using a compressor, then the gas temperature will be much higher than that of the soil immediately downstream of the compressor. The hotter gas will cause higher pressure drops (examine the General Flow equation and see how the pressure varies with the gas ﬂow temperature). Hence, in this case the upstream segment will have a larger pressure drop compared to the downstream segment. Therefore, considering heat transfer effects, the pipe loop should be installed in the upstream portion for maximum beneﬁt. The installation of the pipe loop in the midsection of the pipeline, as in case (c) in Figure 3.8, will not be the optimum location, based on the preceding discussion. It can therefore be concluded that if the gas temperature is fairly constant along the pipeline, the loop should be installed toward the downstream end, as in case (b). If heat transfer is taken into account and the gas temperature varies along the pipeline, with the hotter gas being upstream, the better location for the pipe loop will be on the upstream end, as in case (a). Looping pipes will be explored more in Chapter 5 and in Chapter 10, where we discuss pipeline economics.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 123 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

123

Pressure

1000 psig

800 psig

Flow B

A

Distance

Figure 3.9

Hydraulic pressure gradient for uniform flow.

3.9 HYDRAULIC PRESSURE GRADIENT

Pressure

The hydraulic pressure gradient is a graphical representation of the gas pressures along the pipeline, as shown in Figure 3.9. The horizontal axis shows the distance along the pipeline starting at the upstream end. The vertical axis depicts the pipeline pressures. Since pressure in a gas pipeline is nonlinear compared to liquid pipelines, the hydraulic gradient for a gas pipeline appears to be a slightly curved line instead of a straight line. The slope of the hydraulic gradient at any point represents the pressure loss due to friction per unit length of pipe. As discussed earlier, this slope is more pronounced as we move toward the downstream end of the pipeline, since the pressure drop is larger toward the end of the pipeline. If the ﬂow rate through the pipeline is a constant value (no intermediate injections or deliveries) and pipe size is uniform throughout, the hydraulic gradient appears to be a slightly curved line, as shown in Figure 3.9, with no appreciable breaks. If there are intermediate deliveries or injections along the pipeline, the hydraulic gradient will be a series of broken lines, as indicated in Figure 3.10.

Q A Q1

Figure 3.10

B

Distance Q2

Hydraulic pressure gradient for deliveries and injections.

Copyright 2005 by Taylor & Francis Group, LLC

Q3

2785_C003.fm Page 124 Friday, April 1, 2005 3:09 PM

124

GAS PIPELINE HYDRAULICS

A similar broken hydraulic gradient can also be seen in the case of a pipeline with variable pipe diameters and wall thicknesses, even if the ﬂow rate is constant. Unlike liquid pipelines, the breaks in hydraulic pressure gradient are not as conspicuous in gas pipelines. In a long-distance gas pipeline, due to limitations of pipe pressure, intermediate compressor stations will be installed to boost the gas pressure to the required value so the gas can be delivered at the contract delivery pressure at the end of the pipeline. For example, consider a 200 mi long, NPS 16 pipeline that transports 150 MMSCFD of gas from Compton to a delivery point at Beaumont, as shown in Figure 3.11. Suppose calculations show that 1600 psig pressure is required at Compton in order to deliver gas to Beaumont at 800 psig. If the maximum operating pressure (MOP) of this pipeline is limited to 1350 psig, obviously we will need more than one compressor station. The ﬁrst compressor station will be located at Compton and will provide a pressure of 1350 psig. As gas ﬂows from Compton toward some intermediate location, such as Sheridan, the gas pressure will drop to some value such as 900 psig. At Sheridan, a second compressor station will boost the gas pressure to 1350 psig on its way to the terminus at Beaumont. By installing the second compressor station at Sheridan, pipeline pressures are maintained within MOP limits. The actual location of the intermediate compressor station at Sheridan will depend upon many factors, including pipeline elevation proﬁle, the gas pressure at Sheridan, and the delivery pressure required at Beaumont. The hydraulic pressure gradient in this case is as shown in Figure 3.11. In the preceding discussion, we picked an arbitrary pressure of 900 psig at Sheridan. This gives us an approximate compression ratio of 1350 + 14.7 = 1.492 914.7 which is a reasonable number for centrifugal compressors used in gas pipelines. In reference to Figure 3.11, we will now outline the method of locating the intermediate 1600 psig 1350 psig

1350 psig

900 psig 800 psig

Flow A Compton

Distance

Sheridan

NPS 16 pipeline 200 mi long Figure 3.11

Compton to Beaumont pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

B Beaumont

2785_C003.fm Page 125 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

125

compressor station at Sheridan. In Chapter 4, we will further analyze gas pipelines with multiple compressor stations. Starting at Compton, with an inlet pressure P1 = 1350 psig, we calculate the pipe length L that will cause the pressure to drop to 900 psig, using the General Flow equation. Assuming a ﬂow rate of 150 MMSCFD and a friction factor f = 0.01, we get 1 520 1364.72 − 914.72 150 × 10 = 77.54 0.01 14.7 0.6 × 520 × L × 0.9

0.5

6

× (15.5)2.5

Solving for L, we get L = 109.28 mi Thus, the approximate location of the second compressor station at Sheridan is 109.28 mi from Compton. If we allow the compressor at Sheridan to boost the gas pressure to 1350 psig, the compression ratio is r=

1350 + 14.7 = 1.492 914.7

which is a reasonable ratio for a centrifugal compressor. Therefore, starting at 1350 psig at Sheridan, we calculate the delivery pressure at Beaumont using the General Flow equation for (200 – 109.28) = 90.7 mi of pipe as follows: 1 520 1364.72 − P2 2 150 × 10 = 77.54 14.7 0.6 × 520 × 90.7 × 0.9 0.01 6

0.5

× (15.5)2.5

Solving for P2, we get P2 = 1005.5 psia = 990.82 psig This is more than the required delivery pressure at Beaumont of 800 psig. We could go back and repeat the above calculations, considering slightly lower pressure at Compton, say 1300 psig, in order to get the correct delivery pressure of 800 psig at Beaumont. This is left as an exercise for the reader. Alternatively, we could start with the required delivery pressure of 800 psig at Beaumont and work backward to determine the distance at which the upstream pressure reaches 1350 psig. That would be the location for the Sheridan compressor station. Next, we would determine the pressure at Sheridan, beginning with the 1350 psig upstream pressure at Compton. This would establish the suction pressure of the Sheridan compressor station. Knowing the suction pressure and the discharge pressure at Sheridan, we could calculate the compression ratio required. In Chapter 4 we will discuss multiple compressor stations in more detail.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 126 Friday, April 1, 2005 3:09 PM

126

GAS PIPELINE HYDRAULICS

3.10 PRESSURE REGULATORS AND RELIEF VALVES In a long-distance gas pipeline with intermediate delivery points, there may be a need to regulate the gas pressure at certain delivery points in order to satisfy the customer requirements. Suppose the pressure at the delivery point B in Figure 3.11 is 800 psig, whereas the customer requirement is only 500 psig. Obviously, some means of reducing the gas pressure must be provided so that the customer can utilize the gas for his or her requirements at the correct pressure. This is achieved by means of a pressure regulator that will ensure a constant pressure downstream of the delivery point, regardless of the pressure on the upstream side of the pressure regulator. This concept is further illustrated using the example pipeline shown in Figure 3.12. The main pipeline from A to C is shown along with a branch pipe BE. The ﬂow rate from A to B is 100 MMSCFD, with an inlet pressure of 1200 psig at A. At B, gas is delivered into a branch line BE at the rate of 30 MMSCFD. The remaining volume of 70 MMSCFD is delivered to the pipeline terminus C at a delivery pressure of 600 psig. Based on the delivery pressure requirement of 600 psig at C and a takeoff of 30 MMSCFD at point B, the calculated pressure at B is 900 psig. Starting with 900 psig on the branch line at B, at 30 MMSCFD, gas is delivered to point E at 600 psig. If the actual requirement at E is only 400 psig, a pressure regulator will be installed at E to reduce the delivery pressure by 200 psig. It can be seen from Figure 3.12 that at point D immediately upstream of the pressure regulator, the gas pressure is approximately 600 psig and is regulated to 400 psig downstream at E. If the mainline ﬂow rate changes from 100 MMSCFD to 90 MMSCFD and the delivery at B is maintained at 30 MMSCFD, the gas pressure at B will reduce to a value below 900 psig. Accordingly, the pressure at point D in the branch pipe BE will also reduce to some value below 600 psig. Regardless, due to the pressure regulator, the pressure at E will be maintained at the required 400 psig. However, if for some reason the pressure upstream of the regulator at D falls below 400 psig, the downstream pressure at E cannot be maintained at the original value of 400 psig. The pressure regulator can only reduce the pressure downstream to the required value. It cannot increase the pressure beyond the pressure on the upstream

400 psig E

D 30 MMSCFD 100 MMSCFD

70 MMSCFD C 600 psig

A 1200 psig Figure 3.12

B

Pressure regulation.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 127 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

127

side. If the pressure at D drops to 300 psig, the pressure regulator is ineffective and will remain fully open, and the delivery pressure at E will be 300 psig as well. Example 9 A natural gas pipeline, NPS 16, 0.250 in. wall thickness, 50 mi long, with a branch pipe (NPS 8, 0.250 in. wall thickness, 15 mi long), as shown in Figure 3.13, is used to transport 100 MMSCFD gas (gravity = 0.6 and viscosity = 0.000008 lb/ft-s) from A to B. At B (milepost 20), a delivery of 30 MMSCFD occurs into the branch pipe BE. The delivery pressure at E must be maintained at 300 psig. The remaining volume of 70 MMSCFD is shipped to the terminus C at a delivery pressure of 600 psig. Assume a constant gas temperature of 60°F and a pipeline efﬁciency of 0.95. The base temperature and base pressure are 60°F and 14.7 psia, respectively. The compressibility factor Z = 0.88. a) Using the Panhandle A equation, calculate the inlet pressure required at A. b) Is a pressure regulator required at E? c) If the inlet ﬂow at A drops to 60 MMSCFD, what is the impact in the branch pipeline BE if the ﬂow rate of 30 MMSCFD is maintained? Solution Pipe inside diameter for pipe segment AB and BC = 16 – 2 × 0.25 = 15.5 in. First, we will consider the pipe segment BC and calculate the pressure P1 at B for a 70 MMSCFD ﬂow rate to deliver gas at 600 psig at C. Using Panhandle A Equation 2.55, neglecting elevation effects, 60 + 460 70 × 10 6 = 435.87 × 0.95 14.7

1.0788

P12 − 614.72 0.8539 × 520 × 30 × 0.88 0.6

0.5394

(15.5)2.6182

300 psig E

D NPS 8

30 MMSCFD

100 MMSCFD

A 1200 psig Figure 3.13

NPS 16

70 MMSCFD C 600 psig B

Example problem—pressure regulation.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 128 Friday, April 1, 2005 3:09 PM

128

GAS PIPELINE HYDRAULICS

Solving for pressure at B, we get P1 = 660.39 psia = 645.69 psig Next, considering pipe segment AB, ﬂowing 100 MMSCFD, we calculate the inlet pressure P1 at A using the outlet pressure 660.39 psia we calculated at B. From Panhandle A Equation 2.55, 520 100 × 10 = 435.87 × 0.95 14.7

1.0788

6

P12 − 660.392 0.8539 × 520 × 20 × 0.88 0.6

0.5394

(15.5)2.6182

Solving for the pressure at A, we get P1 = 715.08 psia = 700.38 psig Next, using the pressure 660.39 psia at B, we calculate the outlet pressure of branch BE that ﬂows 30 MMSCFD through the 15 mi NPS 8 pipe. Using Panhandle A Equation 2.55, 520 30 × 10 6 = 435.87 × 0.95 14.7

1.0788

660.392 − P22 0.8539 × 520 × 15 × 0.88 0.6

0.5394

(8.125)2.6182

Solving for the pressure at E, P2 = 544.90 psia = 530.2 psig Since the required delivery pressure at E is 300 psig, a pressure regulator will be required at E. If the ﬂow rate at A drops to 60 MMSCFD and the branch BE ﬂow rate is maintained at 30 MMSCFD, we will calculate the junction pressure at B by using Panhandle A Equation 2.55 for the pipe segment BC, considering a ﬂow rate of 30 MMSCFD and a delivery pressure of 600 psig at C. 520 30 × 10 = 435.87 × 0.95 14.7

1.0788

6

P12 − 614.72 0.8539 × 520 × 30 × 0.88 0.6

0.5394

(15.5)2.6182

Solving for the pressure at B, we get P1 = 624.47 psia = 609.77 psig Using the pressure at B, we calculate the outlet pressure at E on branch BE for a 30 MMSCFD ﬂow rate using Panhandle A Equation 2.55: 520 30 × 10 6 = 435.87 × 0.95 14.7

Copyright 2005 by Taylor & Francis Group, LLC

1.0788

624.472 − P22 0.8539 × 520 × 15 × 0.88 0.6

0.5394

(8.125)2.6182

2785_C003.fm Page 129 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

129

Solving for the pressure P2, we get P2 = 500.76 psia = 486.06 psig This is the new pressure at E. Comparing this pressure with the previously calculated pressure of 530.2 psig, we see that the delivery pressure at E has dropped by 44 psig, approximately. To maintain the delivery pressure of 300 psig at E, a pressure regulator is still required. Therefore, the answers are a) Inlet pressure at A = 700.38 psig. b) A pressure regulator is required at E to reduce the pressure from 530.2 psig to 300 psig. c) Finally, a pressure regulator is required at E to reduce the pressure from 486.1 psig to 300 psig.

3.11 TEMPERATURE VARIATION AND GAS PIPELINE MODELING

Gas temperature

In the preceding sections we assumed the gas temperature to be constant (isothermal) along the length of the pipeline. By assuming isothermal ﬂow, we were able to calculate the pressure drop using constant gas properties such as the compressibility factor. In reality, the temperature of gas in a buried pipeline varies along the length of the pipeline due to heat transfer between the gas and the surrounding soil. If the inlet temperature of the gas is 80°F and the surrounding soil is 60°F, the temperature difference will cause heat loss from the gas to the soil. Additionally, in a longdistance pipeline, the soil temperature can vary along the pipeline. This will cause the gas temperature to vary, as shown in Figure 3.14.

T Ts

Ground temperature

x Distance

Figure 3.14

Temperature variation in a gas pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 130 Friday, April 1, 2005 3:09 PM

130

GAS PIPELINE HYDRAULICS

Generally, if the pipeline is fairly long, the gas temperature will ultimately equal the soil temperature as gas approaches the delivery point. Due to such variation in gas temperature, calculation of pressure drop must be made by considering short lengths of pipe that make up the total pipeline. For example, if the pipeline is 50 mi long, we will subdivide the pipeline into short segments of 1- or 2-mile lengths and apply the General Flow equation for each pipe segment. Starting with the upstream pressure of segment 1, the downstream pressure will be calculated, assuming an average temperature for segment 1. Next, using the calculated downstream pressure as the upstream pressure for segment 2, we calculate the downstream pressure for segment 2. The process is continued until all segments of the pipeline are covered. It must be noted that the variation of temperature from segment to segment must be taken into account to calculate the compressibility factor to be used in the General Flow equation. The calculation of gas temperature at any point along the pipeline by taking into account the heat transfer between the gas and surrounding soil is quite complicated. It does not lend itself easily to manual calculations. The method of calculation will be discussed brieﬂy in this section for information only. To accurately take into account the temperature variations, a suitable gas pipeline hydraulics simulation program must be used, since, as indicated earlier, manual calculation is quite laborious and time consuming. Several commercial simulation programs are available to model steady-state gas pipeline hydraulics. These programs calculate the gas temperature and pressures by taking into consideration variations of soil temperature, pipe burial depth, and thermal conductivities of pipe, insulation, and soil. One such software program is GASMOD, marketed by SYSTEK Technologies, Inc. (www.systek.us). Appendix A includes a sample simulation of a gas pipeline using the GASMOD software. Even though manual calculation of the temperature variation and corresponding pressure drop in a gas pipeline is quite tedious, we will present here the basic equations for reference. Consider a buried pipeline transporting gas from point A to point B. We will analyze a short segment of length ∆L of this pipe as shown in Figure 3.15 and apply the principles of heat transfer to determine how the gas temperature varies along the pipeline. The upstream end of the pipe segment of length ∆L is at a temperature T1 and the downstream end at temperature T2. The average gas temperature in this segment is

Soil temperature Ts

A

Q

T1

T

T2

L 1

Figure 3.15

Analysis of temperature variation.

Copyright 2005 by Taylor & Francis Group, LLC

2

B

2785_C003.fm Page 131 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

131

represented by T. The outside soil temperature at this location is Ts. Assume steadystate conditions and the mass ﬂow rate of gas to be m. The gas ﬂow from the upstream end to the downstream end of the segment causes a temperature drop of ∆T. The heat loss from the gas can be represented by ∆H = −mCp∆T where ∆H m Cp ∆T

= = = =

(3.23)

heat transfer rate, Btu/h mass ﬂow rate of gas, lb/h average speciﬁc heat of gas, Btu/lb/°F temperature difference = T1 – T2 °F

The negative sign in Equation 3.23 indicates loss of heat from upstream temperature T1 to downstream temperature T2. Next, we consider the heat transfer from the gas to the surrounding soil in terms of the overall heat transfer coefﬁcient U and the difference in temperature between the gas and surrounding soil, represented by (T – Ts). Therefore, we can write the following equation for heat transfer: ∆H = U∆A (T – Ts) where U = ∆A = T = Ts = D =

(3.24)

overall heat transfer coefﬁcient, Btu/h/ft2/ °F surface area of pipe for heat transfer = pD∆L average gas temperature in pipe segment, °F average soil temperature surrounding pipe segment, °F pipe inside diameter, ft

Equating the two values of heat transfer rate ∆H from Equation 3.23 and Equation 3.24, we get −mCp∆T = U∆A(T – Ts) Simplifying, we get πUD ∆T ∆L = − T − Ts mCp

(3.25)

Rewriting Equation 3.25 in differential form and integrating, we get

∫

2

1

dT = T − Ts

πUD

∫ − mCp dL 1

(3.26)

2

Integrating and simplifying, we get T2 − Ts = e −θ T1 − Ts

Copyright 2005 by Taylor & Francis Group, LLC

(3.27)

2785_C003.fm Page 132 Friday, April 1, 2005 3:09 PM

GAS PIPELINE HYDRAULICS

Gas temperature

132

Ts

Ground temperature

Distance Figure 3.16

Joule-Thompson effect in gas pipeline.

where

θ=

πUD∆L mCp

(3.28)

Simplifying Equation 3.27 further, we get the downstream temperature of the pipe segment of length ∆L as T2 = Ts + (T1 − Ts )e −θ

(3.29)

It can be seen from Equation 3.29 that as the pipe length increases, the term e −θ approaches zero and the temperature, T2, becomes equal to soil temperature, Ts. Therefore, in a long gas pipeline, the gas temperature ultimately equals the surrounding soil temperature. This is illustrated in Figure 3.14. In the preceding analysis, we made several simplifying assumptions. We assumed that the soil temperature and the overall heat transfer coefﬁcient remained constant and ignored the Joule-Thompson effect as gas expands through a pipeline. In a long pipeline, the soil temperature can actually vary along the pipeline and, therefore, must be taken into account in these calculations. One approach would be to subdivide the pipeline into segments that have constant soil temperatures and perform calculations for each segment separately. The Joule-Thompson effect causes the gas to cool slightly due to expansion. Therefore, in a long pipeline, the gas temperature at the delivery point may fall below that of the ground or soil temperature, as indicated in Figure 3.16.

3.12 LINE PACK As gas ﬂows through a pipeline from point A to point B, the pressures and temperatures vary along the pipeline length. The volume of gas contained in a given length of pipeline is simply the physical volume of the pipe segment. For example, a 1-mile

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 133 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

133

section of NPS 16 pipe can have a physical volume of 7000 ft3. Therefore, this volume represents the volume of the gas in this 1-mile section at the actual gas temperature and pressure. The quantity of gas contained within the pipeline under pressure, measured at standard conditions (generally 14.7 psia and 60°F), is termed the line pack volume. Consider a segment of pipe, of length L, with upstream pressure and temperature of P1 and T1 and downstream values of P2 and T2, respectively. We can calculate the line pack using the gas laws discussed in Chapter 1. Suppose the inside diameter of the pipe is D; then the physical volume Vp of the pipe section is Vp =

π 2 DL 4

(3.30)

This volume is the gas volume at pressures and temperatures ranging from P1, T1 at the upstream end to P2, T2 at the downstream end of the pipe length L. In order to convert this volume to standard conditions of pressure, Pb, and temperature, Tb, we apply the gas law Equation 1.16 as follows: P V PbVb = avg p Z bTb ZavgTavg where Pavg Tavg Zavg Zb

= = = =

(3.31)

average gas pressure in pipe segment average gas temperature in pipe segment average gas compressibility factor at Tavg and Pavg compressibility factor at base conditions = 1.00, approximately

The average pressure, Pavg, is calculated from the upstream and downstream pressures P1 and P2 using Equation 2.14. The average temperature can be taken as the arithmetic mean of the upstream and downstream temperatures T1 and T2. This approach for average temperature will be accurate only if we consider short segments of pipe. From Equation 3.31, solving for line pack Vb at standard conditions, we get T P V Vb = b avg p Pb ZavgTavg

(3.32)

Substituting the value of Vp from Equation 3.30 and simplifying, we get T P Vb = 0.7854 b avg ( D 2 L ) Pb ZavgTavg where Vb = line pack in pipe segment, standard ft3 D = pipe inside diameter, ft L = pipe segment length, ft Other symbols are as deﬁned before.

Copyright 2005 by Taylor & Francis Group, LLC

(3.33)

2785_C003.fm Page 134 Friday, April 1, 2005 3:09 PM

134

GAS PIPELINE HYDRAULICS

Equation 3.33 is modiﬁed in terms of commonly used pipeline units as follows: T P Vb = 28.798 b avg ( D 2 L ) Pb ZavgTavg

(USCS units)

(3.34)

where Vb = line pack in pipe segment, standard ft3 D = pipe inside diameter, in. L = pipe segment length, mi Other symbols are as deﬁned before. The corresponding equation in SI units is T P Vb = 7.855 × 10 −4 b avg ( D 2 L) Pb ZavgTavg

(SI units)

(3.35)

where Vb = line pack in pipe segment, standard m3 D = pipe inside diameter, mm L = pipe segment length, km Other symbols are as deﬁned before. Since the pressure and temperature in a gas pipeline vary along the length, to improve the accuracy of calculations, the line pack volume Vb is calculated for short segments of pipe and summed to obtain the line pack of the entire pipeline. Example 10 A natural gas pipeline is 10 mi long and has an inlet pressure of 1000 psig and outlet pressure of 900 psig when transporting 100 MMSCFD. The base pressure and base temperature are 14.7 psia and 60°F, respectively. If the pipe is NPS 16, 0.250 in. wall thickness, calculate the line pack assuming an average gas temperature of 78°F. Use an average compressibility factor of 0.90. Solution Pipe inside diameter = 16 – 2 × 0.250 = 15.5 in. The average pressure is calculated from Equation 2.14 as follows:

Pavg =

Copyright 2005 by Taylor & Francis Group, LLC

1014.7 × 914.7 2 1014.7 + 914.7 − = 965.56 psia 1014.7 + 914.77 3

2785_C003.fm Page 135 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

135

Using Equation 3.34, we calculate the line pack as follows: 60 + 460 965.56 15.52 × 10 = 4,880,521 standard ft3 Vb = 28.798 14.7 78 + 460 0.90 Therefore, the line pack is 4,880,521 standard ft3. Example 11 A natural gas pipeline is 20 km long and has an inlet pressure of 8000 kPa (gauge) and outlet pressure of 5000 kPa (gauge) when transporting 5 Mm3/day. The base pressure and base temperature are 101 kPa and 15°C, respectively. If the pipe is DN 500, 12 mm wall thickness, calculate the line pack assuming an average gas temperature of 20°C. Use an average compressibility factor of 0.90. Solution Pipe inside diameter = 500 – 2 × 12 = 476 mm The average pressure is calculated from Equation 2.14 as follows: Pavg =

8101 × 5101 2 8101 + 5101 − = 6714.62 kPa (absolute) 8101 + 5101 3

Using Equation 3.35, we calculate the line pack as follows: 15 + 273 6714.62 4762 × 20 Vb = 7.855 × 10 −4 = 258,448 standard m3 0.9 101 20 + 273 Therefore, the line pack is 258,448 standard m3.

3.13 SUMMARY In this chapter we continued to look at the application of the pressure drop equations introduced in Chapter 2. Several piping conﬁgurations, such as pipes in series, pipes in parallel, and gas pipelines with injections and deliveries, were analyzed to determine pressures required and pipe size needed to satisfy certain requirements. The concepts of equivalent length in series piping and equivalent diameter in pipe loops were explained and illustrated using example problems. The hydraulic pressure gradient and the need for intermediate compressor stations to transport given volumes of gas without exceeding allowable pipeline pressures were also covered. The importance of temperature variation in gas pipelines and how it is taken into account in calculating pipeline pressures were introduced with reference to commercial hydraulic simulation models. The method of calculating the line pack volume in a gas pipeline was also explained. In the next chapter, we will discuss compressor stations, compressor performance, and horsepower requirements.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 136 Friday, April 1, 2005 3:09 PM

136

GAS PIPELINE HYDRAULICS

PROBLEMS 1. A pipeline, NPS 14 with 0.250 in. wall thickness, 40 mi long, transports natural gas (speciﬁc gravity = 0.6 and viscosity = 0.000008 lb/ft-s) at a ﬂow rate of 80 MMSCFD at an inlet temperature of 60°F. Assuming isothermal ﬂow and neglecting elevation changes, calculate the inlet pressure required for a delivery pressure of 800 psig. The base pressure and base temperature are 14.7 psia and 60°F, respectively. Use the Colebrook equation with pipe roughness of 0.0007 in. 2. A 100 mi long natural gas pipeline consists of several injections and deliveries. The pipeline is NPS 18, 0.375 in. wall thickness and has an inlet volume of 150 MMSCFD. At points B (milepost 25) and C (milepost 70), 60 MMSCFD and 50 MMSCFD, respectively, are delivered. At D (milepost 90), gas enters the pipeline at 40 MMSCFD. All streams of gas can be assumed to have a speciﬁc gravity of 0.60 and a viscosity of 7.5 × 10−6 lb/ft-s. The pipe is inter-nally coated such that the absolute roughness is 200 µ in. Assume a constant gas ﬂow temperature of 80°F and base pressure and base temperature of 14.7 psia and 60°F, respectively. Use a constant compressibility factor of 0.88 throughout. Neglect elevation differences along the pipeline. a) Using the modiﬁed Colebrook equation, calculate the pressures along the pipeline at points A, B, C, and D for a minimum delivery pressure of 400 psig at the terminus E. b) What diameter pipe will be required for section DE if the required delivery pressure at E is increased to 600 psig? 3. A natural gas pipeline, 210 km long, consists of an inlet stream at A and deliveries at B and C. The pipeline is DN 400, 10 mm wall thickness. At A, the gas enters at a ﬂow rate of 3.5 Mm3/day. At points B (km 20) and C (km 100), gas is delivered at 0.5 Mm3/day and 1.0 Mm3/day, respectively. At D (km 150), gas enters a branch pipe DF (DN 200, 6 mm wall thickness, 10 km long) at a ﬂow rate of 1.0 Mm3/day. The remaining gas volume of 1.0 Mm3/day is delivered to the pipe terminus E. All streams of gas can be assumed to have a speciﬁc gravity of 0.58 and a viscosity of 0.00012 Poise. The pipe’s absolute roughness is 0.015 mm throughout. Assume a constant gas ﬂow temperature of 15°C and base pressure and base temperature of 101 kPa and 15°C, respectively. Use a pipeline efﬁciency of 0.95 and constant compressibility factor of 0.88 throughout. Neglect elevation differences along the pipeline. a) Using the Panhandle B equation, calculate the pressures along the pipeline at A, B, C, and D for a minimum delivery pressure of 30 Bar at terminus E. b) What is the delivery pressure of gas at the end of the branch DF? c) What pipe diameter is needed for the branch DF if the delivery pressure required at F is 40 Bar? 4. A series piping system consists of 10 mi of NPS 16, 0.250 in. wall thickness, connected to 20 mi of NPS 14, 0.250 in. wall thickness and 10 miles of NPS 12, 0.250 in. wall thickness pipes. Calculate the inlet pressure required at the beginning A for a gas ﬂow rate of 85 MMSCFD. Gas is delivered to the terminus B at

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 137 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

137

a delivery pressure of 600 psig. The gas gravity and viscosity are 0.6 and 0.000008 lb/ft-s, respectively. The gas temperature is assumed constant at 60°F. Use a compressibility factor of 0.85 and the General Flow equation with Colebrook friction factor of 0.015. The base temperature and base pressure are 60°F and 14.7 psia, respectively. 5. A gas pipeline consists of two single pipes with a couple of parallel pipes in the middle. The inlet ﬂow rate is 120 MMSCFD. The ﬁrst pipe segment AB is 10 miles long and consists of NPS 16, 0.250 in. wall thickness pipe. The loop BCE is 20 mi long and consists of NPS 14, 0.250 in. wall thickness pipe. The loop BDE is 15 miles long and consists of NPS 12, 0.250 in. wall thickness pipe. The last segment EF is 18 miles long, NPS 16, 0.250 in. wall thickness pipe. Assuming a gas gravity of 0.6, calculate the outlet pressure at F and the pressures at the beginning and the end of the pipe loops and the ﬂow rates through them. The inlet pressure at A = 1000 psig. The gas ﬂowing temperature = 60°F, base temperature = 60°F, and base pressure = 14.73 psia. The compressibility factor Z = 0.90. Use the AGA fully turbulent equation throughout. 6. A natural gas pipeline is 60 km long. The gas ﬂow rate is 5.0 Mm3/day at 20°C. Calculate the minimum diameter required for an inlet and delivery pressure of 8.5 MPa (absolute) and 5 MPa (absolute), respectively. Use the General Flow equation with the modiﬁed Colebrook-White friction factor. The pipe roughness = 0.020 mm. In order to increase the ﬂow rate through the pipeline, the entire line is looped with an identical-diameter pipeline. Assuming the same delivery pressure, calculate the inlet pressure at the new ﬂow rate of 8 Mm3/day. The gas gravity = 0.60 and viscosity = 0.000119 Poise. The compressibility factor Z = 0.90, base temperature = 15°C, and base pressure = 101 kPa. 7. A natural gas pipeline is 50 mi long and has an inlet pressure of 1200 psig and outlet pressure of 890 psig when transporting 120 MMSCFD. The base pressure and base temperature are 14.7 psia and 60°F, respectively. If the pipe is NPS 16, 0.375 in. wall thickness, calculate the line pack assuming an average gas temperature of 75°F. Use an average compressibility factor of 0.85.

REFERENCES 1. Nayyar, M.L., Piping Handbook, 7th ed., McGraw-Hill, New York, 2000. 2. Baumeister, T., Ed., Standard Handbook for Mechanical Engineers, 7th ed., McGrawHill, New York, 1967. 3. Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003. 4. Westaway, C.R. and Loomis, A.W., Cameron Hydraulic Data, 16th ed., IngersollRand, Montvale, NJ, 1981. 5. Flow of Fluids through Valves, Fittings and Pipe, Crane Company, New York, 1976. 6. McCain, W.D. Jr., The Properties of Petroleum Fluids, Petroleum Publishing Company, Tulsa, OK, 1973. 7. Holman, J.P., Thermodynamics, 2nd ed., McGraw-Hill, New York, 1974.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 138 Friday, April 1, 2005 3:09 PM

138

GAS PIPELINE HYDRAULICS

8. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 9. Engineering Data Book, 10th ed., Gas Processors Suppliers Association, Tulsa, OK, 1994. 10. Pipeline Design for Hydrocarbon Gases and Liquids, American Society of Civil Engineers, New York, 1975. 11. Steady Flow in Gas Pipelines, Contract Report No. 10, July 1965, Pipeline Research Council International, Houston, TX, 1965.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 139 Saturday, April 2, 2005 9:48 AM

CHAPTER

4

Compressor Stations In this chapter we will discuss compressor stations that are needed to transport gas in a pipeline. The optimum locations and pressures at which compressor stations operate will be analyzed in relation to pipeline throughputs, allowable pipe pressures, and pipeline topography. Centrifugal and positive displacement compressors used in natural gas transportation will be compared with reference to their performance characteristics and cost. Typical compressor station design and equipment used will be covered. Isothermal, adiabatic, and polytropic compression processes and horsepower required will be discussed with illustrative examples. The discharge temperature of compressed gas, its impact on pipeline throughput and gas cooling will be explained.

4.1 COMPRESSOR STATION LOCATIONS Compressor stations are installed on gas pipelines to provide the pressure needed to transport gas from one location to another. Due to limitations of pipeline pressures, multiple compressor stations may be needed to transport a given volume through a long-distance pipeline. The locations and pressures at which these compressor stations operate are determined by allowable pipe pressures, power available, and environmental and geotechnical factors. Consider a pipeline that is designed to transport 100 MMSCFD of natural gas from Dover to a power plant at Leeds, 50 miles away. According to methods outlined in Chapter 3, we would calculate the pressure required at Dover to ensure delivery of the gas at a pressure of 500 psig at Leeds. This calculated pressure at Dover may be more or less than the maximum allowable pipe pressures. Suppose the maximum allowable operating pressure (MAOP) of the pipeline is 1200 psig, whereas the pressure at Dover is calculated to be 1050 psig. It is clear that there is no violation of pressures and, hence, a single compressor station at Dover would suffice to deliver gas to Leeds at the required delivery pressure. If the pipeline length were 100 miles

139

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 140 Saturday, April 2, 2005 9:48 AM

140

GAS PIPELINE HYDRAULICS

instead, calculations would show that in order to deliver the same quantity of gas to Leeds at the same terminus pressure, the pressure required at Dover would have to be 1580 psig. Obviously, since this is greater than the MAOP, we would need more than one compressor station. As a first step, we will assume that an intermediate compressor station will be needed in addition to the one at Dover. The next question is where would this compressor station be located? A logical location would be the midpoint between Dover and Leeds. For simplicity, assume the pipeline elevation profile is fairly flat and, therefore, elevation differences can be ignored. Having selected the location of the intermediate pump station at the midpoint, Kent, as shown in Figure 4.1, we will proceed to determine the pressures at the compressor stations. Since the MAOP is limited to 1200 psig, assume that the compressor at Dover discharges at this pressure. Due to friction, the gas pressure drops as it travels through the pipeline from Dover to Kent, as indicated in Figure 4.1. Suppose the gas pressure reaches 900 psig at Kent and is boosted to 1200 psig by the compressor at Kent. Therefore, the compressor station at Kent is said to have a suction pressure of 900 psig and a discharge pressure of 1200 psig. The gas continues to move from Kent to Leeds, starting at 1200 psig at Kent. As the gas reaches Leeds, the pressure may or may not be equal to the desired pressure of 500 psig. Therefore, if the desired terminus pressure at Leeds is maintained, the pressure at the discharge of the Kent compressor stations may have to be adjusted. Alternatively, Kent could discharge at the same 1200 psig, but its location along the pipeline may have to be adjusted. We selected the 900 psig suction pressure at the Kent compressor station quite arbitrarily. It could have been 700 psig or 1000 psig. The actual number depends upon the “compression ratio” desired. The compression ratio is simply the

1580 psig 1200 psig

1200 psig

900 psig

500 psig Flow Dover

Distance

Kent

NPS 16 pipeline 100 mi long Figure 4.1

Gas pipeline with two compressor stations.

Copyright 2005 by Taylor & Francis Group, LLC

Leeds

2785_C004.fm Page 141 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

141

ratio of the compressor discharge pressure to its suction pressure, both pressures being expressed in absolute units. Compression ratio r =

Pd Ps

(4.1)

where the suction and discharge pressures Ps and Pd are in absolute units. In the present case, the compression ratio for Kent is r=

1200 + 14.7 = 1.33 900 + 14.7

In the above calculation, we assumed the base pressure to be 14.7 psia. If we had chosen a suction pressure of 700 psig, the compression ratio would be r=

1200 + 14.7 = 1.7 700 + 14.7

An acceptable compression ratio for centrifugal compressors is about 1.5. A larger number requires more compressor horsepower, whereas a smaller compression ratio means less horsepower required. In gas pipelines, it is desirable to keep the average pipeline pressure as high as possible to reduce compression power. Therefore, if the suction pressure at Kent is allowed to fall to 700 psig or lower, the average pressure in the pipeline would be lower than if we chose 900 psig for the suction pressure at Kent. Obviously, there is a tradeoff between the number of compressor stations, the suction pressure, and the compression horsepower required. We will discuss this in more detail later in this chapter. Going back to the example problem above, we concluded that we may have to adjust the location of the Kent compressor station or adjust its discharge pressure to ensure the 500 psig delivery pressure at Leeds. Alternatively, we could leave the intermediate compressor station at the halfway point and discharge at 1200 psig, eventually delivering gas to Leeds at 600 psig. If calculations show that by discharging out of Kent results in 600 psig at Leeds, we have satisfied the minimum pressure requirement at Leeds. However, there is extra energy associated with the extra 100 psig delivery pressure. If the power plant can use this extra energy, then there is no waste. On the other hand, if the power plant requirement is 500 psig maximum, then some means of regulating the pressure must be present at the delivery point. This would mean that 100 psig would be reduced through a pressure regulator or control valve at Leeds and energy would be wasted. Another option would be to keep the Kent compressor at the midpoint but reduce the discharge pressure to a number that would result in the requisite 500 psig at Leeds. Since pressure drop in gas pipelines is nonlinear, remembering our discussion in Chapter 3, Kent discharge pressure may have to be reduced by less than 100 psig to provide the fixed 500 psig delivery pressure at Leeds. This would mean that Dover would operate at 1200 psig discharge, whereas Kent would discharge at 1150 psig. This violates our premise of keeping the average pressure in a gas pipeline as high as possible. However, this is

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 142 Saturday, April 2, 2005 9:48 AM

142

GAS PIPELINE HYDRAULICS

still a solution, and in order to pick the best option, we must compare two or more alternative approaches, factoring in the total horsepower required as well as the cost involved. Moving the Kent compressor station slightly upstream or downstream would change the suction and discharge pressures and, hence, the horsepower required. From a cost standpoint, the change would not be significant. However, the horsepower variation would result in change in energy cost and, therefore, in annual operating cost. We must therefore take into account the capital cost and annual operating cost in order to come up with the optimum solution. An example will illustrate this method. In Chapter 10, we will cover several different cost scenarios when dealing with pipeline economics. Example 1 A natural gas pipeline, 140 miles long from Dover to Leeds, is constructed of NPS 16, 0.250 in. wall thickness pipe, with an MOP of 1200 psig. The gas specific gravity and viscosity are 0.6 and 8 × 10−6 lb/ft-s, respectively. The pipe roughness can be assumed to be 700 µ in., and the base pressure and base temperature are 14.7 psia and 60°F, respectively. The gas flow rate is 175 MMSCFD at 80°F, and the delivery pressure required at Leeds is 800 psig. Determine the number and locations of compressor stations required, neglecting elevation difference along the pipeline. Assume Z = 0.85. Solution We will use the Colebrook-White equation to calculate the pressure drop. The Reynolds number is calculated from Equation 2.34 as follows: R=

0.0004778 × 175 × 10 6 × 0.6 × 14.7 = 11,437,412 15.5 × 8 × 10 −6 × 520

Relative roughness =

700 × 10 −6 = 4.5161 × 10−5 15.5

Using Colebrook-White Equation 2.39, we get the friction factor 4.516 × 10 −5 2.51 = −2 Log10 + 3.7 11, 437, 412 f f

1

Solving for f by successive iteration, we get f = 0.0107 Using General Flow Equation 2.2, we calculate the pressure required at Dover as, neglecting elevation effects, 520 P12 − 814.72 1 175 × 10 6 = 77.54 14.7 0.6 × 540 × 140 × 0.85 0.0107

Copyright 2005 by Taylor & Francis Group, LLC

0.5

× (15.5)2.5

2785_C004.fm Page 143 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

143

1579 psig

Pressure

MOP = 1200 psig

800 psig

175 MMSCFD NPS 16 pipeline 140 mi long Leeds

Dover

Figure 4.2

Distance

Pipeline with 1200 psig MOP.

Solving for the pressure at Dover, we get P1 = 1594 psia = 1579.3 psig It can be seen from Figure 4.2 that since the MOP is 1200 psig, we cannot discharge at 1579.3 psig at Dover. We will need to reduce the discharge pressure at Dover to 1200 psig and install an additional compressor station at some point between Dover and Leeds, as shown in Figure 4.3.

1579 psig MOP = 1200 psig

1200 psig

800 psig

175 MMSCFD

Dover

Distance

Kent

NPS 16 pipeline 140 mi long Figure 4.3

Dover to Leeds pipeline with one compressor station.

Copyright 2005 by Taylor & Francis Group, LLC

Leeds

2785_C004.fm Page 144 Saturday, April 2, 2005 9:48 AM

144

GAS PIPELINE HYDRAULICS

We will initially assume that the intermediate compressor station will be located at Kent, halfway between Dover and Leeds. For the pipe segment from Dover to Kent, we will calculate the suction pressure at the Kent compressor station as follows. Using General Flow Equation 2.2, 520 1214.72 − P22 1 175 × 10 = 77.54 14.7 0.6 × 540 × 70 × 0.85 0.0107

0.5

× (15.5)2.5

6

Solving for the pressure at Kent (suction pressure), P2 = 733 psia = 718 psig At Kent, if we boost the gas pressure from 718 psig to 1200 psig (MOP), the .7 compression ratio at Kent is 1214 = 1.66. This is a reasonable compression ratio for 733 a centrifugal compressor. Next, we will see if the 1200 psig pressure at Kent will give the desired 800 psig delivery pressure at Leeds. Considering the 70 mi segment from Kent to Leeds, using the General Flow equation we get 520 1214.72 − P22 1 175 × 10 = 77.54 14.7 0.6 × 540 × 70 × 0.85 0.0107

0.5

6

× (15.5)2.5

resulting in a pressure at Leeds of P2 = 733 psia = 718 psig This is less than the 800 psig desired. Hence, we must move the location of the Kent compressor station slightly toward Leeds so that the 800 psig delivery pressure can be achieved. We will calculate the distance L required between Kent and Leeds. To achieve this, using General Flow Equation 2.2 520 1214.72 − 814.72 1 175 × 10 6 = 77.54 0.0107 14.7 0.6 × 540 × L × 0.85

0.5

× (15.5)2.5

Solving for length L, we get 1214.72 − 814.72 1214.72 − 7332 = L 70 L = 60.57 miles Therefore, Kent must be located approximately 61 miles from Leeds. We must now recalculate the suction pressure at the Kent compressor station based on the pipe

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 145 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

145

length of 79.43 (140 − 60.57) miles between Dover and Kent. From this suction pressure, we must also check the compression ratio. Using General Flow Equation 2.2 for the pipe segment between Dover and Kent, we get 520 1214.72 − P22 1 175 × 10 6 = 77.54 14.7 0.6 × 540 × 79.43 × 0.85 0.0107

0.5

× (15.5)2.5

Solving for P2, we get 1214.72 − P2 2 1214.72 − 7332 = 79.43 70 or P2 = 645.49 psia = 630.79 psig Therefore, the suction pressure at Kent = 630.79 psig. The compression ratio at Kent = 1214.7 = 1.88. 645.49 The compression ratio is slightly more than the 1.5 we would like to see. However, for now, we will go ahead with this compression ratio. Figure 4.4 shows the revised configuration with the new location of the Kent compressor station.

1579 psig MOP = 1200 psig

1200 psig

631 psig

800 psig

175 MMSCFD

Dover

Figure 4.4

Distance

61 miles Kent NPS 16 pipeline 140 mi long

Dover to Leeds pipeline with relocated Kent compressor station.

Copyright 2005 by Taylor & Francis Group, LLC

Leeds

2785_C004.fm Page 146 Saturday, April 2, 2005 9:48 AM

146

GAS PIPELINE HYDRAULICS

4.2 HYDRAULIC BALANCE In the preceding discussions, we considered each compressor station operating at the same discharge pressure and also considered the same compression ratio. Recalling the definition of compression ratio from Equation 4.1, we can state that each compressor station operates at the same suction and discharge pressures. If there are no intermediate injections or deliveries along the pipeline, as in Example 1, each compressor station is required to compress the same amount of gas. Therefore, with pressures and flow rates being the same, each compressor station will require the same amount of horsepower. This is known as hydraulic balance. In a long pipeline with multiple compressor stations, in which each compressor station adds the same amount of energy to the gas, we say that this is a hydraulically balanced pipeline. One of the advantages of a hydraulically balanced pipeline is that all compression equipment can be identical, which will reduce inventory of spare parts and minimize maintenance. It is much easier and cheaper to maintain five identical compressor stations of 5000 horsepower each than to maintain two 6000 HP and three 5000 HP compressors. Also, in order to pump the same volume through a pipeline, hydraulically balanced compressor stations will require less total horsepower than if the stations were not located for hydraulic balance. We will now discuss the different processes by which gas is compressed, such as isothermal, adiabatic (isentropic), and polytropic compression. After that, we will outline the method for calculating horsepower for a compressor station in the subsequent sections.

4.3 ISOTHERMAL COMPRESSION The isothermal compression process is one in which the gas pressure and volume compressed vary in a way that the temperature remains constant. Isothermal compression requires the least amount of work compared to other forms of compression. This process is of theoretical interest since, in reality, maintaining the temperature constant in a gas compressor is virtually impossible. Figure 4.5 shows the pressure volume diagram for isothermal compression. Point 1 represents the inlet conditions of pressure (P1), volume (V1), and at temperature (T1). Point 2 represents the final compressed conditions of pressure (P2), volume (V2), and at constant temperature (T1). The relationship between pressure, P, and volume, V, in an isothermal process is as follows: PV = C

(4.2)

P1V1 = P2V2

(4.3)

where C is a constant. Therefore, we can state that

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 147 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

147

2 Iso

P2

the

al

Pressure, P

rm

P1V1 = P2V2 pr

oc

es

sP

V=

P1

con

sta nt

1 V2

V1 Volume, V

Figure 4.5

Isothermal compression.

Considering 1 lb of natural gas compressed isothermally, the work done is calculated as follows:

Wi =

where Wi G T1 P1 P2 Loge

= = = = = =

P 53.28 T1 Loge 2 G P1

(USCS units)

(4.4)

isothermal work done, ft-lb/lb of gas gas gravity, dimensionless suction temperature of gas, °R suction pressure of gas, psia discharge pressure of gas, psia natural logarithm to base e (e = 2.718)

( )

The ratio PP21 is also called the compression ratio. In SI units, the work done in isothermal compression of 1 kg of gas is Wi = where Wi T1 P1 P2

= = = =

P 286.76 T1 Loge 2 G P1

isothermal work done, J/kg of gas suction temperature of gas, K suction pressure of gas, kPa absolute discharge pressure of gas, kPa absolute

Other symbols are as defined earlier.

Copyright 2005 by Taylor & Francis Group, LLC

(SI units)

(4.5)

2785_C004.fm Page 148 Saturday, April 2, 2005 9:48 AM

148

GAS PIPELINE HYDRAULICS

Example 2 Natural gas is compressed isothermally at 60°F from an initial pressure of 500 psig to a pressure of 1000 psig. The gas gravity is 0.6. Calculate the work done in compressing 5 lb of gas. Use 14.7 psia and 60°F for the base pressure and temperature, respectively. Solution Using Equation 4.4, the work done per lb of gas is Wi =

1000 + 14.7 53.28 = 31,343 ft-lb/lb (60 + 460) Loge 0.6 500 + 14.7

The total work done in compressing 5 lb of gas is WT = 31,343 × 5 = 156,715 ft-lb

Example 3 Calculate the work done in compressing 2 kg of gas (gravity = 0.65) isothermally at 20°C from 700 kPa to 2000 kPa. Use 101 kPa and 15°C for the base pressure and temperature, respectively. Solution Using Equation 4.5, the work done in isothermal compression of 1 kg of gas is Wi =

2000 + 101 286.76 (273 + 20) Loge = 124,649 J/kg 0.65 700 + 101

Therefore, the total work done in compressing 2 kg of gas is WT = 124,649 × 2 = 249,298 J

4.4 ADIABATIC COMPRESSION The adiabatic compression process is characterized by zero heat transfer between the gas and the surroundings. The terms adiabatic and isentropic are used synonymously, although isentropic really means “constant entropy.” An adiabatic process that is also frictionless is referred to as isentropic. In an adiabatic compression process, the relationship between pressure and volume is as follows: PV γ = C

Copyright 2005 by Taylor & Francis Group, LLC

(4.6)

2785_C004.fm Page 149 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

149

where C g = ratio of specific heats of gas, Cvp Cp = specific heats of gas at constant pressure Cv = specific heats of gas at constant volume C = a constant, different from the one for isothermal compression in Equation 4.2 g is also known as the adiabatic or isentropic exponent for the gas. It ranges in value from 1.2 to 1.4. Therefore, we can state that γ PV = P2V2γ 1 1

(4.7)

Figure 4.6 shows adiabatic compression similar to the P-V diagram for isothermal compression. Considering 1 lb of natural gas compressed adiabatically, the work done is calculated as follows: γ −1 53.28 γ P2 γ − 1 Wa = T1 G γ − 1 P1

where Wa G T1 g P1 P2

= = = = = =

(USCS units)

adiabatic work done, ft-lb/lb of gas gas gravity, dimensionless suction temperature of gas, °R ratio of specific heats of gas, dimensionless suction pressure of gas, psia discharge pressure of gas, psia

P2

2 ati iab Ad

P1V1γ = P2V2γ

Pressure, P

ro cp

ce

P1

ss

PV γ =c ons

tant 1

V2

V1 Volume, V

Figure 4.6

Adiabatic compression.

Copyright 2005 by Taylor & Francis Group, LLC

(4.8)

2785_C004.fm Page 150 Saturday, April 2, 2005 9:48 AM

150

GAS PIPELINE HYDRAULICS

In SI units, the work done in adiabatic compression of 1 kg of gas is γ −1 286.76 γ P2 γ − 1 Wa = T1 G γ − 1 P1

where Wa T1 P1 P2

= = = =

(SI units)

(4.9)

adiabatic work done, J/kg of gas suction temperature of gas, K suction pressure of gas, kPa absolute discharge pressure of gas, kPa absolute

Other symbols are as defined earlier. Example 4 Natural gas is compressed adiabatically from an initial temperature and pressure of 60°F and 500 psig, respectively, to a final pressure of 1000 psig. The gas gravity is 0.6 and the ratio of specific heat is 1.3. Calculate the work done in compressing 5 lb of gas. Use 14.7 psia and 60°F for the base pressure and temperature, respectively. Solution Using Equation 4.8, the work done in adiabatic compression is

Wa =

0.3 1.3 1014.7 1.3 53.28 (60 + 460) 1 − = 33,931 ft-lb/lb 0.6 0.3 514.7

Therefore, the total work done in compressing 5 lb of gas is WT = 33,931 × 5 = 169,655 ft-lb Example 5 Calculate the work done in compressing 2 kg of gas (gravity = 0.65) adiabatically from an initial temperature of 20°C and pressure of 700 kPa to a final pressure of 2000 kPa. The specific heat ratio of gas is 1.4 and the base pressure and base temperature are 101 kPa and 15°C, respectively. Solution Using Equation 4.9, the work done in adiabatic compression of 1 kg of gas is 0.4 1.4 2000 + 101 1.4 286.76 1 Wa = (20 + 273) − = 143,512 J/kg 0.65 0.4 700 + 101

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 151 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

151

Therefore, the total work done in compressing 2 kg of gas is WT = 143,512 × 2 = 287,024 J

4.5 POLYTROPIC COMPRESSION Polytropic compression is similar to adiabatic compression, but there is no requirement of zero heat transfer as in adiabatic compression. In a polytropic process, the relationship between pressure and volume is as follows: PV n = C

(4.10)

where n = polytropic exponent C = a constant, different from the one for isothermal or adiabatic compression in Equation 4.2 and Equation 4.6 Therefore, we can state that n n PV 1 1 = P2V2

(4.11)

Since polytropic compression is similar to adiabatic compression, we can easily calculate the work done in polytropic compression by substituting n for g in Equation 4.8 and Equation 4.9. Example 6 Natural gas is compressed polytropically from an initial temperature and pressure of 60°F and 500 psig, respectively, to a final pressure of 1000 psig. The gas gravity is 0.6 and the base pressure and base temperature are 14.7 psia and 60°F, respectively. Calculate the work done in compressing 5 lb of gas using a polytropic exponent of 1.5. Solution Polytropic compression is similar to adiabatic compression, and, therefore, the same equation can be used for work done, substituting the polytropic exponent n for the adiabatic exponent g (the ratio of specific heat). Using Equation 4.8, the work done in polytropic compression of 1 lb of gas is

Wp =

0.5 1.5 1014.7 1.5 53.28 (60 + 460) 1 − = 35,168 ft-lb/lb 0.6 0.5 514.7

Therefore, the total work done in compressing 5 lb of gas is WT = 35,168 × 5 = 175,840 ft-lb

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 152 Saturday, April 2, 2005 9:48 AM

152

GAS PIPELINE HYDRAULICS

Example 7 Calculate the work done in compressing 2 kg of gas (gravity = 0.65) polytropically from an initial temperature of 20°C and pressure of 700 kPa to a final pressure of 2000 kPa. Use a polytropic exponent of 1.5. The base pressure and base temperature are 101 kPa and 15°C, respectively. Solution Using Equation 4.9, and substituting the polytropic exponent 1.5 in place of g, the work done in polytropic compression is

Wp =

0.5 1.5 2000 + 101 1.5 286.76 1 (20 + 273) − = 146,996 J/kg 0.65 0.5 700 + 101

Therefore, the total work done in compressing 2 kg of gas is WT = 146,996 × 2 = 293,992 J

4.6 DISCHARGE TEMPERATURE OF COMPRESSED GAS In adiabatic or polytropic compression of natural gas, we can determine the final temperature of the gas knowing the initial temperature and initial and final pressures. Using Equation 4.6 for adiabatic compression and the perfect gas law, by eliminating the volume, V, we can write the following: T2 Z1 P2 T = Z P 1 2 1 where T1 = T2 = Z1 = Z2 =

γ −1 γ

(4.12)

suction temperature of gas, °R discharge temperature of gas, °R gas compressibility factor at suction, dimensionless gas compressibility factor at discharge, dimensionless

Other symbols are as defined earlier. Similarly, for polytropic compression, the discharge temperature can be calculated from the following equation: T2 Z1 P2 T = Z P 1 2 1 where all symbols are as defined earlier.

Copyright 2005 by Taylor & Francis Group, LLC

n −1 n

(4.13)

2785_C004.fm Page 153 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

153

Example 8 Gas is compressed adiabatically (g = 1.4) from 60°F suction temperature and a compression ratio of 2.0. Calculate the discharge temperature, assuming Z1 = 0.99 and Z2 = 0.85. Solution Using Equation 4.12, 0.4 T2 0.99 1.4 60 + 460 = 0.85 (2.0) = 1.4198

T2 = 1.4198 × 520 = 738.3°R = 278.3°F

4.7 HORSEPOWER REQUIRED The amount of energy input to the gas by the compressors is dependent upon the pressure of the gas and flow rate. The horsepower (HP), which represents the energy per unit time, also depends upon the gas pressure and the flow rate. As the flow rate increases, the pressure also increases and, hence, the horsepower needed will also increase. Since energy is defined as work done by a force, we can state the power required in terms of the gas flow rate and the discharge pressure of the compressor station. Suppose the gas flow rate is Q measured in standard ft3 per day (SCFD), and the suction and discharge pressures of the compressor station are Ps and Pd , respectively. The compressor station adds the differential pressure of (Pd – Ps) psia to the gas flowing at Q SCFD. Therefore, the rate at which energy is supplied to the gas is (Pd – Ps) × Q × Const1, where Const1 is a constant depending upon the units employed. This is a very simplistic approach, since the gas properties vary with temperature and pressure. Also, the compressibility factor and the type of gas compression (adiabatic or polytropic) must be taken into account. Therefore, the calculation for HP will be approached from another angle in what follows. The head developed by the compressor is defined as the amount of energy supplied to the gas per unit mass of gas. Therefore, by multiplying the mass flow rate of gas by the compressor head, we can calculate the total energy supplied to the gas. Dividing this by compressor efficiency, we will get the horsepower required to compress the gas. The equation for horsepower can be expressed as follows: HP = where HP = compressor horsepower M = mass flow rate of gas, lb/min

Copyright 2005 by Taylor & Francis Group, LLC

M × ∆H η

(4.14)

2785_C004.fm Page 154 Saturday, April 2, 2005 9:48 AM

154

GAS PIPELINE HYDRAULICS

∆H = compressor head, ft-lb/lb h = compressor efficiency, decimal value Another more commonly used formula for compressor horsepower that takes into account the compressibility of gas is as follows: γ −1 Z1 + Z 2 1 P2 γ γ HP = 0.0857 QT − 1 γ − 1 1 2 ηa P1

where HP g Q T1 P1 P2 Z1 Z2 ha

= = = = = = = = =

(4.15)

compressor horsepower ratio of specific heats of gas, dimensionless gas flow rate, MMSCFD suction temperature of gas, °R suction pressure of gas, psia discharge pressure of gas, psia compressibility of gas at suction conditions, dimensionless compressibility of gas at discharge conditions, dimensionless compressor adiabatic (isentropic) efficiency, decimal value

In SI units, the Power equation is as follows: γ −1 Z1 + Z 2 1 P2 γ γ Power = 4.0639 QT − 1 γ − 1 1 2 ηa P1

where Power g Q T1 P1 P2 Z1 Z2 ha

= = = = = = = = =

(4.16)

compression Power, kW ratio of specific heats of gas, dimensionless gas flow rate, Mm3/day suction temperature of gas, K suction pressure of gas, kPa discharge pressure of gas, kPa compressibility of gas at suction conditions, dimensionless compressibility of gas at discharge conditions, dimensionless compressor adiabatic (isentropic) efficiency, decimal value

The adiabatic efficiency ha generally ranges from 0.75 to 0.85. By considering a mechanical efficiency hm of the compressor driver, we can calculate the brake horsepower (BHP) required to run the compressor as follows: BHP =

Copyright 2005 by Taylor & Francis Group, LLC

HP ηm

(4.17)

2785_C004.fm Page 155 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

155

where HP is the horsepower calculated from the preceding equations, taking into account the adiabatic efficiency ha of the compressor. The mechanical efficiency hm of the driver can range from 0.95 to 0.98. The overall efficiency, ho, is defined as the product of the adiabatic efficiency, ha , and the mechanical efficiency, hm: ho = ha × hm

(4.18)

From the adiabatic compression Equation 4.6, eliminating the volume V, the discharge temperature of the gas is related to the suction temperature and the compression ratio by means of the following equation: T2 P2 T = P 1 1

γ −1 γ

(4.19)

The adiabatic efficiency, ha, can also be defined as the ratio of the adiabatic temperature rise to the actual temperature rise. Thus, if the gas temperature due to compression increases from T1 to T2, the actual temperature rise is (T2 – T1). The theoretical adiabatic temperature rise is obtained from the adiabatic pressure– temperature relationship as follows, considering the gas compressibility factors similar to Equation 4.12: T2 Z1 P2 T = Z P 1 2 1

γ −1 γ

(4.20)

or Z P T2 = T1 1 2 Z 2 P1

γ −1 γ

(4.21)

Therefore, the theoretical adiabatic temperature rise is Z P T1 1 2 Z 2 P1

γ −1 γ

− T1

Therefore, the adiabatic efficiency is

ηa =

T1

( )( ) Z1 Z2

P2 P1

γ −1 γ

− T1

T2 − T1

where T2 is the actual discharge temperature of the gas.

Copyright 2005 by Taylor & Francis Group, LLC

(4.22)

2785_C004.fm Page 156 Saturday, April 2, 2005 9:48 AM

156

GAS PIPELINE HYDRAULICS

Simplifying, we get γ −1 T1 Z1 P2 γ − ηa = 1 Z P T − T 2 1 2 1

(4.23)

For example, if the inlet gas temperature is 80°F and the suction and discharge pressures are 800 psia and 1400 psia, respectively, we can calculate the adiabatic efficiency if the outlet temperature is given as 200°F. Using g = 1.4, and from Equation 4.23, the adiabatic efficiency is, assuming compressibility factors to be equal to 1.0, 1.4 −1 80 + 460 1400 1.4 = 0.7802 − ηa = 1 200 − 80 800

(4.24)

Thus, the adiabatic compression efficiency is 0.7802. Example 9 Calculate the compressor horsepower required for an adiabatic compression of 106 MMSCFD gas with inlet temperature of 68°F and 725 psia pressure. The discharge pressure is 1305 psia. Assume the compressibility factors at suction and discharge conditions to be Z1 = 1.0 and Z2 = 0.85, respectively, and the adiabatic exponent g = 1.4, with the adiabatic efficiency ha = 0.8. If the mechanical efficiency of the compressor driver is 0.95, what BHP is required? Calculate the outlet temperature of the gas. Solution From Equation 4.15, the horsepower required is 0.40 1.40 1 + 0.85 1 1305 1.40 ( ) HP = 0.0857 × 106 68 460 + 1 − 2 0.8 725 = 3550 0.40

Using Equation 4.17, we calculate the driver horsepower required based on a mechanical efficiency of 0.95: BHP required =

3550 = 3737 0.95

The outlet temperature of the gas is found from Equation 4.23 after transposing as follows: T2 = (68 + 460) ×

( 0.185 )( 1305 725 ) 0.8

0.4 1.4

− 1 + (68 + 460) = 786.46°R = 326.46°F

The discharge temperature of the gas is 326.46°F.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 157 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

157

Example 10 Natural gas at 3 Mm3/day and 20°C is compressed isentropically (g = 1.4) from a suction pressure of 5 MPa absolute to a discharge pressure of 9 MPa absolute in a centrifugal compressor with an isentropic efficiency of 0.80. Calculate the compressor power required, assuming the compressibility factors at suction and discharge conditions to be Z1 = 0.95 and Z2 = 0.85, respectively. If the mechanical efficiency of the compressor driver is 0.95, what is the driver power required? Calculate the outlet temperature of the gas. Solution From Equation 4.16, the power required is 0.40 0.95 + 0.885 1 9 1.40 1.40 1 Power = 4.0639 × 3 ( ) 20 273 + − 0.8 5 = 2572 kW 2 0.40

Power = 2572 kW Using Equation 4.17, we calculate the driver power required as follows: Driver power required =

2572 = 2708 kW 0.95

The outlet temperature of the gas is found from Equation 4.23 as follows: 0.4 20 + 273 0.95 9 1.4 × T2 = − 1 + (20 + 273) = 410.94 K = 137.94°C 0.8 0.85 5

4.8 OPTIMUM COMPRESSOR LOCATIONS In the foregoing discussion, we looked at a two-compressor station configuration for gas deliveries from Dover to the Leeds power plant. In this section, we will consider various locations of the intermediate compressor stations on a long-distance gas transmission pipeline to arrive at the optimum locations, taking into account the overall horsepower required. In Section 4.2, we discussed hydraulic balance. The advantage of locating the intermediate compressor station, such that the same amount of energy is added to the gas at each compressor station, was pointed out. In the next example, we will analyze optimum compressor locations by considering both hydraulically balanced and unbalanced compressor station locations. Example 11 A gas transmission pipeline is 240 mi long, NPS 30, 0.500 in. wall thickness, with an origin compressor station at Payson and two intermediate compressor stations

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 158 Saturday, April 2, 2005 9:48 AM

158

GAS PIPELINE HYDRAULICS

MOP = 1400 psig

1400 psig

600 psig

900 MMSCFD

Payson m.p. 0

Williams m.p. 80

Snowflake m.p. 160

Douglas m.p. 240

NPS 30 pipeline 240 mi long Figure 4.7

Gas pipeline with three compressor stations.

tentatively located at Williams (milepost 80) and Snowflake (milepost 160), as shown in Figure 4.7. There are no intermediate flow deliveries or injections, and the inlet flow rate of 900 MMSCFD at Payson equals the delivery flow rate at Douglas. The delivery pressure required at Douglas is 600 psig and the MOP of the pipeline is 1400 psig throughout. Neglect the effects of elevation and assume constant gas flow temperature of 80°F and constant values of transmission factor F = 20 and compressibility factor Z = 0.85 throughout the pipeline. The gas gravity = 0.6, base pressure = 14.7 psia, and base temperature = 60°F. Use a polytropic compression coefficient of 1.38 and a compression efficiency of 0.9. Solution Neglecting the effects of elevation, we could calculate for each of the three segments— Payson to Williams, Williams to Snowflake, and Snowflake to Douglas—the downstream pressure starting with an upstream pressure of 1400 psig. Thus, using the General Flow equation for the Payson to Williams segment, we would calculate the downstream pressure at Williams starting with a pressure of 1400 psig at Payson. This downstream pressure is actually the suction pressure at the Williams compressor station. Next, in a similar fashion, we would calculate the downstream pressure at Snowflake, for the second segment from Williams to Snowflake, based on an upstream pressure of 1400 psig at Williams. This downstream pressure is actually the suction pressure at the Snowflake compressor station. Finally, we would calculate the downstream pressure at Douglas, for the third segment from Snowflake to Douglas, based on an upstream pressure of 1400 psig at Snowflake. This final pressure is the delivery pressure at the Douglas terminus. We have thus calculated the suction pressures at each of the two intermediate compressor stations at Williams and Snowflake and also calculated the final delivery pressure at Douglas. This pressure calculated at Douglas may or may not be equal to the desired delivery pressure of 600 psig, since we performed a forward calculation going from Payson to Douglas. Therefore, since the delivery pressure is usually a desired or contracted value, we will have to adjust the location of the last compressor station at Snowflake to achieve the desired delivery pressure at Douglas.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 159 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

159

Another approach would be to perform a backward calculation starting at Douglas and proceeding toward Payson. In this case, we will start with segment 3 and calculate the location of the Snowflake compressor station that will result in an upstream pressure of 1400 psig at Snowflake. Thus, we locate the Snowflake compressor station that will cause a discharge pressure of 1400 psig at Snowflake and a delivery pressure of 600 psig at Douglas. Having located the Snowflake compressor station, we can now recalculate the suction pressure at Snowflake by considering the pipe segment 2 and using an upstream pressure of 1400 psig at Williams. We will not have to repeat calculations for segment 1, since the location of Williams has not changed and, therefore, the suction pressure at Williams will remain the same as the previously calculated value. We have thus been able to determine the pressures along the pipeline with the given threecompressor-station configuration such that the desired delivery pressure at Douglas has been achieved and each compressor station discharges at an MOP value of 1400 psig. But are these the optimum locations of the intermediate compressor stations Williams and Snowflake? And are all compressor stations in hydraulic balance? We can state that these compressor stations are optimized and are in hydraulic balance only if each compressor station operates at the same compression ratio and, therefore, adds the same amount of horsepower to the gas at each compressor station. The locations of Williams and Snowflake may not result in the same suction pressures even though the discharge pressures are the same. Therefore, chances are that Williams might be operating at a lower compression ratio than Snowflake or Payson, or vice versa, which will not result in hydraulic balance. However, if the compression ratios are close enough that the required compressor sizes are the same, we could still be in hydraulic balance and the stations could be at optimum locations. Next, perform the actual calculations and determine how much tweaking of the compressor station locations is required to optimize these stations. First, we will perform the backward calculations for segment 3, starting with a downstream pressure of 600 psig at Douglas and an upstream pressure of 1400 psig at Snowflake. With these constraints, we will calculate the pipe length, L, miles between Snowflake and Douglas. Using General Flow Equation 2.4, neglecting elevations, 520 1414.72 − 614.772 900 × 10 = 38.77 × 20.0 14.7 0.6 × 540 × L × 0.85 6

0.5

(29)2.5

Solving for pipe length, we get L = 112.31 mi Therefore, in order to discharge at 1400 psig at Snowflake and deliver gas at 600 psig at Douglas, the Snowflake compressor station will be located at a distance of 112.31 mi upstream of Douglas—or at milepost (240 – 112.31) = 127.69 measured from Payson. Next, keeping the location of the Williams compressor at milepost 80, we will calculate the downstream pressure at Snowflake for pipe segment 2 starting at 1400 psig at Williams. This calculated pressure will be the suction pressure of the Snowflake compressor station.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 160 Saturday, April 2, 2005 9:48 AM

160

GAS PIPELINE HYDRAULICS

Using General Flow Equation 2.4, neglecting elevations, 1414.72 − P2 2 520 900 × 10 = 38.77 × 20.0 14.7 0.6 × 540 × 47.69 × 0.85

0.5

6

(29)2.5

where the pipeline segment length between Williams and Snowflake was calculated as 127.69 – 80 = 47.69 mi Solving for suction pressure at Snowflake, we get P2 = 1145.42 psia = 1130.72 psig Therefore, the compression ratio at Snowflake is

1414.7 = 1.24. 1145.42

Next, for pipe segment 1 between Payson and Williams, we will calculate the downstream pressure at Williams, starting at 1400 psig at Payson. This calculated pressure will be the suction pressure of the Williams compressor station. Using General Flow Equation 2.4, neglecting elevations, 1414.72 − P2 2 520 900 × 10 = 38.77 × 20.0 14.7 0.6 × 540 × 80 × 0.85

0.5

6

(29)2.5

Solving for suction pressure at Williams, we get P2 = 919.20 psia = 904.5 psig Therefore, the compression ratio at Williams =

1414.7 = 1.54 919.2

Therefore, from the foregoing calculations, the compressor station at Williams requires a compression ratio r = 1.54, whereas the compressor station at Snowflake requires a compression ratio r = 1.24. Obviously, this is not a hydraulically balanced compressor station system. Further, we do not know what the suction pressure is at the Payson compressor station. If we assume that Payson receives gas at approximately the same suction pressure as Williams (905 psig), both the Payson and Williams compressor stations will have the same compression ratio of 1.54. In this case, the Snowflake compressor station will be the odd one, operating at a compression ratio of 1.24. How do we balance these compressor stations? One way would be to obtain the same compression ratios for all three compressor stations by simply relocating the Snowflake compressor station toward Douglas such that its suction pressure will drop from 1131 psig to 905 psig while keeping the discharge at Snowflake at 1400 psig. This will then ensure that all three compressor stations will be operating at the following suction and discharge pressures and compression ratios: Suction pressure Ps = 904.5 psig Discharge pressure Pd = 1400 psig Compression ratio r =

Copyright 2005 by Taylor & Francis Group, LLC

1400 + 14.7 = 1.54 904.5 + 14.7

2785_C004.fm Page 161 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

161

MOP = 1400 psig

1400 psig

905 psig ∆P 600 psig 900 MMSCFD

Payson m.p. 0

Williams m.p. 80

Snowflake m.p. 160

Douglas m.p. 240

NPS 30 pipeline 240 mi long Figure 4.8

Pressure regulation at Douglas.

However, because the Snowflake compressor station is now located closer to Douglas than before (127.69), the discharge pressure of 1400 psig at Snowflake will result in a higher delivery pressure at Douglas than the required 600 psig, as shown in Figure 4.8. If the additional pressure at Douglas can be tolerated by the customer, then there will be no problem. But if the customer requires no more than 600 psig, we have to reduce the delivery pressure to 600 psig by installing a pressure regulator at Douglas, as shown in Figure 4.8. Therefore, by balancing the compressor station locations, we have also created a problem of getting rid of the extra pressure at the delivery point. Pressure regulation means wasted horsepower. The advantage of the balanced compressor stations vs. the negative aspect of the pressure regulation must be factored into the decision process. To illustrate this pressure regulation scenario, we will now determine the revised location of the Snowflake compressor station for hydraulic balance. We will calculate the length of pipe segment 2 by assuming 1400 psig discharge pressure at Williams and a suction pressure of 904.5 psig at Snowflake. Using General Flow Equation 2.4, neglecting elevations, 520 1414.72 − 919.222 900 × 10 6 = 38.77 × 20.0 14.7 0.6 × 540 × L × 0.85

0.5

(29)2.5

Solving for pipe length for segment 2, we get L = 80 mi Therefore, the Snowflake compressor station should be located at a distance of 80 mi from Williams or at milepost 160. We could have arrived at this without the above calculations, since elevations are neglected and the Payson to Williams pressure

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 162 Saturday, April 2, 2005 9:48 AM

162

GAS PIPELINE HYDRAULICS

profile will be the same as the pressure profile from Williams to Snowflake. With the Snowflake compressor station located at milepost 160, and discharging at 1400 psig, we conclude that the delivery pressure at Douglas will also be 904.5 psig, since all three pipe segments are hydraulically the same. We see that the delivery pressure at Douglas is approximately 305 psig more than the desired pressure. As indicated earlier, a pressure regulator will be required at Douglas to reduce the delivery pressure to 600 psig. We can compare the hydraulically balanced scenario with the previously calculated case where Payson and Williams operate at a compression ratio of 1.54 and Snowflake operates at lower compression ratio of 1.24. By applying approximate cost per installed horsepower, we can compare these two cases. First, using Equation 4.15, calculate the horsepower required at each compressor station, assuming polytropic compression and a compression ratio of 1.54 for a balanced compressor station: 0.38 1 + 0.85 1 1.38 (80 + 460) HP = 0.0857 × 900 × (1.54) 1.38 − 1 = 19,627 2 0.9 0.38

Therefore, the total horsepower required in the hydraulically balanced case is Total HP = 3 × 19,627 = 58,881 At a cost of $2000 per installed HP, Total HP cost = $2000 × 58,881 = $117.76 million In the hydraulically unbalanced case, the Payson and Williams compressor stations will operate at a compression ratio of 1.54 each, whereas the Snowflake compressor station will require a compression ratio of 1.24. Using Equation 4.15, the horsepower required at the Snowflake compressor station is 0.38 1 + 0.85 1 1.38 1.38 − 1 = 9487 ( ) ( . ) HP = 0.0857 × 900 × 80 460 + 1 24 2 0.9 0.38

Therefore, the total horsepower required in the hydraulically unbalanced case is Total HP = (2 × 19,627) + 9487 = 48,741 At a cost of $2000 per installed HP, Total HP cost = $2000 × 48,741 = $97.48 million The hydraulically balanced case requires 10,140 (58,881 − 48,741) HP more and will cost approximately $20.28 ($117.76 − $97.48) million more. In addition to the extra HP cost, the hydraulically balanced case will require a pressure regulator that will waste energy and result in extra equipment cost. Therefore, the advantages of using identical components, by reducing spare parts and inventory in the hydraulically balanced case, must be weighed against the additional cost. It may not be worth spending the extra $20 million to obtain this benefit. The preferred solution in this case is for the Payson and Williams compressor stations to be identical (compression ratio = 1.54) and the Snowflake compressor station to be a smaller one (compression ratio = 1.24), requiring the lower compression ratio and horsepower, to provide the required 600 psig delivery pressure at Douglas.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 163 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

163

1296 psia

1080 psia

900 psia

Q

Flow = Q Compression ratio = 1.2 Compression ratio = 1.2 Overall compression ratio = 1.44 Figure 4.9

Compressors in series.

4.9 COMPRESSORS IN SERIES AND PARALLEL When compressors operate in series, each unit compresses the same amount of gas but at different compression ratios, such that the overall pressure increase of the gas is achieved in stages, as shown in Figure 4.9. It can be seen from Figure 4.9 that the first compressor compresses gas from a suction pressure of 900 psia to 1080 psia at a compression ratio of 1.2. The second compressor takes the same volume and compresses it from 1080 psia to a discharge pressure of 1080 × 1.2 = 1296 psia. Thus, the overall compression ratio of the two identical compressors in series is 1296/900 = 1.44. We have thus achieved the increase in pressure in two stages. At the end of each compression cycle, the gas temperature rises to some value calculated in accordance with Equation 4.19. Therefore, with multiple stages of compression, unless the gas is cooled between stages, the final gas temperature may be too high. High gas temperatures are not desirable, since the throughput capability of a gas pipeline decreases with gas flow temperature. Therefore, with compressors in series, the gas is cooled to the original suction temperature between each stage of compression, such that the final temperature at the end of all compressors in series is not exceedingly high. Suppose the calculated discharge temperature of a compressor is 232°F, starting at a 70°F suction temperature and with a compression ratio of 1.4. If two of these compressors were in series and there were no cooling between compressions, the final gas temperature would reach approximately (232 + 460)(232 + 460) = 903.5°R = 443.5°F 70 + 460 This is too high a temperature for pipeline transportation. On the other hand, if we cool the gas back to 70°F before compressing it through the second compressor, the final temperature of the gas coming out of the second compressor will be 232°F, approximately. We will discuss compressors in series in more detail in the subsequent section.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 164 Saturday, April 2, 2005 9:48 AM

164

GAS PIPELINE HYDRAULICS

1260 psia

300 MMSCFD

900 MMSCFD

300 MMSCFD

900 psia 900 MMSCFD

300 MMSCFD

Compression ratio = 1.4 Figure 4.10

Compressors in parallel.

Compressors are installed in parallel so that large volumes necessary can be provided by multiple compressors, each producing the same compression ratio. Three identical compressors with compression ratio of 1.4 can be used to provide a 900 MMSCFD gas flow from a suction pressure of 900 psia. In this example, each compressor will compress 300 MMSCFD from 900 psia to a discharge pressure of P2 = 900 × 1.4 = 1260 psia This is illustrated schematically in Figure 4.10. Unlike compressors in series, the discharge temperature of the gas coming out of the parallel bank of compressors will not be high, since the gas does not undergo multiple compression ratios. The gas temperature on the discharge side of each parallel compressor will be the same as that of a single compressor with the same compression ratio. Therefore, three parallel compressors, each compressing the same volume of gas at a compression ratio of 1.4, will have a final discharge temperature of 232°F, starting from a suction temperature of 70°F. Gas cooling is required at these temperatures in order to achieve efficient gas transportation and also operate at temperatures not exceeding the limits of the pipe coating material. Generally, pipe coating requires the gas temperature not to exceed 140 to 150°F. The compression ratio was defined earlier as the ratio of the discharge pressure to the suction pressure. The higher the compression ratio, the higher will be the gas discharge temperature, in accordance with Equation 4.19. Consider a suction temperature of 80°F and the suction and discharge pressures of 900 psia and 1400 psia, respectively. The compression ratio is 1400/900 = 1.56. Using Equation 4.19, the discharge temperature will be T2 1400 80 + 460 = 900

1.3−1 1.3

T2 = 598.36°R or 138.36°F If the compression ratio is increased to 2.0, the discharge temperature will become 173.67°F. It can be seen that the discharge temperature of the gas increases

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 165 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

165

considerably with the compression ratio. Since the throughput capacity of a gas pipeline decreases with gas temperature, we must find a way to reduce the high gas temperature resulting from gas compression. In previous chapters, we solved many problems with a constant gas inlet temperature of 80°F. In order to maintain throughput, cooling should be provided on the discharge of the compressor. We prefer centrifugal compressors used in gas pipeline applications to have a compression ratio of 1.5 to 2.0; there may be instances in which higher compression ratios are required due to lower gas receipt pressures and higher pipeline discharge pressures to enable a given volume of gas to be transported through a pipeline. Reciprocating compressors are designed to provide higher compression ratios. However, manufacturers limit maximum compression ratios to a range of 4 to 6. This is due to high forces that are exerted on the compressor components, which cause expensive material requirements as well as complicated safety needs. Suppose a compressor is required to provide gas at 1500 psia from gas that is received at 200 psia. This requires an overall compression ratio of 7.5. Since this is beyond the acceptable range of compression ratios, we will have to provide this compression in stages. If we provide the necessary pressure by using two compressors in series, each compressor will require to be at a compression ratio of 7.5 , or approximately 2.74. The first compressor raises the pressure from 200 psia to 200 × 2.74 = 548 psia. The second compressor will then boost the gas pressure from 548 psia to 548 × 2.74 = 1500 psia, approximately. In general, if n compressors are installed in series to achieve the required compression ratio r, we can state that each compressor will operate at a compression ratio of 1

r = (rt ) n

(4.25)

where r = compression ratio, dimensionless rt = overall compression ratio, dimensionless n = number of compressors in series It has been found that by providing the overall compression ratio by means of identical compressors in series, power requirements will be minimized. Thus, in the preceding example, we assumed that two identical compressors in series, each providing a compression ratio of 2.74 resulting in an overall compression ratio of 7.5, will be a better option than if we had a compressor with a compression ratio of 3.0 in series with another compressor with a compression ratio of 2.5. To illustrate this further, if an overall compression ratio of 20 were required and we were to use three compressors in series, the most economical option would be to use identical compres1 sors, each with a compression ratio of (20) 3 = 2.71. Example 12 A compressor station with multiple compressors in series is to provide a gas discharge pressure of 1500 psia. The gas inlet pressure and temperature are 100 psia and 80°F, respectively. How many compressors in series will be required if the discharge temperature is limited to 250°F? The ratio of specific heats g = 1.4.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 166 Saturday, April 2, 2005 9:48 AM

166

GAS PIPELINE HYDRAULICS

Solution The overall compression ratio is r=

1500 = 15.0 100

Since this is more than the maximum recommended compression ratio of 4 to 6, we need two or more compressors in series. Initially, consider two compressors in series. The compression ratio for each compressor is 1

r = (15) 2 = 3.873 This is acceptable, but the discharge temperature needs to be checked. From Equation 4.19, the discharge temperature for the first compressor is 0.4

T2 = (80 + 460)(3.8) 1.4 = 790.76°R or 330.76°F. This temperature is higher than the 250°F allowable. Therefore, we will need to consider three stages of compression. Using three compressors in series, the compression ratio is 1

r = (15) 3 = 2.466 Therefore, the discharge temperature for each compressor is 0.4

T2 = (80 + 460)(2.466) 1.4 = 698.86°R or 239°F. Since this is less than 250°F allowed, the three compressors in series are the choice. However, the gas must be cooled to the initial inlet temperature of 80°F between each compressor to limit discharge temperatures to 239°F.

4.10 TYPES OF COMPRESSORS—CENTRIFUGAL AND POSITIVE DISPLACEMENT Compressors used in natural gas transportation systems are either positive displacement (PD) type or centrifugal (CF) type. Positive displacement compressors generate the pressure required by trapping a certain volume of gas within the compressor and increasing the pressure by reduction of volume. The high-pressure gas is then released through the discharge valve into the pipeline. Piston-operated reciprocating compressors fall within the category of positive displacement compressors. These compressors

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 167 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

167

have a fixed volume and are able to produce high compression ratios. Centrifugal compressors, on the other hand, develop the pressure required by the centrifugal force due to rotation of the compressor wheel that translates the kinetic energy into pressure energy of the gas. Centrifugal compressors are more commonly used in gas transmission systems due to their flexibility. Centrifugal compressors have lower installed cost and lower maintenance expenses. They can handle larger volumes within a small area compared to positive displacement compressors. They also operate at high speeds and are of balanced construction. However, centrifugal compressors have less efficiency than positive displacement compressors. Positive displacement compressors have flexibility in pressure range, have higher efficiency, and can deliver compressed gas at a wide range of pressures. They are also not very sensitive to the composition of the gas. Positive displacement compressors have pressure ranges up to 30,000 psi and range from very low HP to more than 20,000 HP per unit. Positive displacement compressors can be single stage or multistage, depending upon the compression ratio required. The compression ratio per stage for positive displacement compressors is limited to 4.0, because higher ratios cause higher discharge pressures, which affect the valve life of positive displacement compressors. Heat exchangers are used between stages of compression so that the compressed heated gas is cooled to the original suction temperature before being compressed in the next stage. The HP required in a positive displacement compressor is usually estimated from charts provided by the compressor manufacturer. The following equation can be used for large slow-speed compressors with compression ratios greater than 2.5 and for gas specific gravity of 0.65. BHP = 22rNQF where BHP r N Q F

= = = = = = = =

(4.26)

brake horsepower compression ratio per stage number of stages gas flow rate, MMSCFD at suction temperature and 14.4 psia factor that depends on the number of compression stages 1.0 for single-stage compression 1.08 for two-stage compression 1.10 for three-stage compression

In Equation 4.26, the constant 22 is changed to 20 when gas gravity is between 0.8 and 1.0. Also, for compression ratios between 1.5 and 2.0, the constant 22 is replaced with a number between 16 and 18. Example 13 Calculate the BHP required to compress 5 MMSCFD gas at 14.4 psia and 70°F, with an overall compression ratio of 7, considering two-stage compression.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 168 Saturday, April 2, 2005 9:48 AM

168

GAS PIPELINE HYDRAULICS

Solution Considering two identical stages, the compression ratio per stage = 7.0 = 2.65. Using Equation 4.26, we get BHP = 22 × 2.65 × 2 × 5 × 1.08 = 629.64 Centrifugal compressors can be a single-wheel or single-stage compressor or multiwheel or multistage compressor. Single-stage centrifugal compressors have a volume range of 100 to 150,000 ft3/min at actual conditions (ACFM). Multistage centrifugal compressors handle a volume range of 500 to 200,000 ACFM. The operational speeds of centrifugal compressors range from 3000 to 20,000 r/min. The upper limit of speed will be limited by the wheel tip speed and stresses induced in the impeller. Advances in technology have produced compressor wheels operating at speeds in excess of 30,000 r/min. Centrifugal compressors are driven by electric motors, steam turbines, or gas turbines. Sometimes speed increasers are used to increase the speeds necessary to generate the pressure.

4.11 COMPRESSOR PERFORMANCE CURVES

e lin

e

The performance curve of a centrifugal compressor that can be driven at varying speeds typically shows a graphic plot of the inlet flow rate in actual cubic feet per minute (ACFM) against the head or pressure generated at various percentages of the design speed. Figure 4.11 shows a typical centrifugal compressor performance curve or performance map.

Head, ft-lb/lb

surg

100% speed 90% 80% 70% 60%

Flow rate, ACFM Figure 4.11

Typical centrifugal compressor performance curve.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 169 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

169

The limiting curve on the left-hand side is known as the surge line, and the corresponding curve on the right side is known as the stone wall limit. Generally, the performance of a centrifugal compressor follows the “affinity laws.” According to the affinity laws, as the rotational speed of the centrifugal compressor is changed, the inlet flow and head vary as the speed and the square of the speed, respectively, as indicated in the following equations. For compressor speed change, Q2 N 2 = Q1 N1 H2 N 2 = H1 N1

(4.27) 2

(4.28)

where Q1, Q2 = initial and final flow rates H1, H2 = initial and final heads N1, N2 = initial and final compressor speeds In addition, the horsepower for compression varies as the cube of the speed change as follows: HP2 N 2 = HP1 N1

3

(4.29)

An example problem of using the affinity laws to predict the performance of a centrifugal compressor is illustrated next. Example 14 The compressor head and volume flow rate for a centrifugal compressor at 18,000 rpm are as follows: Flow Rate, Q

Head, H

ACFM 360 450 500 600 700 730

ft-lb/lb 10,800 10,200 9700 8200 5700 4900

Using the affinity laws, determine the performance of this compressor at a speed of 20,000 rpm.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 170 Saturday, April 2, 2005 9:48 AM

170

GAS PIPELINE HYDRAULICS

Solution The ratio of speed is 20000 = 1.11 18000 The multiplier for the flow rate is 1.11 and the multiplier for the head is (1.11)2 or 1.232. Using the affinity laws, the performance of the centrifugal compressor at 20,000 rpm is as follows: Flow Rate, Q

Head, H

ACFM 399.6 499.5 555.0 666.0 777.0 810.0

ft-lb/lb 13,306 12,566 11,950 10,102 7,022 6,037

Next, we will explore how the head developed by a centrifugal compressor is calculated from the suction and discharge pressures, the compressibility factor, and the polytropic or adiabatic exponent. The calculation for the actual or inlet flow rate (ACFM) from the standard flow rate will also be illustrated. Finally, knowing the maximum head that can be generated per stage, the number of stages needed will be calculated. Suppose a centrifugal compressor is used to raise the gas pressure from 800 psia to 1440 psia starting at a suction temperature of 70°F and gas flow rate of 80 MMSCFD. The average compressibility factor from the suction to the discharge side is 0.95. The compressibility factor at the inlet is assumed to be 1.0, and the polytropic exponent is 1.3. Gas gravity is 0.6. The head generated by the compressor is calculated as 0.3 1.3 1440 1.3 53.28 H= × 0.95 × (70 + 460) 1 − = 28,146 ft-lb/lb 0.6 0.3 800

The actual flow rate at inlet conditions is calculated using the gas law as Qact =

80 × 14.7 × 1.0 70 + 460 10 6 = 1040.5 ft3/min (ACFM) × × 800 60 + 460 24 × 60

If this particular compressor, according to vendor data, can produce a maximum head per stage of 10,000 ft-lb/lb, the number of stages required to produce the required head is n=

Copyright 2005 by Taylor & Francis Group, LLC

28146 = 3 , approximately. 10000

2785_C004.fm Page 171 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

171

Next, suppose that this compressor has a maximum design speed of 16,000 rpm. The actual operating speed necessary for the three-stage compressor is, according to the affinity laws, 28,146 = 15, 498 3 × 10, 000

N act = 16, 000

Therefore, in order to generate 28,146 ft-lb/lb of head at a gas flow rate of 1040.5 ACFM, this three-stage compressor must run at a speed of 15,498 rpm.

4.12 COMPRESSOR STATION PIPING LOSSES As the gas enters the suction side of the compressor, it flows through a complex piping system within the compressor station. Similarly, the compressed gas leaving the compressor traverses the compressor station discharge piping system that consists of valves and fittings before entering the main pipeline on its way to the next compressor station or delivery terminus. This is illustrated in Figure 4.12. It can be seen from Figure 4.12 that at the compressor station boundary A on the suction side, the gas pressure is P1. This pressure drops to a value Ps at the compressor suction, as the gas flows through the suction piping from A to B. This suction piping, consisting of valves, fittings, filters, and meters, causes a pressure drop of ∆Ps to occur. Therefore, the actual suction pressure at the compressor is Ps = P1 – ∆Ps

(4.30)

where Ps = compressor suction pressure, psia P1 = compressor station suction pressure, psia ∆Ps = pressure loss in compressor station suction piping, psi At the compressor, the gas pressure is raised from Ps to Pd through a compression ratio r as follows: r=

P1

Pd Ps

(4.31)

A

B

Ps

Pd Flow

Figure 4.12

Compressor station suction and discharge piping.

Copyright 2005 by Taylor & Francis Group, LLC

P2

2785_C004.fm Page 172 Saturday, April 2, 2005 9:48 AM

172

GAS PIPELINE HYDRAULICS

where r = compression ratio, dimensionless Pd = compressor discharge pressure, psia The compressed gas then flows through the station discharge piping and loses pressure until it reaches the station discharge valve at the boundary D of the compressor station. If the station discharge pressure is P2, we can write P2 = Pd – ∆Pd

(4.32)

where P2 = compressor station discharge pressure, psia ∆Pd = pressure loss in compressor station discharge piping, psi Generally, the values of ∆Ps and ∆Pd range from 5 to 15 psi. Example 15 A compressor station on a gas transmission pipeline has the following pressures at the station boundaries. The station suction pressure = 850 psia, and the station discharge pressure = 1430 psia. The pressure losses in the suction piping and discharge piping are 5 psi and 10 psi, respectively. Calculate the compression ratio of this compressor station. Solution From Equation 4.30, the compressor suction pressure is Ps = 850 – 5 = 845 psia Similarly, the compressor discharge pressure is Pd = 1430 + 10 = 1440 psia Therefore, the compression ratio is

r=

1440 = 1.70 845

4.13 COMPRESSOR STATION SCHEMATIC A typical compressor station schematic showing the arrangement of the valves, piping, and the compressor itself is shown in Figure 4.13.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 173 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

173

To next compressor station

From previous compressor station A

B MOV

MOV

Pd

Ps

Compressors Figure 4.13

Compressor station schematic.

4.14 SUMMARY We discussed compressing a gas to generate the pressure needed to transport the gas from one point to another along a pipeline. An important parameter known as the compression ratio determines the horsepower required to compress a certain volume of gas and also influences the discharge temperature of the gas exiting the compressor. In a long-distance gas transmission pipeline, the method of locating intermediate compressor stations and minimizing energy lost was discussed. Hydraulically balanced and optimized compressor station locations were also discussed. Calculation of isothermal, adiabatic, and polytropic compression processes was explained and illustrated with sample problems. The HP required for a given compression ratio and calculation of the gas discharge temperature were explained. The different types of compressors, such as positive displacement and centrifugal, were explained, along with their advantages and disadvantages. The need for configuring compressors in series and parallel was explored. The centrifugal compressor performance curve was discussed, and the effect of rotational speed on the flow rate and head using the affinity laws was illustrated with examples. Finally, the impact of the compressor station yard piping pressure drops and how they affect the compression ratio and horsepower were discussed.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 174 Saturday, April 2, 2005 9:48 AM

174

GAS PIPELINE HYDRAULICS

PROBLEMS 1. A natural gas pipeline 120 mi long from Dover to Leeds is constructed of NPS 14 and .250 in. wall thickness pipe, with an MOP of 1400 psig. The gas specific gravity and viscosity are 0.6 and 8 × 10−6 lb/ft-s, respectively. The pipe roughness can be assumed to be 600 µin., and the base pressure and base temperature are 14.7 psia and 60°F, respectively. The gas flow rate is 120 MMSCFD at 70°F, and the delivery pressure required at Leeds is 700 psig. Determine the number and locations of compressor stations required, neglecting elevation difference along the pipeline. Assume Z = 0.90. 2. Calculate the compressor horsepower required for an adiabatic compression of 80 MMSCFD gas with inlet temperature of 70°F and 800 psia pressure. The discharge pressure is 1400 psia. Assume the compressibility factors at suction and discharge conditions to be Z1 = 0.95 and Z2 = 0.88, respectively, and the adiabatic exponent g = 1.3, with the adiabatic efficiency ha = 0.82. If the mechanical efficiency of the compressor driver is 0.94, what BHP is required? Also, calculate the outlet temperature of the gas. 3. Natural gas at 4 Mm3/day and 24°C is compressed isentropically (g = 1.3) from a suction pressure of 6.2 MPa to a discharge pressure of 9.4 MPa in a centrifugal compressor with an isentropic efficiency of 0.82. Calculate the compressor power required, assuming the compressibility factors at suction and discharge conditions to be Z1 = 0.96 and Z2 = 0.87, respectively. If the mechanical efficiency of the compressor driver is 0.94, what is the driver power required? Also, calculate the outlet temperature of the gas. 4. Determine the horsepower required to compress natural gas in a pipeline at a flow rate of 350 MMSCFD and at a compression ratio of 1.6, discharging at 1400 psig pressure. The suction temperature is 80°F. The base temperature and base pressure are 60°F and 14.7 psia, respectively. The gas specific gravity is 0.65, and the compression efficiency is 0.85. What is the discharge temperature of the gas, assuming a polytropic compression exponent of 1.39? The compressibility factor Z = 1.0 at suction conditions and Z = 0.86 at discharge conditions. 5. Determine the horsepower required to compress natural gas in a pipeline at a flow rate of 500 MMSCFD and at a compression ratio of 1.4, discharging at 1200 psia pressure. The suction temperature is 70°F. The base temperature and base pressure are 60°F and 14.7 psia, respectively. The gas specific gravity is 0.6, and assume a compression efficiency of 0.9. What is the discharge temperature of the gas, assuming the polytropic compression coefficient of 1.38? Z = 1.0 at suction conditions and Z = 0.86 at discharge conditions. 6. A gas transmission pipeline is 220 mi long, NPS 24, 0.500 in. wall thickness, and runs from Taylor to Jenks. There is an origin compressor station at Taylor and two intermediate compressor stations at Trent (milepost 70) and Beaver (milepost 130). There are no intermediate flow deliveries or injections, and the inlet flow rate of 500 MMSCFD at Taylor equals the delivery flow rate at Jenks. The delivery pressure required at Jenks is 700 psig, and the MOP of the pipeline is 1440 psig throughout. Neglect the effects of elevation, and assume a constant gas flow temperature of 70°F and constant values of transmission factor F = 20 and

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 175 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

175

compressibility factor Z = 0.87 throughout the pipeline. The gas gravity = 0.6, base pressure = 14.7 psia, and base temperature = 60°F. Use a polytropic compression coefficient of 1.4 and a compression efficiency of 0.85. Determine the best locations for the intermediate compressor stations at Trent and Beaver. If the flow rate drops to 350 MMSCFD, will a single intermediate compressor station be sufficient at the reduced flow rate? 7. A compressor station with multiple compressors in series is to provide a gas discharge pressure of 1400 psia. The gas inlet pressure and temperature are 200 psia and 70°F, respectively. How many compressors in series will be required if the discharge temperature is limited to 200°F? The ratio of specific heats g = 1.3. 8. Calculate the BHP required to compress 8 MMSCFD gas at 14.4 psia and 80°F, with an overall compression ratio of 8, considering two-stage compression. 9. The compressor head and volume flow rates for a centrifugal compressor at 15,000 rpm are as follows: Q–ACFM H–ft-lb/lb

720 5400

900 5100

1000 4900

1200 4100

1400 2800

Using the affinity laws, determine the performance of this compressor at a speed of 12,000 rpm. 10. A compressor station on a gas transmission pipeline has the following pressures at the station boundaries: station suction pressure = 840 psig and station discharge pressure = 1410 psig. The pressure losses in the suction piping and discharge piping are 6 psi and 12 psi, respectively. Calculate the compression ratio of this compressor station.

REFERENCES 1. Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003. 2. Westaway, C.R. and Loomis, A.W., Cameron Hydraulic Data, 16th ed., Ingersoll-Rand, Montvale, NJ, 1981. 3. Flow of Fluids through Valves, Fittings and Pipe, Crane Company, New York, 1976. 4. McCain, W.D. Jr., The Properties of Petroleum Fluids, Petroleum Publishing Company, Tulsa, OK, 1973. 5. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 6. Engineering Data Book, 10th ed., Gas Processors Suppliers Association, Tulsa, OK, 1994.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 176 Saturday, April 2, 2005 9:48 AM

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 177 Friday, April 1, 2005 3:30 PM

CHAPTER

5

Pipe Loops versus Compression

In this chapter we will explore the need for installing pipe loops in order to increase the throughput in a gas pipeline. Looping will be compared to another means of increasing pipeline capacity, such as installing compressor stations. The advantages and disadvantages of looping pipes vs. adding compressor stations will be discussed.

5.1 PURPOSE OF A PIPE LOOP The purpose of a pipe loop that is installed in a segment of a pipeline is to essentially reduce the amount of pressure drop in that section of pipe. By doing so, the overall pressure drop in the pipeline will be reduced. This, in turn, will result in an increased pipeline flow rate at the same inlet pressure. Alternatively, if the flow rate is kept constant, reduction in total pressure required will cause a reduction in pumping horsepower. This is illustrated in Figure 5.1. The pipe loop can be constructed of the same-diameter pipe as the main pipeline, or in some cases it can be of a different size. As we have seen in the analysis of parallel pipes in Chapter 3, the same diameter of pipe loop will result in equal volumes of gas flow in the main pipe as well as the loop. Thus, an NPS 20 pipe looped with an identical NPS 20 pipe segment will reduce the flow to one-half its original value in each pipe. If the loop is larger or smaller in diameter compared to the main pipeline, the volume distribution will not be equal. An NPS 20 pipe looped with an NPS 16 pipe will result in approximately 64% of the flow rate going through the larger-diameter pipe and 36% through the smaller-diameter pipe.

177

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 178 Friday, April 1, 2005 3:30 PM

178

GAS PIPELINE HYDRAULICS

without loop

Pressure

with loop

loop C

A

Figure 5.1

B

Effect of pipe loop.

5.2 PURPOSE OF COMPRESSION We have seen in Chapter 3 and Chapter 4 that installing intermediate compressor stations along a pipeline will increase the flow rate and also reduce the operating pressure in a long gas transmission pipeline. The installation of the intermediate compressor station will result in additional operational and maintenance issues in comparison with pipe loops. Sometimes, additional compression is installed to increase flow rate in preference to looping the pipeline, since looping will involve additional permitting and right-of-way issues and could cost considerably more than adding the new compressor station. Installation of an intermediate compressor to increase flow rate is illustrated in Figure 5.2.

1440 psig

without second compressor station

with

seco

nd co

mpre

ssor

statio

n

800 psig Flow Q = 100 MMSCFD

Dover

NPS 16 pipeline

Kent Distance

Figure 5.2

Adding a compressor station.

Copyright 2005 by Taylor & Francis Group, LLC

Leeds

2785_C005.fm Page 179 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

179

5.3 INCREASING PIPELINE CAPACITY Consider an existing pipeline that is currently limited by the operating pressure that is close to the MAOP of the pipeline. Suppose the capacity of an NPS 16 pipeline is 100 MMSCFD and the discharge pressure at the originating compressor station is 1440 psig, as shown in Figure 5.2. It can be seen that, at the given flow rate and discharge pressure, the delivery pressure is 800 psig. If the pipeline flow rate is increased to 120 MMSCFD without changing the originating pressure of 1440 psig, the increased flow will cause greater pressure drop and, hence, the delivery pressure at the pipeline terminus will drop to some value such as 600 psig. The reduced delivery pressure may or may not be acceptable to the customer receiving the gas. However, we cannot increase the discharge pressure at the beginning of the pipeline to compensate for the drop in delivery pressure because the pressure is already at the MAOP level. How can we increase the flow rate and still provide the same delivery pressure as before? By installing an intermediate compressor station as shown in Figure 5.2, we can pump the increased volume approximately halfway and then boost the pressure at the new compressor station to the same MAOP level for ultimate delivery to the pipeline terminus at 800 psig, as before. This is illustrated in Figure 5.2. Thus, we have been able to achieve the increased pipeline capacity of 120 MMSCFD by installing an additional compressor station at approximately the halfway point along the pipeline. Suppose we want to increase the flow rate further without changing the discharge pressure or the delivery pressure. It is clear that we could install additional intermediate compressor stations as needed to achieve the increased throughput, while maintaining the same delivery pressure. This is illustrated in Figure 5.3, where two additional compressor stations have been installed to increase the pipeline throughput while maintaining the desired delivery pressure at the pipeline terminus.

1440 psig

800 psig

Q Dover

Figure 5.3

NPS 16 pipeline

Multiple compressor stations.

Copyright 2005 by Taylor & Francis Group, LLC

Kent Distance

Leeds

2785_C005.fm Page 180 Friday, April 1, 2005 3:30 PM

180

GAS PIPELINE HYDRAULICS

Pha

se 1

Ph

as

Pressure

e

2

Q

=

MOP = 1440 psig Q=

238

.41

28

ini

MM

SCF

D

tia

8.

41

lQ

M

=

18

800 psig

8.4

M

1M

SC

MS

FD

CF

D

Q 800 psig Windsor m.p. 0

Avon m.p. 37.55

Hart m.p. 57.33

Cardiff m.p. 100

NPS 16 pipeline 100 mi long Figure 5.4

Windsor to Cardiff pipeline.

However, there is a limit to the number of compressor stations that can be installed in a given pipeline system, since the HP required continues to increase with flow rate and, hence, the capital cost and operating costs increase as well. At some point, the cost increases at a very high rate compared to the increase in flow rate. Each pipe size has a particular volume that can be economically transported based upon cost. An additional factor that must be taken into consideration as the flow rate is increased is the resulting higher velocity. As indicated in Chapter 2, the gas velocity must be well below the erosional velocity for the pipe. Example 1 A natural gas pipeline is 100 mi long and is constructed of NPS 16 and 0.250 in. wall thickness and runs from Windsor to Cardiff, as shown in Figure 5.4. 1. Neglecting elevation effects, calculate the maximum throughput capability of this pipeline, based upon an MAOP of 1440 psig and a delivery pressure of 800 psig at Cardiff. The suction pressure at Windsor is 800 psig. 2. Determine the requirement for two expansion scenarios. The phase 1 expansion will increase pipeline throughput by 50 MMSCFD and phase 2 will increase throughput by another 50 MMSCFD. In each case, calculate the number of compressor stations and HP required. The gas flow velocities must be checked to ensure that they are within erosional limits. 3. Also estimate the approximate cost for each of these cases, using an overall installed cost of $2000 per HP for the compressor stations. 4. Compare these expansion cases using pipe loop instead of compression. Thus, for phase 1, instead of building intermediate compressor stations, calculate the amount

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 181 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

181

of pipe loop needed to reduce the pressure drop at the higher flow rate. Similarly, for the phase 2 flow rate, calculate the looping necessary to maintain pressures without adding compressor stations. Estimate the cost of the expansion scenarios using pipe loops instead of compressor stations, based upon an overall installed cost of $500,000 per mile of loop. Assume a transmission factor of 20, gas flow temperature of 80°F, and compressibility factor of 0.85 throughout. Additional data are as follows: gas gravity = 0.6, ratio of specific heats = 1.4, base temperature = 60°F, and base pressure = 14.7 psia. The compressor isentropic efficiency = 0.8, and the mechanical efficiency of the compressor driver is 0.95. Solution 1. First, determine the initial capacity, considering one compressor station at Windsor providing the pressure of 1440 psig needed for delivery pressure of 800 psig at Cardiff. Using General Flow Equation 2.4, we calculate the initial capacity, Q, of the pipeline as follows: 520 1454.72 − 814.72 Q = 38.77 × 20 14.7 0.6 × 540 × 100 × 0.85

0.5

(15.5)2.5 = 188,410,280 SCFD

The HP required is calculated from Equation 4.15: 0.40 1 + 0.85 1 1454.7 1.40 1.40 − 1 = 6357 HP = 0.0857 × 188.41 (540) 0.40 2 0.8 814.7

BHP required =

6357 = 7064 0.95

Checking gas velocities using Equation 2.26, the gas velocity at Windsor is 188.41 × 10 6 14.7 540 u1 = 0.002122 520 1454.7 = 17.46 ft/s 15.52 The velocity at Cardiff is 188.41 × 10 6 14.7 540 u1 = 0.002122 520 814.7 = 31.18 ft/s 15.52 The erosion velocity from Equation 2.31 is

umax = 100

Copyright 2005 by Taylor & Francis Group, LLC

0.85 × 10.73 × 540 = 58.94 ft/s 29 × 0.6 × 814.7

2785_C005.fm Page 182 Friday, April 1, 2005 3:30 PM

182

GAS PIPELINE HYDRAULICS

2. Next, we will calculate the compressor station requirement for the phase 1 flow rate of Q = 188.41 + 50 = 238.41 MMSCFD Assume that an additional compressor station is needed for this flow rate. This will be located at Avon at a distance of L miles from Cardiff, such that a discharge pressure of 1440 psig at Avon will produce a delivery pressure of 800 psig at Cardiff. We will calculate the value of the pipe length, L , using General Flow Equation 2.4 as follows: 520 1454.72 − 814.72 238.41 × 10 6 = 38.77 × 20 14.7 0.6 × 540 × L × 0.85

0.5

(15.5)2.5

Solving for L, we get

L = 62.45 mi Next, calculate the suction pressure at Avon using 1440 psig at Windsor and considering a pipe length of 37.55 (100 − 62.45) mi between Windsor and Avon. Using General Flow Equation 2.4, we get 520 1454.72 − P22 238.41 × 10 = 38.77 × 20 . 14 7 0 . 6 × 540 × 37 . 55 × 0 . 85

0.5

6

(15.5)2.5

Solving for the suction pressure at Avon, we get P2 = 1114.85 psia = 1100.15 psig Therefore, the compression ratio at Avon is r=

1454.7 = 1.30 1114.85

This is a satisfactory compression ratio for a centrifugal compressor. The HP required at Windsor and Avon for phase 1 will be calculated using Equation 4.15. For Windsor, assuming the compressibility factor at suction is 1.0, 0.40 1 + 0.85 1 1454.7 1.40 1.40 − 1 = 8044 HP = 0.0857 × 238.41 (540) 0.40 2 0.8 814.7

Therefore, the BHP required at Windsor for phase 1 =

8044 = 8468. 0.95

Similarly, the HP required at Avon is 0.40 1.40 1 + 0.85 1 1.40 − 1 = 3476 ( . ) ( ) HP = 0.0857 × 238.41 540 1 30 2 0.8 0.40

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 183 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

Therefore, the BHP required at Avon =

183

3476 = 3659. 0.95

The total compressor HP required at both compressor stations for the phase 1 flow rate of 238 MMSCFD is 8468 + 3659 = 12,127 HP Therefore, the incremental HP for phase 1 is ∆HP = 12,127 – 7064 = 5063 HP 3. This represents the additional compressor HP required for phase 1 for the extra 50 MMSCFD flow rate. The cost of this incremental HP, based on $2000 per installed HP, is ∆Cost = 5063 × 2000 = $10.13 million Next, check the gas velocity at the increased flow rate in phase 1 from Equation 2.26. The velocity at Cardiff is 238.41 × 10 6 14.7 540 u1 = 0.002122 520 814.7 = 39.45 ft/s 15.52 This velocity is acceptable, since it is less than the erosion velocity of 58.94 ft/s calculated earlier. The velocity at Windsor at the higher pressure of 1440 psig will be lower and, hence, less than the erosion velocity. Next, consider the phase 2 flow rate of Q = 238.41 + 50 = 288.41 MMSCFD Since phase 2 occurs after phase 1, where the Avon compressor station is already built, we might have to install one compressor station between Windsor and Avon and another between Avon and Cardiff. If we consider this phase independent of phase 1, we could probably install two additional compressor stations between Windsor and Cardiff to handle the phase 2 flow of 288.41 MMSCFD. For now, we will consider a compressor station at Jenks between Windsor and Avon and another one at Hart located between Avon and Cardiff. We will calculate the distance of L miles from Hart to Cardiff, such that a discharge pressure of 1440 psig at Hart will produce a delivery pressure of 800 psig at Cardiff. The value of L is calculated, using General Flow Equation 2.4, as we did before for locating the Avon compressor station, as follows: 520 1454.72 − 814.72 288.41 × 10 6 = 38.77 × 20 14.7 0.6 × 540 × L × 0.85

0.5

(15.5)2.5

Solving for L, we get L = 42.67 mi Therefore, the Hart compressor station will be located at a distance of 19.78 (62.45 − 42.67) mi from Avon. The suction pressure at Hart is calculated next.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 184 Friday, April 1, 2005 3:30 PM

184

GAS PIPELINE HYDRAULICS

Using General Flow Equation 2.4, we get 520 1454.72 − P22 288.41 × 10 6 = 38.77 × 20 14.7 0.6 × 540 × 19.78 × 0.85

0.5

(15.5)2.5

Solving for the suction pressure at Hart, we get P2 = 1201.24 psia = 1186.54 psig Therefore, the compression ratio at Hart is r=

1454.7 = 1.21 1201.24

This is a satisfactory compression ratio. Before determining the location of the Jenks compressor station between Windsor and Avon, calculate the suction pressure at Avon, assuming Jenks doesn’t exist and that the Windsor compressor station pumps directly into Avon, as in phase 1. The suction pressure at Avon, considering 1440 psig at Windsor, is calculated using General Flow Equation 2.4: 520 1454.72 − P2 2 288.41 × 10 6 = 38.77 × 20 14.7 0.6 × 540 × 37.55 × 0.85

0.5

(15.5)2.5

Solving for the suction pressure at Avon, we get P2 = 915.54 psia = 900.84 psig Therefore, the compression ratio at Avon is r=

1454.7 = 1.59 915.54

This is a satisfactory compression ratio. Therefore, for phase 2, we will need only two compressor stations besides Windsor, Avon at milepost 37.55 and Hart at milepost 57.33. Next, calculate the total HP required at Windsor, Avon, and Hart at phase 2 flow rates. The HP required at Windsor is, using Equation 4.15, 0.40 1 + 0.85 1 1454.7 1.40 1.40 − HP = 0.0857 × 288.41 540 1 ( ) 2 0.8 814.7 = 9731 0.40

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 185 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

185

Therefore, the BHP required at Windsor for phase 2 =

9731 = 10,243. 0.95

Similarly, the HP required at Avon is 0.40 1 + 0.85 1 1.40 HP = 0.0857 × 288.41 (1.59) 1.40 − 1 = 7652 (540) 2 0.8 0.40

Therefore, the BHP required at Avon =

7652 = 8055. 0.95

The HP required at Hart is 0.40 1 + 0.85 1 1.40 HP = 0.0857 × 288.41 (540) (1.21) 1.40 − 1 = 3023 0.40 2 0.8

Therefore, the BHP required at Hart =

3023 = 3182. 0.95

The total compressor HP required at all three compressor stations for phase 2 is 10,243 + 8055 + 3182 = 21,480 HP The incremental HP for phase 2 compared to phase 1 is ∆HP = 21,480 – 12,127 = 9353 HP This represents the additional compression HP required for phase 2 compared to phase 1, for the additional 50 MMSCFD flow rate. The cost of this incremental HP, based on $2000 per installed HP, is ∆Cost = 9353 × 2000 = $18.71 million Next, check the velocity at increased flow rates in phase 2 from Equation 2.26. The velocity at Cardiff is 288.41 × 10 6 14.7 540 u1 = 0.002122 520 814.7 = 47.72 ft/s 15.52 This velocity is acceptable, since it is less than the erosion velocity. The velocity at higher pressures will be well within the limits. 4. In the preceding analysis, we accomplished the increase in flow rates for phase 1 and phase 2 by adding intermediate compressor stations. The capital cost for phase 1 expansion was $10.13 million and for the phase 2 expansion was an additional $18.71 million. Next, we will explore the two expansions by installing pipe loops without additional intermediate compressor stations. For phase 1, assume that L miles of the pipe near Cardiff will be looped. The reason we picked this section is because in Chapter 3, we found that looping close to the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 186 Friday, April 1, 2005 3:30 PM

186

GAS PIPELINE HYDRAULICS

downstream end is more beneficial than looping near the upstream end, as long as the flowing temperature was constant. Following the methodology of Chapter 3, we will determine the equivalent diameter of the pipe loop as follows: Assuming the loop to be of the same diameter as the main piping and L1 = L2, using Equation 3.18, we get Const1 = 1.0 Therefore, the equivalent diameter, using Equation 3.17, is 1/ 5

1 + 1 2 De = D1 1

= 1.32 D1 = 1.32 × 15.5 = 20.46 in.

Considering L miles of pipe of inside diameter 20.46 in., calculate the upstream pressure at the beginning of the loop as shown in Figure 5.5. The downstream pressure at Cardiff is 800 psig, and the upstream pressure P is unknown. Using General Flow Equation 2.4, 520 P 2 − 814.72 238.41 × 10 6 = 38.77 × 20 14.7 0.6 × 540 × L × 0.85

0.5

(20.46)2.5

(5.1)

There are two unknowns, P and L, in Equation 5.1. We need another equation to solve for both variables. For this, the pipe segment from Windsor to the start of the loop will be examined. Considering 1440 psig at Windsor, calculate the downstream pressure P at the beginning of the loop for a pipe length of (100 – L). Using General Flow Equation 2.4, 520 1454.72 − P 2 238.41 × 10 6 = 38.77 × 20 14.7 0.6 × 540 × (100 − L ) × 0.85

0.5

(15.5)2.5

(5.2)

MOP = 1440 psig

Phas

P = 962 psig

= 23

8 MM

Pressure

e 1Q

SCF

D 800 psig

Loop - 50.03 mi

238 MMSCFD 800 psig Windsor m.p. 0

NPS 16 pipeline 100 mi long Figure 5.5

Pipe loop for phase 1.

Copyright 2005 by Taylor & Francis Group, LLC

Cardiff m.p. 100

2785_C005.fm Page 187 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

187

Eliminating P from Equation 5.1 and Equation 5.2, we solve for L as follows: L = 50.03 mi Substituting this value of L in Equation 5.1 and solving for P, P = 976.76 psia = 962.06 psig Therefore, for phase 1, without an intermediate compressor station, flow increase can be achieved by looping 50.03 mi of pipe upstream of Cardiff. The installed cost of this pipe loop is 50.03 × $500,000 = $25.02 million In addition to this cost of pipe loop, we must also include the increased horsepower requirement at Windsor for the phase 1 flow rate. Since the discharge pressure at Windsor is still 1440 psig as before, the HP is the same as that calculated earlier. The incremental HP is (8468 – 7064) = 1404 HP. At $2000 per installed HP, the extra cost for incremental HP is 2000 × 1404 = $2.81 million Thus, for phase 1, the cost of looping pipe upstream of Cardiff and increased HP cost at Windsor compressor station is $25.02 + $2.81 = $27.83 million This compares with $10.13 million calculated earlier for phase 1 using a compressor station at Avon. Even though at first sight the looping appears to be a more expensive option, we must also consider the increased operating cost when adding a compressor station. The annual operating cost for the compressor station can be estimated considering the fuel consumption, operating and maintenance costs, and other costs. In Chapter 10, we will discuss more details of capital cost, operating cost, and cost of service. For now, we will only look at capital costs. For phase 2, at a flow rate of 288.41 MMSCFD, similarly calculate the amount of pipe loop needed, without adding any intermediate compression. The length of loop L required is calculated as follows: Using General Flow Equation 2.4, 520 P 2 − 814.72 288.41 × 10 = 38.77 × 20 . . 14 7 0 6 × 540 × L × 0 . 85 6

0.5

(20.46)2.5

(5.3)

Considering 1440 psig at Windsor, calculate the downstream pressure P at the beginning of the loop for a pipe length of (100 – L). Using General Flow Equation 2.4, 520 1454.72 − P 2 288.41 × 10 6 = 38.77 × 20 14.7 0.6 × 540 × (100 − L ) × 0.85

0.5

(15.5)2.5

(5.4)

Eliminating P from Equation 5.3 and Equation 5.4, we solve for the loop length L as L = 76.26 mi

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 188 Friday, April 1, 2005 3:30 PM

188

GAS PIPELINE HYDRAULICS

MOP = 1440 psig

Pressure

P = 1130 psig Pha

se 2

Q=

288

MM

SCF

D

800 psig

Loop - 76.26 mi

288 MMSCFD 800 psig Windsor m.p. 0

Cardiff m.p. 100 NPS 16 pipeline 100 mi long

Figure 5.6

Pipe loop for phase 2.

This is shown in Figure 5.6. Substituting this value of L in Equation 5.4 and solving for P, P = 1144.54 psia = 1129.84 psig Therefore, the installed cost of this pipe loop is 76.26 × $500,000 = $38.13 million In addition to the cost of pipe loop, we must include the increased horsepower requirement at Windsor for phase 2. Since the discharge pressure at Windsor is still 1440 psig as before, the HP is the same as that calculated earlier. The incremental HP is (10,243 – 8468) = 1775 HP more than that required for phase 1. At $2000 per installed HP, the incremental cost is 1775 × $2000 = $3.55 million compared to phase 1. Thus, for phase 2 the total incremental cost of additional looping over phase 1 and increased HP at the Windsor compressor station is ($38.13 – $25.02) + $3.55 = $16.66 million The costs of the initial case and the two expansion scenarios for the compressor station option and the pipe loop option are summarized in Table 5.1 and Table 5.2. Table 5.1 Windsor to Cardiff Pipeline Expansion—Compressor Station Option Phase

Flow, MMSCFD

Initial 1 2

188.41 238.41 288.41

Copyright 2005 by Taylor & Francis Group, LLC

Compressor BHP Required Windsor Avon Hart 7,064 8,468 10,243

— 3,659 8,055

— — 3,182

Compression Cost, $ million

Incremental Cost, $ million

14.13 24.25 42.96

— 10.13 18.71

2785_C005.fm Page 189 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

189

Table 5.2 Windsor to Cardiff Pipeline Expansion—Pipe Loop Option Pipe Compressor Pipe Compression Loop Total Incremental Flow, BHP Loop, Cost, Cost, Cost, Cost, Phase MMSCFD Windsor mi $ million $ million $ million $ million Initial 1 2

188.41 238.41 288.41

7,064 8,468 10,243

— 50.03 76.26

14.13 16.94 20.49

— 25.02 38.13

14.13 41.96 58.62

— 27.83 16.66

5.4 REDUCING POWER REQUIREMENTS In an existing pipeline for a given flow rate, we can calculate the HP required based upon the number of compressor stations, their suction and discharge pressures, and flow rate. Suppose we are interested in reducing the HP required and, hence, the annual operating cost of the pipeline. If the flow rate is not reduced, the only way power consumption can be reduced is to reduce the overall pressure drop between compressor stations. If the pipeline is 100 miles long and at a flow rate of 100 MMSCFD, an origin compressor station and an intermediate compressor station are required, each station operating at 900 psia suction and 1400 psia discharge pressures. The HP required will depend upon the compression ratio of (1400/900) or 1.56. Since the flow rate is constant, HP can be reduced by increasing the suction pressure or decreasing the discharge pressure, both of which reduce the compression ratio. Since the objective is to operate a gas pipeline at the highest possible pressure for efficiency, we will not reduce the discharge pressure. That leaves us the option of only increasing the suction pressure. Suction pressure can be increased by reducing the pressure drop in the pipeline segment upstream of the compressor station. Since the flow rate and pipe diameter are fixed, the pressure drop in a pipe segment can be decreased by installing a pipe loop. Therefore, looping a segment of pipeline, thereby reducing the pressure drop, will result in a decrease in HP and annual operating cost. We will illustrate this using an example.

Example 2 A natural gas (specific gravity = 0.60) pipeline is 130 mi long and is constructed of NPS 20, 0.500 in. wall thickness pipeline (MAOP = 1440 psig) that runs from Anaheim to Ventura. At a flow rate of 300 MMSCFD, an intermediate compressor at Brentwood (milepost 70) is needed. Calculate the total HP required for both compressor stations. In order to reduce the power consumption by 30% at the present flow rate, it is proposed to loop the pipeline. Calculate the extent of looping required. For simplicity, use the General Flow equation with a transmission factor F = 20 and compressibility factor of 0.90. The gas flow temperature is 60°F and base pressure is 14.7 psia. The base temperature is 60°F. The delivery pressure required at Ventura is 800 psig. The discharge pressure at Anaheim is 1440 psig, and the suction pressure is 900 psig. Use 80% isentropic efficiency and 95% mechanical efficiency for compressors. The gas specific heat ratio g = 1.4.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 190 Friday, April 1, 2005 3:30 PM

190

GAS PIPELINE HYDRAULICS

Solution Using General Flow Equation 2.4, calculate the discharge pressure required at the Brentwood compressor station for 800 psig delivery pressure at Ventura. 520 P12 − 814.72 300 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × 60 × 0.9

0.5

(19.0)2.5

Solving for the discharge pressure at Brentwood, P1 = 1216 psia Next, calculate the suction pressure at Brentwood, applying the General Flow equation to the pipeline segment 70 mi long between Anaheim and Brentwood. 520 1454.72 − P22 300 × 10 = 38.77 × 20 14.7 0.6 × 520 × 70 × 0.9

0.5

6

(19.0)2.5

Solving for the suction pressure at Brentwood, P2 = 1079.25 psia The compression ratio at Anaheim is r=

1454.7 = 1.59 914.7

The HP required at Anaheim is calculated using Equation 4.15: 0.40 1 + 0.9 1 1454.7 1.40 1.40 520 ( ) − HP = 0.0857 × 300 1 2 0.8 914.7 = 7876 0.40

where the compressibility factor for suction conditions is assumed to be 1.0. Considering a mechanical efficiency of 95%, the BHP required at Anaheim is BHP =

7876 = 8291 0.95

Similarly, calculate the compression ratio and BHP for the Brentwood compressor station. The compression ratio at Brentwood is r=

1216 = 1.127 1079.25

The HP required at Brentwood is 0.40 1.40 1.9 1 HP = 0.0857 × 300 (520) (1.127) 1.40 − 1 = 1931 0.40 2 0.8

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 191 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

191

and the BHP is BHP =

1931 = 2033 0.95

By looping the pipe segment between Anaheim and Brentwood using NPS 20 pipe, the flow rate through each pipe will be one-half the inlet flow of 300 MMSCFD at Anaheim. The suction pressure at Brentwood is calculated using the General Flow equation as 520 1454.72 − P22 150 × 10 = 38.77 × 20 0.6 × 520 × 70 × 0.9 . 14 7

0.5

(19.0)2.5

6

Solving for P2, we get P2 = 1371 psia Since this pressure is more than the discharge pressure of 1216 psia calculated earlier for Brentwood, we conclude that the Brentwood station will not be needed if we loop the entire 70 mi pipe segment from Anaheim to Brentwood. This would reduce the total BHP required to 8291 from (8291 + 2033) calculated earlier for the Anaheim and Brentwood compressor stations. The reduction in BHP is ∆BHP =

2033 = 0.197, or 19.7% 8291 + 2033

Since the objective is to reduce the power consumption by 30%, we must do more than just loop the pipe segment between Anaheim and Brentwood. We will recalculate the discharge pressure at Anaheim without the Brentwood compressor station, such that the delivery pressure at Ventura is 800 psig. The reduced discharge pressure at Anaheim due to the 70 mi pipe loop will reduce the compression ratio and, hence, the HP at Anaheim. Using the General Flow equation for the pipe segment between Anaheim and Brentwood, the downstream pressure at Brentwood must equal the 1216 psia calculated earlier to ensure 800 psig delivery at Ventura. Therefore, considering half the total flow rate through each NPS 20 pipe section of the loop, 520 P12 − 12162 150 × 10 = 38.77 × 20 14.7 0.6 × 520 × 70 × 0.9 6

Solving for the discharge pressure at Anaheim, we get P1 = 1310 psia

Copyright 2005 by Taylor & Francis Group, LLC

0.5

(19.0)2.5

2785_C005.fm Page 192 Friday, April 1, 2005 3:30 PM

192

GAS PIPELINE HYDRAULICS

This reduced discharge pressure at Anaheim causes the compression ratio to be r=

1310 = 1.432 914.7

The revised HP at Anaheim is calculated as 0.40 1.40 1.9 1 HP = 0.0857 × 300 (1.432) 1.40 − 1 = 6003 (520) 0.40 2 0.8

BHP =

6003 = 6319 0.95

Therefore, the total reduction in HP is (8291 + 2033) – 6319 = 4005 Or the percentage reduction in HP is 4005 = 0.39 or 39% 8291 + 2033 This is well above the 30% reduction in power required. If we reduce the loop pipe length slightly from 70 mi, we will realize the required 30% reduction. This is left as an exercise for the reader.

5.5 LOOPING IN DISTRIBUTION PIPING Another example of pipe loops is as follows. Consider a distribution piping system as shown in Figure 5.7. Gas enters the pipeline at A at a flow rate of 60 MMSCFD, and after making gas deliveries at B of 20 MMSCFD and at C of 30 MMSCFD, the remaining 10 MMSCFD of gas proceeds to D, where an additional 10 MMSCFD enters the pipeline, which is delivered to the terminus at E. The last segment of pipe has a flow rate of (60 − 20 − 30 + 10) or 20 MMSCFD. Suppose it is desired to bring in an extra 10 MMSCFD gas at D so that the delivery at E is increased to 30 MMSCFD. If the delivery pressure at E is to remain the same at 600 psig, it is clear that the pressure at D will need to be increased to handle the extra flow rate in pipe segment DE. This, in turn, will raise all pressures upstream of D. Thus, the pressures 60 MMSCFD

40 MMSCFD B

A

20 MMSCFD Figure 5.7

Distribution piping.

Copyright 2005 by Taylor & Francis Group, LLC

10 MMSCFD C

30 MMSCFD

20 MMSCFD D

10 MMSCFD

600 psig E

2785_C005.fm Page 193 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

193

at A, B, and C will all increase, resulting in an increased HP requirement at A. However, by looping the section DE, we can maintain all pressures the same as before. Assume that for the initial case, where the injection at D is 10 MMSCFD, the pressure at D is 900 psig. The delivery pressure at E is to be maintained constant at 600 psig. If the entire length of pipe DE is looped with an identical pipe size, the equivalent diameter Deq is such that at 30 MMSCFD, the pressure drop in the diameter Deq is the same as the pressure drop in the original pipe diameter D at 20 MMSCFD. From General Flow Equation 2.4, the flow rate is directly proportional to the square root of (P12 – P22) and also to the pipe diameter raised to the power of 2.5, keeping everything else the same. P1 and P2 are the upstream and downstream pressures in a pipe segment. Since we want the upstream and downstream pressures for the pipe segment DE to be the same at both flow rates, at 20 MMSCFD,

(

)

(

)

20 = C P12 − P22

0.5

D 2.5

(A)

Deq 2.5

(B)

and at 30 MMSCFD, 30 = C P12 − P22

0.5

where C is a constant for the pipe segment. By dividing one equation by the other, we get Deq D

2.5

=

30 = 1.5 20

The equivalent diameter is 1

Deq = D(1.5) 2.5 If the initial pipe size of DE was 12.00 in. inside diameter, we need an equivalent diameter of 1

Deq = 12 × (1.5) 2.5 = 14.11 in. Next, we need to determine the loop diameter required that will produce the equivalent diameter just calculated. Since the pressure drop in each pipe loop is the same, if Q1 and Q2 represent the flow rates in the main pipe and loop respectively, Q1 + Q2 = 30

Copyright 2005 by Taylor & Francis Group, LLC

(5.5)

2785_C005.fm Page 194 Friday, April 1, 2005 3:30 PM

194

GAS PIPELINE HYDRAULICS

and Q1 Q2 30 = = 2.5 2.5 Deq (12) ( D)2.5

(5.6)

where D is the loop diameter to be calculated and the main pipe is 12.00 in. diameter. Solving for Q1, we get 12 Q1 = 30 14.11

2.5

= 20.01 MMSCFD

and the flow rate through the loop is Q2 = 30 − 20.01 = 9.99 MMSCFD Therefore, from Equation 5.6, 9.99 30 = 2.5 D (14.11)2.5 Solving for D, we get 1

9.99 2.5 D= × 14.11 = 9.09 in. 30 Therefore, by looping the entire length DE of the existing 12 in. diameter pipe with a pipe having an inside diameter of 9.09 in., we will maintain the same pressure at all points as before. A slightly different case of looping is one in which the inlet flow at A needs to be increased so that the increased volume can be delivered at B, while keeping all pressures the same as before. Suppose the delivery volume at B needs to be increased to 30 MMSCFD, without changing other deliveries or receipt. The inlet volume at A will increase from 60 MMSCFD to 70 MMSCFD, and the delivery volume at B will increase from 20 MMSCFD to 30 MMSCFD. We will loop the section AB such that the pressure at A and B remain the same as before, so that the volumes and pressures at all points downstream of B remain the same. This will be illustrated by calculating the pressures and the size of the pipe loop required in the next example. Example 3 In a gas distribution pipeline, 60 MMSCFD enters the pipeline at A, as shown in Figure 5.8. If the delivery at B is increased from 20 MMSCFD to 30 MMSCFD by increasing the inlet flow at A, keeping all downstream flow rates the same, calculate the looping necessary for section AB to ensure pressures are not changed throughout the pipeline. Pipe AB is NPS 14, 0.250 in. wall thickness; BC is NPS 12, 0.250 in.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 195 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

70 MMSCFD

40 MMSCFD B

A

20 MMSCFD

10 MMSCFD

30 MMSCFD

600 psig

D

C

30 MMSCFD Figure 5.8

195

E

10 MMSCFD

Looping a distribution piping.

wall thickness; CD is NPS 10, 0.250 in. wall thickness; and DE is NPS 12, 0.250 in. wall thickness. The delivery pressure at E is fixed at 600 psig. The pipe lengths are as follow: AB = 12 mi BC = 18 mi CD = 20 mi DE = 8 mi The gas gravity is 0.60, and the flow temperature is 60°F. The compressibility factor and transmission factor can be assumed to be 0.85 and 20, respectively, throughout the pipeline. The base pressure and base temperature are 14.7 psia and 60°F, respectively. Solution First, the pressures at A and B must be calculated for the initial flow rates. Starting at E, for a delivery pressure of 600 psig at E, the pressures at D, C, and B will be calculated sequentially. Applying the General Flow equation for the 8 mi section DE of inside diameter 12.25 in. and at a flow rate of 20 MMSCFD, 520 PD2 − 614.72 20 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × 8 × 0.85

0.5

(12.25)2.5

Solving for PD, we get PD = 618.02 psia = 603 psig Next, calculate the pressure at C, considering 10 MMSCFD flow through the 20 mi section of pipe CD, of inside diameter 10.25 in.: 520 PC2 − 618.022 10 × 10 = 38.77 × 20 14.7 0.6 × 520 × 20 × 0.85 6

Solving for PC, we get PC = 623.04 psia = 608.34 psig

Copyright 2005 by Taylor & Francis Group, LLC

0.5

(10.25)2.5

2785_C005.fm Page 196 Friday, April 1, 2005 3:30 PM

196

GAS PIPELINE HYDRAULICS

Similarly, the pressure at B is calculated considering 40 MMSCFD flow through the 18 mi section of pipe BC, with an inside diameter 12.25 in.: 520 PB2 − 623.04 2 40 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × 18 × 0.85

0.5

(12.25)2.5

Solving for PB, we get PB = 651.90 psia = 637.20 psig Next, calculate the pressure at A considering 60 MMSCFD flow through the 12 mi section of pipe AB, with an inside diameter 13.5 in.: 520 PA2 − 651.90 2 60 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × 12 × 0.85

0.5

(13.5)2.5

Solving for PA, we get PA = 677.45 psia = 662.75 psig Therefore, the pipe section AB, when flowing 60 MMSCFD of gas, has the following pressures: PA = 677.45 psia PB = 651.90 psia When the delivery rate at B is increased from 20 MMSCFD to 30 MMSCFD, the flow rate in pipe segment AB increases from 60 MMSCFD to 70 MMSCFD. Since the pressures at A and B are to remain the same as before, the pipe segment AB must be looped to reduce the pressure drop at the higher flow rate. We will assume the entire 12 mi length will be looped. Next, we calculate the equivalent diameter required for segment AB, using the General Flow equation, so the pressures at A and B are the same as before. 520 677.452 − 651.90 2 70 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × 12 × 0.85

0.5

( D)2.5

Solving for the diameter, D = 14.36 in. The equivalent diameter of the looped line AB must be 14.36 in. to keep pressures the same as calculated. From Equation 3.17 and Equation 3.18, the diameter of the loop can be calculated, knowing the equivalent diameter just calculated. From Equation 3.17, 1

1 + Const1 2 5 14.36 = 13.50 Const1

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 197 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

197

Solving for Const1 Const1 = 5.99 From Equation 3.18, and since L1 = L2 = 12 mi, 13.5 5.99 = D2

5

Solving for the pipe loop diameter D2, D2 = 6.6 in. Therefore, the pipe section AB must be looped with a pipe of inside diameter 6.6 in. for the entire length of 12 mi. We could also increase the loop diameter and reduce the pipe length that is looped to get the same effect. For example, increasing the loop diameter to 10 in. will reduce the length of looping needed. Suppose we decide on an NPS 10, 0.250 in. wall thickness pipe for the loop length of L mi. upstream of B. The equivalent diameter will be calculated using Equation 3.17 and Equation 3.18. From Equation 3.18, 13.5 Const1 = 10.25

5

= 1.9908

and from Equation 3.17, the equivalent diameter is 1

1 + 1.9908 2 5 = 15.89 in. De = 13.50 1.9908 The pressure at the start of the loop will be calculated from General Flow Equation 2.4: 520 P 2 − 651.90 2 70 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × L × 0.85

0.5

(15.89)2.5

Simplifying, P2 – 651.90 2 = 1705L

(5.7)

Next, consider the unlooped portion of pipe AB from A to the starting point of the loop. Using General Flow Equation 2.4, we get 677.452 − P 2 520 70 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × (12 − L ) × 0.85

Copyright 2005 by Taylor & Francis Group, LLC

0.5

(13.5)2.5

2785_C005.fm Page 198 Friday, April 1, 2005 3:30 PM

198

GAS PIPELINE HYDRAULICS

Simplifying, 677.452 – P2 = 3851.92 (12 – L)

(5.8)

Eliminating P from Equation 5.7 and Equation 5.8 and solving for L, we get L = 5.71 mi Therefore, by looping the existing NPS 14 pipe from A to B with an identical NPS 14 pipe, 5.71 mi long (measured upstream from B), the pressures will be the same as before the increased delivery volume at B.

5.6 SUMMARY We discussed two ways to increase the throughput of a gas pipeline: using intermediate compressor stations and installing pipe loops. With intermediate compressor stations, the flow rate can be increased to fully utilize pipe MAOP. However, adding compressor stations causes increased capital cost as well as annual operating and maintenance costs. On the other hand, by installing a pipe loop, the effective diameter of the pipe is increased, resulting in a lower pressure drop. Therefore, additional flow rate can be realized without installing an intermediate compressor station. Looping an existing pipeline causes increase in capital but very little increase in operating and maintenance costs compared to installing intermediate compressor stations. We also discussed how the HP required can be reduced by installing a pipe loop. On distribution piping, an example of increasing delivery rate to certain locations using pipe loops, without changing pipe pressures in the rest of the pipeline, was also illustrated.

PROBLEMS 1. A natural gas pipeline from Compton to Merced is 100 mi long and is constructed of NPS 14, 0.250 in. wall thickness. The pipeline elevation profile is essentially flat. The MAOP of the pipeline is 1280 psig. The gas delivery pressure at Merced is 600 psig. What is the maximum pipeline throughput with an origin compressor station at Compton? The gas gravity is 0.6 and gas flowing temperature is 80°F. Use the Colebrook equation for pressure drop with a friction factor of 0.01. The compressibility factor can be assumed to be constant at 0.88. If the flow rate increases by 50 MMSCFD, calculate the increased HP required at Compton and the HP required at an intermediate compressor station at Vale. Instead of the intermediate compressor station at Vale, a portion of the pipe is looped. What length of NPS 14 loop will be needed? The base pressure and base temperature are 14.7 psia and 60°F, respectively. The compressor isentropic efficiency = 0.8, and the mechanical efficiency of the compressor driver is 0.95. 2. A natural gas (specific gravity = 0.60) pipeline is 120 mi long and is constructed of NPS 20, 0.500 in. wall thickness (MAOP = 1000 psig), and runs from Akers to Coburn. At a flow rate of 250 MMSCFD, an intermediate compressor at

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 199 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

199

Bradley (milepost 65) is required. Calculate the total HP required. In order to reduce the power consumption by 20% at the present flow rate, it is proposed to loop the pipeline. Calculate the length of looping required. Use the Panhandle A equation with 95% efficiency. The compressibility factor can be assumed constant at 0.90. The gas flow temperature is 70°F, and the base pressure and base temperature are 14.7 psia and 60°F, respectively. The delivery pressure required at Coburn is 750 psig. The discharge pressure at Akers is 1000 psig, and the suction pressure is 850 psig. Use 80% isentropic efficiency and 95% mechanical efficiency for the compressors. 3. In a gas distribution pipeline, similar to that shown in Figure 5.8, gas enters the pipeline at A at a flow rate of 50 MMSCFD. At B and C, deliveries of 10 MMSCFD and 20 MMSCFD are made. At D, an additional volume of gas at 15 MMSCFD enters the pipeline. Calculate the pressures at the various pipe nodes A, B, C, and D, considering a delivery pressure of 500 psig at E. If the incoming volume at D is increased to 25 MMSCFD and all pressures are to remain the same, how much of the pipe DE should be looped? The pipe lengths are as follows: AB: NPS 12, 0.250 length = 18 mi BC: NPS 10, 0.250 length = 24 mi CD: NPS 8, 0.250 length = 16 mi DE: NPS 12, 0.250 length = 20 mi

REFERENCES 1. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 2. Engineering Data Book, 10th ed., Gas Processors Suppliers Association, Tulsa, OK, 1994. 3. Pipeline Design for Hydrocarbon Gases and Liquids, American Society of Civil Engineers, New York, 1975.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 200 Friday, April 1, 2005 3:30 PM

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 201 Friday, April 1, 2005 3:29 PM

CHAPTER

6

Pipe Analysis In this chapter we will discuss the mechanical strength needed for a pipeline transporting gas. We will analyze the impact of pipe diameter, wall thickness, material of construction, and specific safety requirements dictated by design codes and state and federal regulations. Also covered will be testing requirements and classification of pipelines based upon their proximity to human dwellings and industrial establishments and population density. The importance of mainline block valves and calculation of blowdown time to isolate sections of a gas pipeline will also be discussed.

6.1 PIPE WALL THICKNESS In Chapter 3 we calculated the pressure needed to transport a given volume of gas through a pipeline. The internal pressure in a pipe causes the pipe wall to be stressed, and if allowed to reach the yield strength of the pipe material, it could cause permanent deformation of the pipe and ultimate failure. Obviously, the pipe should have sufficient strength to handle the internal pressure safely. In addition to the internal pressure due to gas flowing through the pipe, the pipe might also be subjected to external pressure. External pressure can result from the weight of the soil above the pipe in a buried pipeline and also by the loads transmitted from vehicular traffic in areas where the pipeline is located below roads, highways, and railroads. The deeper the pipe is buried, the higher will be the soil load on the pipe. However, the pressure transmitted to the pipe due to vehicles above ground will diminish with the depth of the pipe below the ground surface. Thus, the external pressure due to vehicular loads on a buried pipeline that is 6 ft below ground will be less than that on a pipeline that is at a depth of 4 ft. In most cases involving buried pipelines transporting gas and other compressible fluids, the effect of the internal pressure is more than that of external loads. Therefore, the necessary minimum wall thickness will be dictated by the internal pressure in a gas pipeline. The minimum wall thickness required to withstand the internal pressure in a gas pipeline will depend upon the pressure, pipe diameter, and pipe material. The larger 201

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 202 Friday, April 1, 2005 3:29 PM

202

GAS PIPELINE HYDRAULICS

the pressure or diameter, the larger would be the wall thickness required. Higherstrength steel pipes will require less wall thickness to withstand the given pressure compared to low-strength materials. The commonly used formula to determine the wall thickness for internal pressure is known as Barlow’s equation. This equation has been modified to take into account design factors and type of pipe joints (seamless, welded, etc.) and is incorporated into design codes such as DOT Code of Federal Regulations Part 192 and ASME B31.8 Standards. See Chapter 9 for a full list of design codes and standards used in the design, construction, and operation of gas pipelines.

6.2 BARLOW’S EQUATION When a circular pipe is subject to internal pressure, the pipe material at any point will have two stress components at right angles to each other. The larger of the two stresses is known as the hoop stress and acts along the circumferential direction. Hence, it is also called the circumferential stress. The other stress is the longitudinal stress, also known as the axial stress, which acts in a direction parallel to the pipe axis. Figure 6.1 shows a cross section of a pipe subject to internal pressure. An element of the pipe wall material is shown with the two stresses Sh and Sa in perpendicular directions. Both stresses will increase as the internal pressure is increased. As will be shown shortly, the hoop stress Sh is the larger of the two stresses and, hence, will govern the minimum wall thickness required for a given internal pressure. In its basic form, Barlow’s equation relates the hoop stress in the pipe wall to the internal pressure, pipe diameter, and wall thickness as follows: Sh =

PD 2t

(6.1)

L

Axial stress - Sa

Axial stress - Sa

Sh Pressure - P

Diameter - D

Figure 6.1

Stresses in pipe subject to internal pressure.

Copyright 2005 by Taylor & Francis Group, LLC

s - Sh

tres

ps Hoo

Sh

2785_C006.fm Page 203 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

where Sh = P = D = t =

203

hoop or circumferential stress in pipe material, psi internal pressure, psi pipe outside diameter, in. pipe wall thickness, in.

Similar to Equation 6.1, the axial (or longitudinal) stress, Sa, is given by the following equation: Sa =

PD 4t

(6.2)

Note that in these equations the pipe diameter used is the outside diameter, not the inside diameter as we used in Chapter 2 and Chapter 3. For example, consider an NPS 20 pipe, 0.500 in. wall thickness, that is subject to an internal gas pressure of 1200 psig. The pipe wall material will be stressed in the circumferential direction by the hoop stress given by Equation 6.1 as follows: Sh =

1200 × 20 = 24,000 psig 2 × 0.500

and in accordance with Equation 6.2, the axial stress in the pipe wall is Sa =

1200 × 20 = 12,000 psig 4 × 0.500

Barlow’s equation is valid only for thin-walled cylindrical pipes. Most pipelines transporting gases and liquids generally fall in this category. There are instances in which pipes carrying gases and petroleum liquids, subject to high external loads, such as deep submarine pipelines, may be classified as thick-walled pipes. The governing equations for such thick-walled pipes are different and more complex. We will introduce these formulas for information only. 6.3 THICK-WALLED PIPES Consider a thick-walled pipe with an outside diameter DO and inside diameter of Di, subject to an internal pressure of P. The greatest stress in the pipe wall will be found to occur in the circumferential direction near the inner surface of the pipe. This stress can be calculated from the following equation: Smax =

(

P Do2 + Di2

(D

2 o

−D

2 i

)

)

(6.3)

The pipe wall thickness is t=

Copyright 2005 by Taylor & Francis Group, LLC

Do − Di 2

(6.4)

2785_C006.fm Page 204 Friday, April 1, 2005 3:29 PM

204

GAS PIPELINE HYDRAULICS

Rewriting Equation 6.3 in terms of outside diameter and wall thickness, we get D 2 + ( Do − 2t )2 Smax = P o2 2 Do − ( Do − 2t ) Simplifying further,

Smax

PDo = 2t

( ) ( ) ( )

t 1 − Do + 2 1 + Dt o

2

t Do

(6.5)

In the limiting case, a thin-walled pipe is one in which the wall thickness is very small compared to the diameter Do. In this case (t /D) is small compared to 1 and, therefore, can be neglected in Equation 6.5. Therefore, the approximation for thinwalled pipes from Equation 6.5 becomes Smax =

PDo 2t

which is the same as Barlow’s Equation 6.1 for hoop stress. Example 1 A gas pipeline is subject to an internal pressure of 1400 psig. It is constructed of steel pipe with 24 in. outside diameter and 0.75 in. wall thickness. Calculate the maximum hoop stress in the pipeline, considering both the thin-walled approach and the thick-walled equation. What is the error in assuming that the pipe is thin walled? Solution Pipe inside diameter = 24 – 2 × 0.75 = 22.5 in. From Equation 6.1 for thin-walled pipe, Barlow’s equation gives the maximum hoop stress as Sh =

1400 × 24 = 22,400 psig 2 × 0.75

Considering the thick-walled pipe formula Equation 6.3, Smax =

1400(24 2 + 22.52 ) = 21,723 psig (24 2 − 22.52 )

Therefore, by assuming thin-walled pipe, the hoop stress is overestimated by approximately 22, 400 − 21, 723 = 0.0312 or 3.12% 21, 723

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 205 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

205

6.4 DERIVATION OF BARLOW’S EQUATION Since Barlow’s equation is the basic equation for pipes under internal pressure, it is appropriate to understand how the formula is derived, which is the subject of this section. Consider a circular pipe of length L, outside diameter D, and wall thickness t as shown in Figure 6.1. We consider the cross section of one-half portion of this pipe. The pipe is subject to an internal pressure of P psig. Within the pipe material, the hoop stress Sh and the axial stress Sa act at right angles to each other as shown. Considering the one-half section of the pipe, for balancing the forces in the direction of the hoop stress Sh, we can say that Sh, acting on the two rectangular areas L × t , balances the internal pressure on the projected area D × L. Therefore, P × D × L = Sh × L × t × 2

(6.6)

Solving for Sh , we get the derivation of Equation 6.1 as Sh =

PD 2t

Now we will look at the balancing of longitudinal forces. The internal pressure P acting on the cross-sectional area of pipe π4 D 2 produces the bursting force. This is balanced by the axial resisting force Sa acting on the area π Dt . Therefore,

π 2 D = Sa × π Dt 4

(6.7)

Solving for Sa, we get the derivation of Equation 6.2 as Sa =

PD 4t

It can be seen from the preceding equations that the hoop stress is twice the axial stress and, therefore, is the governing stress. Consider a pipe with 20 in. outside diameter and 0.500 in. wall thickness subject to an internal pressure of 1000 psig. From Barlow’s Equation 6.1 and Equation 6.2, we calculate the hoop stress and axial stress as follows: Sh =

1000 × 20 = 20,000 psig 2 × 0.500

Sa =

1000 × 20 = 10,000 psig 4 × 0.500

Therefore, we are able to determine the stress levels in the pipe material for a given internal pressure, pipe diameter, and wall thickness. If the above-calculated

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 206 Friday, April 1, 2005 3:29 PM

206

GAS PIPELINE HYDRAULICS

values are within the stress limits of the pipe material, we can conclude that the NPS 20 pipe with 0.500 in. wall thickness is adequate for the internal pressure of 1000 psig. The yield stress of the pipe material represents the stress at which the pipe material yields and undergoes permanent deformation. Therefore, we must ensure that the stress calculations above do not come dangerously close to the yield stress. Frequently, we have to solve the reverse problem of determining the wall thickness of a pipeline for a given pressure. For example, suppose the pipe is constructed of steel with a yield strength of 52,000 psi and we are required to determine what wall thickness is needed for NPS 20 pipe to withstand 1400 psig internal pressure. If we are allowed to stress the pipe material to no more than 60% of the yield stress, we can easily calculate the minimum wall thickness required using Equation 6.1, as follows: 0.6 × 52, 000 =

1400 × 20 2t

Here, we have equated the hoop stress per Barlow’s equation to 60% yield strength of the pipe material. Solving for pipe wall thickness, we get t = 0.4487 in. Suppose we used the nearest standard wall thickness of 0.500 in. The actual hoop stress can then be calculated from Barlow’s equation as Sh =

1400 × 20 = 28,000 psi 2 × 0.5

28,000 Therefore, the pipe will be stressed to 52 ,000 = 0.54 or 54% of yield stress, which is less than the 60% we started with. Incidentally, the actual axial or longitudinal stress in the preceding example will be one-half the hoop stress or 14,000 psi. Therefore, in this basic example, we used Barlow’s equation to calculate the pipe wall thickness required for a NPS 20 pipe to withstand an internal pressure of 1400 psig without stressing the pipe material beyond 60% of its yield strength. In the foregoing, we arbitrarily picked 60% of the yield stress of pipe material to calculate the pipe wall thickness. We did not use 100% of the yield stress because, in this case, the pipe material would yield at the given pressure, which obviously cannot be allowed. In design, we generally use a design factor that is a number less than 1.00 that represents the fraction of the yield stress of the pipe material that the pipe can be stressed to. Gas pipelines are designed with various design factors ranging from 0.4 to 0.72. This means that the pipe hoop stress is allowed to be between 40 and 72% of the yield strength of pipe material. The actual percentage will depend on various factors and will be discussed shortly. The yield stress used in the calculation of pipe wall thickness is called the specified minimum yield strength (SMYS) of pipe material. Thus, in the preceding example, we calculated the pipe wall thickness based on a design factor of 0.6 or allowed the pipe stress to go up to 60% of the SMYS.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 207 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

207

Table 6.1 Pipe Material and Yield Strength Pipe Material API 5LX Grade

Specified Minimum Yield Strength (SMYS), psi

X42 X46 X52 X56 X60 X65 X70 X80 X90

42,000 46,000 52,000 56,000 60,000 65,000 70,000 80,000 90,000

6.5 PIPE MATERIAL AND GRADE Steel pipes used in gas pipeline systems generally conform to API 5L and 5LX specifications. These are manufactured in grades ranging from X42 to X90 with SMYS, as shown in Table 6.1. Sometimes API 5L grade B pipe with 35,000 psi SMYS is also used in certain installations.

6.6 INTERNAL DESIGN PRESSURE EQUATION We indicated earlier in this chapter that Barlow’s equation, in a modified form, is used in designing gas pipelines. The following form of Barlow’s equation is used in design codes for petroleum transportation systems to calculate the allowable internal pressure in a pipeline based upon given diameter, wall thickness, and pipe material. P=

2tSEFT D

(6.8)

where P = internal pipe design pressure, psig D = pipe outside diameter, in. t = pipe wall thickness, in. S = specified minimum yield strength (SMYS) of pipe material, psig E = seam joint factor, 1.0 for seamless and submerged arc welded (SAW) pipes. F = design factor, usually 0.72 for cross-country gas pipelines, but can be as low as 0.4, depending on class location and type of construction T = temperature deration factor = 1.00 for temperatures below 250°F It must be noted that in the foregoing, we used the outside diameter of the pipe and not the inside diameter as used in Chapter 2 and Chapter 3 for pressure drop calculations.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 208 Friday, April 1, 2005 3:29 PM

208

GAS PIPELINE HYDRAULICS

Table 6.2 Pipe Seam Joint Factors Specification ASTM A53

ASTM ASTM ASTM ASTM ASTM ASTM ASTM ASTM

A106 A134 A135 A139 A211 A333 A333 A381

ASTM A671 ASTM A672 ASTM A691 API 5L

API 5LX

API 5LS

Seam Joint Factor (E)

Pipe Class Seamless Electric Resistance Welded Furnace Lap Welded Furnace Butt Welded Seamless Electric Fusion Arc Welded Electric Resistance Welded Electric Fusion Welded Spiral Welded Pipe Seamless Welded Double Submerged Arc Welded Electric-Fusion-Welded Electric-Fusion-Welded Electric-Fusion-Welded Seamless Electric Resistance Welded Electric Flash Welded Submerged Arc Welded Furnace Lap Welded Furnace Butt Welded Seamless Electric Resistance Welded Electric Flash Welded Submerged Arc Welded Electric Resistance Welded Submerged Arc Welded

1 1 0.8 0.6 1 0.8 1 0.8 0.8 1 1 1 1 1 1 1 1 1 1 0.8 0.6 1 1 1 1 1 1

The seam joint factor E used in Equation (6.8) varies with the type of pipe material and welding employed. Seam joint factors are given in Table 6.2 for the most commonly used pipe and joint types. The internal design pressure calculated from Equation (6.8) is known as the maximum allowable operating pressure (MAOP) of the pipeline. This term has been shortened to maximum operating pressure (MOP) in recent years. Throughout this book we will use MOP and MAOP interchangeably. The design factor F has values ranging from 0.4 to 0.72, as mentioned earlier. Table 6.3 lists the values of the design factor based upon class locations. The class locations, in turn, depend on the population density in the vicinity of the pipeline. Table 6.3 Design Factors for Steel Pipe

Copyright 2005 by Taylor & Francis Group, LLC

Class Location

Design Factor, F

1 2 3 4

0.72 0.60 0.50 0.40

2785_C006.fm Page 209 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

209

220 yards pipe centerline 220 yards

1 mi Figure 6.2

Class location unit.

6.7 CLASS LOCATION The following definitions of class 1 through class 4 are taken from DOT 49 CFR, Part 192 (see Reference section for details). The class location unit (CLU) is defined as an area that extends 220 yards on either side of the center line of a 1-mi section of pipe, as indicated in Figure 6.2. Class 1 Offshore gas pipelines are Class 1 locations. For onshore pipelines, any class location unit that has 10 or fewer buildings intended for human occupancy is termed Class 1. Class 2 This is any class location unit that has more than 10 but fewer than 46 buildings intended for human occupancy. Class 3 This is any class location unit that has 46 or more buildings intended for human occupancy or an area where the pipeline is within 100 yards of a building or a playground, recreation area, outdoor theatre, or other place of public assembly that is occupied by 20 or more people at least 5 days a week for 10 weeks in any 12-month period. The days and weeks need not be consecutive. Class 4 This is any class location unit where buildings with four or more stories above ground exist.

The temperature deration factor T is equal to 1.00 up to gas temperature 250°F, as indicated in Table 6.4. Table 6.4 Temperature Deration Factors Temperature °F

°C

Deration Factor T

250 or less 300 350 400 450

121 or less 149 177 204 232

1.000 0.967 0.033 0.900 0.867

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 210 Friday, April 1, 2005 3:29 PM

210

GAS PIPELINE HYDRAULICS

Example 2 A gas pipeline is constructed of API 5L X65 steel, NPS 16, 0.250 in. wall thickness. Calculate the MAOP of this pipeline for class 1 through class 4 locations. Use a temperature deration factor of 1.00. Solution Using Equation 6.8, the MAOP is given by P=

2 × 0.250 × 65, 000 × 1.0 × 0.72 × 1.0 = 1462.5 psig for class 1 16

Similarly, MAOP = 1462.5 ×

0.6 = 1218.8 psig for class 2 0.72

MAOP = 1462.5 ×

0.5 = 1015.62 psig for class 3 0.72

MAOP = 1462.5 ×

0.4 = 812.5 psig for class 4 0.72

6.8 MAINLINE VALVES Mainline valves are installed in gas pipelines so that portions of the pipeline can be isolated for hydrostatic testing and maintenance. Valves are also necessary to separate sections of pipe and minimize gas loss that can occur due to pipe rupture from construction damage. Design codes specify the spacing of these valves based upon class location, which in turn depends on the population density around the pipeline. The following lists the maximum spacing between mainline valves in gas transmission piping. These are taken from ASME B31.8 code. Class Location 1 2 3 4

Valve Spacing 20 15 10 5

miles miles miles miles

It can be seen from the preceding that the valve spacing is shorter as the pipeline traverses high-population areas. This is necessary as a safety feature to protect the inhabitants in the vicinity of the pipeline by restricting the amount of gas that might escape due to rupture of the pipeline. These mainline valves must be full-opening, through-conduit type valves such that scraper pigs and inspection tools can pass through these valves without any obstruction. Therefore, ball valves and gate valves are used of the welded construction rather than flanged type. Buried valves have

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 211 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

211

extended stems with elevated valve operators located above ground, with lubrication and bleed lines brought above ground for easy access and maintenance.

6.9 HYDROSTATIC TEST PRESSURE When a pipeline is designed to operate at a certain MOP, it must be tested to ensure that it is structurally sound and can withstand safely the internal pressure before being put into service. Generally, gas pipelines are hydrotested with water by filling the test section of the pipe with water and pumping the pressure up to a value higher than the MAOP and holding it at this test pressure for a period of 4 to 8 hours. The magnitude of the test pressure is specified by design code, and it is usually 125% of the operating pressure. Thus, a pipeline designed to operate continuously at 1000 psig will be hydrotested to a minimum pressure of 1250 psig. Consider a pipeline NPS 24, with 0.375 in. wall thickness, constructed of API 5L X65 pipe. Using a temperature deration factor of 1.00, we calculate the MOP of this pipeline from Equation 6.8 for class 1 location as follows: P=

2 × 0.375 × 65, 000 × 1.0 × 0.72 × 1.0 = 1462.5 psig 24

Since the pipe fittings and valves will be ANSI 600, we will establish an MOP of 1440 psig for this pipeline. Therefore, the hydrotest pressure will be 1.25 × 1440 = 1800 psig If the pipeline is designed to be below ground, the test pressure is held constant for a period of 8 hours, and it is thoroughly checked for leaks. Above-ground pipelines are tested for a period of 4 hours. If the design factor used in the MOP calculation is 0.72 (class 1), the hoop stress is allowed to reach 72% of the SMYS of pipe material. Testing this pipe at 125% of MOP will result in the hoop stress reaching a value of 1.25 × 0.72 = 0.90 or 90% of SMYS. Thus, by hydrotesting the pipe at 1.25 times the operating pressure, we are stressing the pipe material to 90% of the yield strength. Generally, the hydrotest pressure is given such that the hoop stress has a range of values, such as 90 to 95% of SMYS. Therefore, in the preceding example, the minimum and maximum hydrotest pressures will be as follows: Minimum hydrotest pressure = 1.25 × 1440 = 1800 psig Maximum hydrotest pressure = 1800 × (95/90) = 1900 psig It can be seen from Equation 6.1 that the 1800 psig internal pressure will cause a hoop stress of Sh =

Copyright 2005 by Taylor & Francis Group, LLC

1800 × 24 = 57,600 psi 2 × 0.375

2785_C006.fm Page 212 Friday, April 1, 2005 3:29 PM

212

GAS PIPELINE HYDRAULICS

Dividing this hoop stress by the SMYS, we get the lower limit of the hydrotest pressure as 57, 600 = 0.89 = 89% of SMYS 65, 000 Similarly, by proportion, the maximum hydrotest pressure of 1900 psig will cause a hoop stress of Sh =

1900 × 57, 600 = 60,800 = 94% of SMYS 1800

Therefore, in this example, the hydrotest envelope of 1800 to 1900 psig is equivalent to stressing the pipe in the range of 89 to 94% of SMYS. In the preceding analysis we have not taken into consideration the pipeline elevation profile in calculating the hydrotest pressures. Generally, a long pipeline is divided into test sections and the hydrotest pressures are established for each section, taking into account the elevations along the pipeline profile. The reason for subdividing the pipeline into sections for hydrotesting will be evident from the following example. Consider, for example, a pipeline 50 mi long with an elevation profile as shown in Figure 6.3. The elevation of the starting point, Norwalk, is 300 ft, whereas the pipeline terminus, Lakewood, is at an elevation of 1200 ft. If the entire 50 mi length of the pipeline were filled with water for hydrotesting, the static pressure difference between the two ends due to elevation will be as follows: Pressure difference = (1200 – 300) × 0.433 = 389.7 psig The factor 0.433 is the conversion factor from feet of water to pressure in psig. It can be seen that the pipe section at the low elevation point at Norwalk will be at a higher pressure than the pipeline at the high elevation end at Lakewood by almost 390 psig. Therefore, if we pump the water in the line to the required hydrotest

Hydrotest pressure: 1800 psig

1410 psig

1200 ft 390 psig Pipeline elevation profile 300 ft Norwalk Figure 6.3

50 mi

Pipeline with elevation profile—impact on hydrotest.

Copyright 2005 by Taylor & Francis Group, LLC

Lakewood

2785_C006.fm Page 213 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

213

pressure of 1800 psig at Norwalk, the corresponding water pressure at Lakewood will be 1800 – 390 = 1410 psig Conversely, if we pump the water in the line to the required hydrotest pressure of 1800 psig at Lakewood, the corresponding water pressure at Norwalk will be 1800 + 390 = 2190 psig This is shown in Figure 6.3. The pressure of 2190 psig at Norwalk will result in a hoop stress of Sh =

2190 × 24 = 70,080 psi 2 × 0.375

This is equivalent to 70, 080 = 1.08 = 108% of SMYS 65, 000 Obviously, we have exceeded the yield strength of the pipe material, and this is not acceptable. On the other hand, with 1800 psig test pressure at Norwalk, the corresponding test pressure at Lakewood is calculated to be 1410 psig. Even though the pipe section at the low end at Norwalk has the requisite test pressure (125% MOP), the pipe section at the higher elevation at Lakewood will see only 1410 × 125 = 98% MOP 1800 This will not be an acceptable hydrotest, because we have not been able to test the entire pipeline at the correct hydrotest pressure, which must be at least 125% of the MAOP. The solution to this dilemma is to break the length of 50 mi into several sections such that each section can be tested separately at the required test pressure. These test sections will have smaller elevation differences between the ends of the test sections. Therefore, each section will be hydrotested to pressures close to the required minimum pressure. Figure 6.4 shows such a pipeline subdivided into sections suitable for hydrotesting. Using the hydrotest envelope of 90 to 95% of SMYS, we will be able to adjust the test pressures for each section such that even with some elevation difference between the ends of each test section, the hydrotest pressures may be close to the required pressures. This will not be possible if we have one single test section with significant elevation difference between the two ends, as illustrated in Figure 6.4. Table 6.5 through Table 6.13 list the internal design pressure and hydrostatic test pressure for various pipe diameters and pipe materials ranging from X42 to X90.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 214 Friday, April 1, 2005 3:29 PM

214

GAS PIPELINE HYDRAULICS

Hydrotest pressure profile

Pipeline elevation profile San Juan Figure 6.4

250 km

Cadiz

Hydrotesting by subdividing pipeline.

Example 3 A gas pipeline, NPS 20, 0.500 in. wall thickness, is constructed of API 5L X52 pipe. (a) Calculate the design pressures for class 1 through class 4 locations. (b) What is the range of hydrotest pressures for each of these class locations? Assume joint factor = 1.00 and temperature deration factor = 1.00. Solution Using Equation 6.8, the internal design pressure is P=

2 × 0.500 × 52, 000 × 1.00 × 1.0 × F = 2600 F 20

where F = design factor = 0.72 for class 1 Therefore, the design pressures for class 1 through class 4 are as follows: Class Class Class Class

1 2 3 4

= = = =

2600 2600 2600 2600

× × × ×

0.72 0.60 0.50 0.40

= = = =

1872 1560 1300 1040

psig psig psig psig

The range of hydrotest pressures is such that the hoop stress will be between 90 and 95% of SMYS.

Copyright 2005 by Taylor & Francis Group, LLC

Wall Thickness in.

4.5

0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500

6.625

8.625

10.75

12.75

14.00

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09

42000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 3185 4529 5873 7137 2282 2556 3944 5131 1753 1942 2258 2847 1407 1727 2054 2813 1186 1565 1779 1926 2372 1080 1348 1620 1888 2160

2654 3774 4894 5947 1902 2130 3286 4275 1461 1619 1882 2372 1172 1439 1711 2344 988 1304 1482 1605 1976 900 1123 1350 1573 1800

2212 3145 4079 4956 1585 1775 2739 3563 1217 1349 1568 1977 977 1199 1426 1953 824 1087 1235 1337 1647 750 936 1125 1311 1500

1770 2516 3263 3965 1268 1420 2191 2850 974 1079 1254 1582 781 960 1141 1563 659 870 988 1070 1318 600 749 900 1049 1200

3982 5662 7342 8921 2853 3195 4930 6413 2191 2428 2822 3559 1758 2159 2567 3516 1482 1957 2224 2407 2965 1350 1685 2025 2360 2700

4203 5976 7749 9416 3011 3373 5204 6769 2313 2563 2979 3756 1856 2279 2709 3712 1565 2065 2347 2541 3129 1425 1778 2138 2491 2850

4424 6291 8157 9912 3170 3550 5477 7126 2435 2698 3136 3954 1953 2399 2852 3907 1647 2174 2471 2675 3294 1500 1872 2250 2622 3000

Copyright 2005 by Taylor & Francis Group, LLC

215

(continued )

2785_C006.fm Page 215 Friday, April 1, 2005 3:29 PM

Pipe Material API 5L X42 Diameter in.

PIPE ANALYSIS

Table 6.5 Pipeline Internal Design Pressures and Test Pressures

Pipe Material API 5L X42

SMYS

Wall Thickness in.

Weight lb/ft

16.00

0.250 0.312 0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750

42.05 52.27 62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25

18.00

20.00

22.00

24.00

26.00

Copyright 2005 by Taylor & Francis Group, LLC

42000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 945 1179 1418 1652 1890 840 1048 1260 1468 1680 943 1134 1321 1512 1699 1031 1375 1718 2062 945 1101 1260 1416 1575 1890 872 1163 1454 1745

788 983 1181 1377 1575 700 874 1050 1224 1400 786 945 1101 1260 1416 859 1145 1432 1718 788 918 1050 1180 1313 1575 727 969 1212 1454

656 819 984 1147 1313 583 728 875 1020 1167 655 788 918 1050 1180 716 955 1193 1432 656 765 875 984 1094 1313 606 808 1010 1212

525 655 788 918 1050 467 582 700 816 933 524 630 734 840 944 573 764 955 1145 525 612 700 787 875 1050 485 646 808 969

1181 1474 1772 2065 2363 1050 1310 1575 1835 2100 1179 1418 1652 1890 2124 1289 1718 2148 2577 1181 1377 1575 1770 1969 2363 1090 1454 1817 2181

1247 1556 1870 2180 2494 1108 1383 1663 1937 2217 1245 1496 1744 1995 2242 1360 1814 2267 2720 1247 1453 1663 1869 2078 2494 1151 1535 1918 2302

1313 1638 1969 2294 2625 1167 1456 1750 2039 2333 1310 1575 1835 2100 2360 1432 1909 2386 2864 1313 1530 1750 1967 2188 2625 1212 1615 2019 2423

GAS PIPELINE HYDRAULICS

Diameter in.

2785_C006.fm Page 216 Friday, April 1, 2005 3:29 PM

216

Table 6.5 Pipeline Internal Design Pressures and Test Pressures (Continued)

32.00

34.00

36.00

42.00

110.64 146.85 182.73 218.27 118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

810 1080 1350 1620 756 1008 1260 1512 709 945 1181 1418 667 889 1112 1334 630 840 1050 1260 540 720 900 1080 1440

675 900 1125 1350 630 840 1050 1260 591 788 984 1181 556 741 926 1112 525 700 875 1050 450 600 750 900 1200

563 750 938 1125 525 700 875 1050 492 656 820 984 463 618 772 926 438 583 729 875 375 500 625 750 1000

450 600 750 900 420 560 700 840 394 525 656 788 371 494 618 741 350 467 583 700 300 400 500 600 800

1013 1350 1688 2025 945 1260 1575 1890 886 1181 1477 1772 834 1112 1390 1668 788 1050 1313 1575 675 900 1125 1350 1800

1069 1425 1781 2138 998 1330 1663 1995 935 1247 1559 1870 880 1174 1467 1760 831 1108 1385 1663 713 950 1188 1425 1900

1125 1500 1875 2250 1050 1400 1750 2100 984 1313 1641 1969 926 1235 1544 1853 875 1167 1458 1750 750 1000 1250 1500 2000

217

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 217 Friday, April 1, 2005 3:29 PM

30.00

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

PIPE ANALYSIS

28.00

Pipe Material API 5L X46 Diameter in. 4.5

6.625

8.625

10.75

12.75

16.00

Copyright 2005 by Taylor & Francis Group, LLC

0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500 0.250 0.312

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09 42.05 52.27

46000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 3489 4961 6433 7816 2500 2800 4319 5619 1920 2127 2473 3118 1540 1892 2249 3081 1299 1714 1948 2109 2598 1183 1476 1774 2068 2366 1035 1292

2907 4134 5361 6514 2083 2333 3599 4683 1600 1773 2061 2598 1284 1576 1874 2567 1082 1429 1624 1758 2165 986 1230 1479 1723 1971 863 1076

2423 3445 4467 5428 1736 1944 3000 3902 1333 1477 1717 2165 1070 1314 1562 2140 902 1191 1353 1465 1804 821 1025 1232 1436 1643 719 897

1938 2756 3574 4342 1389 1555 2400 3122 1067 1182 1374 1732 856 1051 1249 1712 722 952 1082 1172 1443 657 820 986 1149 1314 575 718

4361 6201 8041 9770 3125 3499 5399 7024 2400 2659 3091 3898 1926 2365 2811 3851 1624 2143 2435 2637 3247 1479 1845 2218 2585 2957 1294 1615

4603 6545 8488 10313 3298 3694 5699 7414 2533 2807 3263 4114 2033 2496 2968 4065 1714 2262 2571 2783 3427 1561 1948 2341 2728 3121 1366 1704

4845 6890 8934 10856 3472 3888 5999 7804 2667 2955 3435 4331 2140 2627 3124 4279 1804 2381 2706 2930 3608 1643 2050 2464 2872 3286 1438 1794

GAS PIPELINE HYDRAULICS

14.00

Wall Thickness in.

2785_C006.fm Page 218 Friday, April 1, 2005 3:29 PM

218

Table 6.6 Pipeline Internal Design Pressures and Test Pressures

24.00

26.00

28.00

30.00

Copyright 2005 by Taylor & Francis Group, LLC

1294 1508 1725 767 957 1150 1340 1533 861 1035 1206 1380 1551 941 1255 1568 1882 863 1005 1150 1293 1438 1725 796 1062 1327 1592 739 986 1232 1479 690

1078 1256 1438 639 797 958 1117 1278 718 863 1005 1150 1293 784 1045 1307 1568 719 838 958 1077 1198 1438 663 885 1106 1327 616 821 1027 1232 575

863 1005 1150 511 638 767 893 1022 574 690 804 920 1034 627 836 1045 1255 575 670 767 862 958 1150 531 708 885 1062 493 657 821 986 460

1941 2261 2588 1150 1435 1725 2010 2300 1292 1553 1809 2070 2327 1411 1882 2352 2823 1294 1508 1725 1939 2156 2588 1194 1592 1990 2388 1109 1479 1848 2218 1035

2048 2387 2731 1214 1515 1821 2122 2428 1363 1639 1910 2185 2456 1490 1986 2483 2980 1366 1591 1821 2047 2276 2731 1261 1681 2101 2521 1171 1561 1951 2341 1093

2156 2513 2875 1278 1595 1917 2234 2556 1435 1725 2010 2300 2585 1568 2091 2614 3136 1438 1675 1917 2154 2396 2875 1327 1769 2212 2654 1232 1643 2054 2464 1150 (continued )

2785_C006.fm Page 219 Friday, April 1, 2005 3:29 PM

22.00

1553 1809 2070 920 1148 1380 1608 1840 1033 1242 1447 1656 1861 1129 1505 1882 2258 1035 1206 1380 1551 1725 2070 955 1274 1592 1911 887 1183 1479 1774 828

219

20.00

62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25 110.64 146.85 182.73 218.27 118.65

PIPE ANALYSIS

18.00

0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375

Pipe Material API 5L X46 Diameter in.

32.00

34.00

36.00

42.00

Weight lb/ft

0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

46000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 1104 1380 1656 776 1035 1294 1553 731 974 1218 1461 690 920 1150 1380 591 789 986 1183 1577

920 1150 1380 647 863 1078 1294 609 812 1015 1218 575 767 958 1150 493 657 821 986 1314

767 958 1150 539 719 898 1078 507 676 846 1015 479 639 799 958 411 548 685 821 1095

613 767 920 431 575 719 863 406 541 676 812 383 511 639 767 329 438 548 657 876

1380 1725 2070 970 1294 1617 1941 913 1218 1522 1826 863 1150 1438 1725 739 986 1232 1479 1971

1457 1821 2185 1024 1366 1707 2048 964 1285 1607 1928 910 1214 1517 1821 780 1040 1301 1561 2081

1533 1917 2300 1078 1438 1797 2156 1015 1353 1691 2029 958 1278 1597 1917 821 1095 1369 1643 2190

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

SMYS

Wall Thickness in.

2785_C006.fm Page 220 Friday, April 1, 2005 3:29 PM

220

Table 6.6 Pipeline Internal Design Pressures and Test Pressures (Continued)

4.5

6.625

8.625

10.75

12.75

14.00

Wall Thickness in. 0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09

52000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 3944 5608 7272 8836 2826 3165 4883 6352 2170 2405 2796 3525 1741 2138 2542 3483 1468 1938 2202 2384 2936 1337 1669 2006 2337 2674

3286 4673 6060 7363 2355 2637 4069 5293 1809 2004 2330 2937 1451 1782 2119 2902 1224 1615 1835 1987 2447 1114 1391 1671 1948 2229

2739 3894 5050 6136 1962 2198 3391 4411 1507 1670 1941 2448 1209 1485 1766 2419 1020 1346 1529 1656 2039 929 1159 1393 1623 1857

2191 3115 4040 4909 1570 1758 2713 3529 1206 1336 1553 1958 967 1188 1412 1935 816 1077 1224 1325 1631 743 927 1114 1299 1486

4930 7010 9090 11045 3532 3956 6103 7940 2713 3006 3494 4406 2177 2673 3178 4353 1835 2423 2753 2981 3671 1671 2086 2507 2922 3343

5203 7399 9595 11658 3728 4176 6443 8381 2864 3173 3689 4651 2298 2822 3355 4595 1937 2557 2906 3146 3875 1764 2202 2646 3084 3529

5477 7788 10100 12272 3925 4395 6782 8822 3014 3340 3883 4896 2419 2970 3531 4837 2039 2692 3059 3312 4078 1857 2318 2786 3246 3714

Copyright 2005 by Taylor & Francis Group, LLC

221

(continued )

2785_C006.fm Page 221 Friday, April 1, 2005 3:29 PM

Pipe Material API 5L X52 Diameter in.

PIPE ANALYSIS

Table 6.7 Pipeline Internal Design Pressures and Test Pressures

Pipe Material API 5L X52 Diameter in. 16.00

18.00

20.00

22.00

26.00

Copyright 2005 by Taylor & Francis Group, LLC

Weight lb/ft

0.250 0.312 0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750

42.05 52.27 62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25

52000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 1170 1460 1755 2045 2340 1040 1298 1560 1818 2080 1168 1404 1636 1872 2104 1276 1702 2127 2553 1170 1363 1560 1753 1950 2340 1080 1440 1800 2160

975 1217 1463 1704 1950 867 1082 1300 1515 1733 973 1170 1363 1560 1753 1064 1418 1773 2127 975 1136 1300 1461 1625 1950 900 1200 1500 1800

813 1014 1219 1420 1625 722 901 1083 1262 1444 811 975 1136 1300 1461 886 1182 1477 1773 813 947 1083 1218 1354 1625 750 1000 1250 1500

650 811 975 1136 1300 578 721 867 1010 1156 649 780 909 1040 1169 709 945 1182 1418 650 757 867 974 1083 1300 600 800 1000 1200

1463 1825 2194 2556 2925 1300 1622 1950 2272 2600 1460 1755 2045 2340 2630 1595 2127 2659 3191 1463 1704 1950 2192 2438 2925 1350 1800 2250 2700

1544 1927 2316 2698 3088 1372 1713 2058 2399 2744 1541 1853 2159 2470 2776 1684 2245 2807 3368 1544 1799 2058 2314 2573 3088 1425 1900 2375 2850

1625 2028 2438 2841 3250 1444 1803 2167 2525 2889 1622 1950 2272 2600 2922 1773 2364 2955 3545 1625 1894 2167 2435 2708 3250 1500 2000 2500 3000

GAS PIPELINE HYDRAULICS

24.00

SMYS

Wall Thickness in.

2785_C006.fm Page 222 Friday, April 1, 2005 3:29 PM

222

Table 6.7 Pipeline Internal Design Pressures and Test Pressures (Continued)

32.00

34.00

36.00

42.00

110.64 146.85 182.73 218.27 118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

1003 1337 1671 2006 936 1248 1560 1872 878 1170 1463 1755 826 1101 1376 1652 780 1040 1300 1560 669 891 1114 1337 1783

836 1114 1393 1671 780 1040 1300 1560 731 975 1219 1463 688 918 1147 1376 650 867 1083 1300 557 743 929 1114 1486

696 929 1161 1393 650 867 1083 1300 609 813 1016 1219 574 765 956 1147 542 722 903 1083 464 619 774 929 1238

557 743 929 1114 520 693 867 1040 488 650 813 975 459 612 765 918 433 578 722 867 371 495 619 743 990

1254 1671 2089 2507 1170 1560 1950 2340 1097 1463 1828 2194 1032 1376 1721 2065 975 1300 1625 1950 836 1114 1393 1671 2229

1323 1764 2205 2646 1235 1647 2058 2470 1158 1544 1930 2316 1090 1453 1816 2179 1029 1372 1715 2058 882 1176 1470 1764 2352

1393 1857 2321 2786 1300 1733 2167 2600 1219 1625 2031 2438 1147 1529 1912 2294 1083 1444 1806 2167 929 1238 1548 1857 2476

223

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 223 Friday, April 1, 2005 3:29 PM

30.00

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

PIPE ANALYSIS

28.00

Pipe Material API 5L X56 Diameter in. 4.5

6.625

8.625

10.75

12.75

16.00

Copyright 2005 by Taylor & Francis Group, LLC

0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500 0.250 0.312

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09 42.05 52.27

56000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 4247 6039 7831 9516 3043 3408 5258 6841 2337 2590 3011 3796 1875 2303 2738 3751 1581 2087 2372 2568 3162 1440 1797 2160 2517 2880 1260 1572

3539 5033 6526 7930 2536 2840 4382 5701 1948 2158 2509 3163 1563 1919 2282 3126 1318 1739 1976 2140 2635 1200 1498 1800 2098 2400 1050 1310

2949 4194 5438 6608 2113 2367 3652 4750 1623 1798 2091 2636 1302 1599 1901 2605 1098 1449 1647 1783 2196 1000 1248 1500 1748 2000 875 1092

2359 3355 4351 5286 1691 1893 2921 3800 1299 1439 1673 2109 1042 1279 1521 2084 878 1160 1318 1427 1757 800 998 1200 1398 1600 700 874

5309 7549 9789 11894 3804 4260 6573 8551 2922 3237 3763 4745 2344 2879 3423 4688 1976 2609 2965 3210 3953 1800 2246 2700 3146 3600 1575 1966

5604 7968 10333 12555 4015 4497 6938 9026 3084 3417 3972 5009 2474 3039 3613 4949 2086 2754 3129 3388 4173 1900 2371 2850 3321 3800 1663 2075

5899 8388 10876 13216 4226 4734 7303 9501 3246 3597 4181 5272 2605 3199 3803 5209 2196 2899 3294 3566 4392 2000 2496 3000 3496 4000 1750 2184

GAS PIPELINE HYDRAULICS

14.00

Wall Thickness in.

2785_C006.fm Page 224 Friday, April 1, 2005 3:29 PM

224

Table 6.8 Pipeline Internal Design Pressures and Test Pressures

22.00

24.00

26.00

28.00

1890 2202 2520 1120 1398 1680 1958 2240 1258 1512 1762 2016 2266 1375 1833 2291 2749 1260 1468 1680 1888 2100 2520 1163 1551 1938 2326 1080 1440 1800 2160

1575 1835 2100 933 1165 1400 1631 1867 1048 1260 1468 1680 1888 1145 1527 1909 2291 1050 1224 1400 1574 1750 2100 969 1292 1615 1938 900 1200 1500 1800

1313 1530 1750 778 971 1167 1360 1556 874 1050 1224 1400 1574 955 1273 1591 1909 875 1020 1167 1311 1458 1750 808 1077 1346 1615 750 1000 1250 1500

1050 1224 1400 622 777 933 1088 1244 699 840 979 1120 1259 764 1018 1273 1527 700 816 933 1049 1167 1400 646 862 1077 1292 600 800 1000 1200

2363 2753 3150 1400 1747 2100 2447 2800 1572 1890 2202 2520 2832 1718 2291 2864 3436 1575 1835 2100 2360 2625 3150 1454 1938 2423 2908 1350 1800 2250 2700

2494 2906 3325 1478 1844 2217 2583 2956 1660 1995 2325 2660 2990 1814 2418 3023 3627 1663 1937 2217 2492 2771 3325 1535 2046 2558 3069 1425 1900 2375 2850

2625 3059 3500 1556 1941 2333 2719 3111 1747 2100 2447 2800 3147 1909 2545 3182 3818 1750 2039 2333 2623 2917 3500 1615 2154 2692 3231 1500 2000 2500 3000

Copyright 2005 by Taylor & Francis Group, LLC

225

(continued )

2785_C006.fm Page 225 Friday, April 1, 2005 3:29 PM

20.00

62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25 110.64 146.85 182.73 218.27

PIPE ANALYSIS

18.00

0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750

Pipe Material API 5L X56 Diameter in. 30.00

32.00

34.00

36.00

42.00

Weight lb/ft

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

56000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 1008 1344 1680 2016 945 1260 1575 1890 889 1186 1482 1779 840 1120 1400 1680 720 960 1200 1440 1920

840 1120 1400 1680 788 1050 1313 1575 741 988 1235 1482 700 933 1167 1400 600 800 1000 1200 1600

700 933 1167 1400 656 875 1094 1313 618 824 1029 1235 583 778 972 1167 500 667 833 1000 1333

560 747 933 1120 525 700 875 1050 494 659 824 988 467 622 778 933 400 533 667 800 1067

1260 1680 2100 2520 1181 1575 1969 2363 1112 1482 1853 2224 1050 1400 1750 2100 900 1200 1500 1800 2400

1330 1773 2217 2660 1247 1663 2078 2494 1174 1565 1956 2347 1108 1478 1847 2217 950 1267 1583 1900 2533

1400 1867 2333 2800 1313 1750 2188 2625 1235 1647 2059 2471 1167 1556 1944 2333 1000 1333 1667 2000 2667

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

SMYS

Wall Thickness in.

2785_C006.fm Page 226 Friday, April 1, 2005 3:29 PM

226

Table 6.8 Pipeline Internal Design Pressures and Test Pressures (Continued)

4.5

6.625

8.625

10.75

12.75

14.00

Wall Thickness in. 0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09

60000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 4550 6470 8390 10195 3260 3652 5634 7329 2504 2775 3226 4067 2009 2467 2934 4019 1694 2236 2541 2751 3388 1543 1925 2314 2697 3086

3792 5392 6992 8496 2717 3043 4695 6108 2087 2312 2688 3389 1674 2056 2445 3349 1412 1864 2118 2293 2824 1286 1605 1929 2247 2571

3160 4493 5827 7080 2264 2536 3912 5090 1739 1927 2240 2824 1395 1713 2037 2791 1176 1553 1765 1911 2353 1071 1337 1607 1873 2143

2528 3595 4661 5664 1811 2029 3130 4072 1391 1542 1792 2259 1116 1371 1630 2233 941 1242 1412 1528 1882 857 1070 1286 1498 1714

5688 8088 10488 12744 4075 4565 7042 9162 3130 3469 4032 5084 2512 3084 3667 5023 2118 2795 3176 3439 4235 1929 2407 2893 3371 3857

6004 8537 11071 13452 4302 4818 7434 9671 3304 3661 4256 5366 2651 3256 3871 5302 2235 2951 3353 3630 4471 2036 2541 3054 3558 4071

6320 8987 11653 14160 4528 5072 7825 10180 3478 3854 4480 5649 2791 3427 4074 5581 2353 3106 3529 3821 4706 2143 2674 3214 3746 4286

Copyright 2005 by Taylor & Francis Group, LLC

227

(continued )

2785_C006.fm Page 227 Friday, April 1, 2005 3:29 PM

Pipe Material API 5L X60 Diameter in.

PIPE ANALYSIS

Table 6.9 Pipeline Internal Design Pressures and Test Pressures

Pipe Material API 5L X60 Diameter in. 16.00

18.00

20.00

22.00

26.00

Copyright 2005 by Taylor & Francis Group, LLC

Weight lb/ft

0.250 0.312 0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750

42.05 52.27 62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25

60000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 1350 1685 2025 2360 2700 1200 1498 1800 2098 2400 1348 1620 1888 2160 2428 1473 1964 2455 2945 1350 1573 1800 2023 2250 2700 1246 1662 2077 2492

1125 1404 1688 1967 2250 1000 1248 1500 1748 2000 1123 1350 1573 1800 2023 1227 1636 2045 2455 1125 1311 1500 1686 1875 2250 1038 1385 1731 2077

938 1170 1406 1639 1875 833 1040 1250 1457 1667 936 1125 1311 1500 1686 1023 1364 1705 2045 938 1093 1250 1405 1563 1875 865 1154 1442 1731

750 936 1125 1311 1500 667 832 1000 1165 1333 749 900 1049 1200 1349 818 1091 1364 1636 750 874 1000 1124 1250 1500 692 923 1154 1385

1688 2106 2531 2950 3375 1500 1872 2250 2622 3000 1685 2025 2360 2700 3035 1841 2455 3068 3682 1688 1967 2250 2529 2813 3375 1558 2077 2596 3115

1781 2223 2672 3114 3563 1583 1976 2375 2768 3167 1778 2138 2491 2850 3203 1943 2591 3239 3886 1781 2076 2375 2670 2969 3563 1644 2192 2740 3288

1875 2340 2813 3278 3750 1667 2080 2500 2913 3333 1872 2250 2622 3000 3372 2045 2727 3409 4091 1875 2185 2500 2810 3125 3750 1731 2308 2885 3462

GAS PIPELINE HYDRAULICS

24.00

SMYS

Wall Thickness in.

2785_C006.fm Page 228 Friday, April 1, 2005 3:29 PM

228

Table 6.9 Pipeline Internal Design Pressures and Test Pressures (Continued)

32.00

34.00

36.00

42.00

110.64 146.85 182.73 218.27 118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

1157 1543 1929 2314 1080 1440 1800 2160 1013 1350 1688 2025 953 1271 1588 1906 900 1200 1500 1800 771 1029 1286 1543 2057

964 1286 1607 1929 900 1200 1500 1800 844 1125 1406 1688 794 1059 1324 1588 750 1000 1250 1500 643 857 1071 1286 1714

804 1071 1339 1607 750 1000 1250 1500 703 938 1172 1406 662 882 1103 1324 625 833 1042 1250 536 714 893 1071 1429

643 857 1071 1286 600 800 1000 1200 563 750 938 1125 529 706 882 1059 500 667 833 1000 429 571 714 857 1143

1446 1929 2411 2893 1350 1800 2250 2700 1266 1688 2109 2531 1191 1588 1985 2382 1125 1500 1875 2250 964 1286 1607 1929 2571

1527 2036 2545 3054 1425 1900 2375 2850 1336 1781 2227 2672 1257 1676 2096 2515 1188 1583 1979 2375 1018 1357 1696 2036 2714

1607 2143 2679 3214 1500 2000 2500 3000 1406 1875 2344 2813 1324 1765 2206 2647 1250 1667 2083 2500 1071 1429 1786 2143 2857

229

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 229 Friday, April 1, 2005 3:29 PM

30.00

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

PIPE ANALYSIS

28.00

4.5

6.625

8.625

10.75

12.75

16.00

Copyright 2005 by Taylor & Francis Group, LLC

0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500 0.250 0.312

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09 42.05 52.27

65000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 4930 7010 9090 11045 3532 3956 6103 7940 2713 3006 3494 4406 2177 2673 3178 4353 1835 2423 2753 2981 3671 1671 2086 2507 2922 3343 1463 1825

4108 5841 7575 9204 2943 3297 5086 6617 2261 2505 2912 3672 1814 2228 2648 3628 1529 2019 2294 2484 3059 1393 1738 2089 2435 2786 1219 1521

3423 4868 6312 7670 2453 2747 4238 5514 1884 2088 2427 3060 1512 1856 2207 3023 1275 1682 1912 2070 2549 1161 1449 1741 2029 2321 1016 1268

2739 3894 5050 6136 1962 2198 3391 4411 1507 1670 1941 2448 1209 1485 1766 2419 1020 1346 1529 1656 2039 929 1159 1393 1623 1857 813 1014

6162 8762 11362 13806 4415 4945 7629 9925 3391 3758 4368 5507 2721 3341 3973 5442 2294 3028 3441 3726 4588 2089 2607 3134 3652 4179 1828 2282

6504 9249 11993 14573 4660 5220 8053 10477 3580 3966 4611 5813 2872 3527 4193 5744 2422 3196 3632 3933 4843 2205 2752 3308 3855 4411 1930 2408

6847 9736 12624 15340 4906 5494 8477 11028 3768 4175 4853 6119 3023 3713 4414 6047 2549 3365 3824 4140 5098 2321 2897 3482 4058 4643 2031 2535

GAS PIPELINE HYDRAULICS

14.00

Wall Thickness in.

2785_C006.fm Page 230 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures

Pipe Material API 5L X65 Diameter in.

230

Table 6.10

22.00

24.00

26.00

28.00

2194 2556 2925 1300 1622 1950 2272 2600 1460 1755 2045 2340 2630 1595 2127 2659 3191 1463 1704 1950 2192 2438 2925 1350 1800 2250 2700 1254 1671 2089 2507

1828 2130 2438 1083 1352 1625 1894 2167 1217 1463 1704 1950 2192 1330 1773 2216 2659 1219 1420 1625 1827 2031 2438 1125 1500 1875 2250 1045 1393 1741 2089

1523 1775 2031 903 1127 1354 1578 1806 1014 1219 1420 1625 1827 1108 1477 1847 2216 1016 1184 1354 1522 1693 2031 938 1250 1563 1875 871 1161 1451 1741

1219 1420 1625 722 901 1083 1262 1444 811 975 1136 1300 1461 886 1182 1477 1773 813 947 1083 1218 1354 1625 750 1000 1250 1500 696 929 1161 1393

2742 3196 3656 1625 2028 2438 2841 3250 1825 2194 2556 2925 3288 1994 2659 3324 3989 1828 2130 2438 2740 3047 3656 1688 2250 2813 3375 1567 2089 2612 3134

2895 3373 3859 1715 2141 2573 2998 3431 1927 2316 2698 3088 3470 2105 2807 3509 4210 1930 2249 2573 2892 3216 3859 1781 2375 2969 3563 1654 2205 2757 3308

3047 3551 4063 1806 2253 2708 3156 3611 2028 2438 2841 3250 3653 2216 2955 3693 4432 2031 2367 2708 3044 3385 4063 1875 2500 3125 3750 1741 2321 2902 3482

Copyright 2005 by Taylor & Francis Group, LLC

231

(continued )

2785_C006.fm Page 231 Friday, April 1, 2005 3:29 PM

20.00

62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25 110.64 146.85 182.73 218.27

PIPE ANALYSIS

18.00

0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750

30.00

32.00

34.00

36.00

42.00

Weight lb/ft

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

65000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 1170 1560 1950 2340 1097 1463 1828 2194 1032 1376 1721 2065 975 1300 1625 1950 836 1114 1393 1671 2229

975 1300 1625 1950 914 1219 1523 1828 860 1147 1434 1721 813 1083 1354 1625 696 929 1161 1393 1857

813 1083 1354 1625 762 1016 1270 1523 717 956 1195 1434 677 903 1128 1354 580 774 967 1161 1548

650 867 1083 1300 609 813 1016 1219 574 765 956 1147 542 722 903 1083 464 619 774 929 1238

1463 1950 2438 2925 1371 1828 2285 2742 1290 1721 2151 2581 1219 1625 2031 2438 1045 1393 1741 2089 2786

1544 2058 2573 3088 1447 1930 2412 2895 1362 1816 2270 2724 1286 1715 2144 2573 1103 1470 1838 2205 2940

1625 2167 2708 3250 1523 2031 2539 3047 1434 1912 2390 2868 1354 1806 2257 2708 1161 1548 1935 2321 3095

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

SMYS

Wall Thickness in.

2785_C006.fm Page 232 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures (Continued)

Pipe Material API 5L X65 Diameter in.

232

Table 6.10

Pipe Material API 5L X70 Diameter in. 4.5

6.625

8.625

10.75

12.75

14.00

Wall Thickness in. 0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09

70000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 5309 7549 9789 11894 3804 4260 6573 8551 2922 3237 3763 4745 2344 2879 3423 4688 1976 2609 2965 3210 3953 1800 2246 2700 3146 3600

4424 6291 8157 9912 3170 3550 5477 7126 2435 2698 3136 3954 1953 2399 2852 3907 1647 2174 2471 2675 3294 1500 1872 2250 2622 3000

3687 5242 6798 8260 2642 2958 4565 5938 2029 2248 2613 3295 1628 1999 2377 3256 1373 1812 2059 2229 2745 1250 1560 1875 2185 2500

2949 4194 5438 6608 2113 2367 3652 4750 1623 1798 2091 2636 1302 1599 1901 2605 1098 1449 1647 1783 2196 1000 1248 1500 1748 2000

6636 9436 12236 14868 4755 5325 8216 10689 3652 4047 4704 5931 2930 3598 4278 5860 2471 3261 3706 4012 4941 2250 2808 3375 3933 4500

7005 9960 12916 15694 5019 5621 8673 11282 3855 4271 4965 6261 3093 3798 4516 6186 2608 3442 3912 4235 5216 2375 2964 3563 4152 4750

7373 10484 13596 16520 5283 5917 9129 11876 4058 4496 5227 6590 3256 3998 4753 6512 2745 3624 4118 4458 5490 2500 3120 3750 4370 5000

Copyright 2005 by Taylor & Francis Group, LLC

233

(continued )

2785_C006.fm Page 233 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures

PIPE ANALYSIS

Table 6.11

16.00

18.00

20.00

22.00

26.00

Copyright 2005 by Taylor & Francis Group, LLC

Weight lb/ft

0.250 0.312 0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750

42.05 52.27 62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25

70000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 1575 1966 2363 2753 3150 1400 1747 2100 2447 2800 1572 1890 2202 2520 2832 1718 2291 2864 3436 1575 1835 2100 2360 2625 3150 1454 1938 2423 2908

1313 1638 1969 2294 2625 1167 1456 1750 2039 2333 1310 1575 1835 2100 2360 1432 1909 2386 2864 1313 1530 1750 1967 2188 2625 1212 1615 2019 2423

1094 1365 1641 1912 2188 972 1213 1458 1699 1944 1092 1313 1530 1750 1967 1193 1591 1989 2386 1094 1275 1458 1639 1823 2188 1010 1346 1683 2019

875 1092 1313 1530 1750 778 971 1167 1360 1556 874 1050 1224 1400 1574 955 1273 1591 1909 875 1020 1167 1311 1458 1750 808 1077 1346 1615

1969 2457 2953 3441 3938 1750 2184 2625 3059 3500 1966 2363 2753 3150 3541 2148 2864 3580 4295 1969 2294 2625 2951 3281 3938 1817 2423 3029 3635

2078 2594 3117 3633 4156 1847 2305 2771 3229 3694 2075 2494 2906 3325 3737 2267 3023 3778 4534 2078 2422 2771 3114 3464 4156 1918 2558 3197 3837

2188 2730 3281 3824 4375 1944 2427 2917 3399 3889 2184 2625 3059 3500 3934 2386 3182 3977 4773 2188 2549 2917 3278 3646 4375 2019 2692 3365 4038

GAS PIPELINE HYDRAULICS

24.00

SMYS

Wall Thickness in.

2785_C006.fm Page 234 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures (Continued)

Pipe Material API 5L X70 Diameter in.

234

Table 6.11

32.00

34.00

36.00

42.00

110.64 146.85 182.73 218.27 118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

1350 1800 2250 2700 1260 1680 2100 2520 1181 1575 1969 2363 1112 1482 1853 2224 1050 1400 1750 2100 900 1200 1500 1800 2400

1125 1500 1875 2250 1050 1400 1750 2100 984 1313 1641 1969 926 1235 1544 1853 875 1167 1458 1750 750 1000 1250 1500 2000

938 1250 1563 1875 875 1167 1458 1750 820 1094 1367 1641 772 1029 1287 1544 729 972 1215 1458 625 833 1042 1250 1667

750 1000 1250 1500 700 933 1167 1400 656 875 1094 1313 618 824 1029 1235 583 778 972 1167 500 667 833 1000 1333

1688 2250 2813 3375 1575 2100 2625 3150 1477 1969 2461 2953 1390 1853 2316 2779 1313 1750 2188 2625 1125 1500 1875 2250 3000

1781 2375 2969 3563 1663 2217 2771 3325 1559 2078 2598 3117 1467 1956 2445 2934 1385 1847 2309 2771 1188 1583 1979 2375 3167

1875 2500 3125 3750 1750 2333 2917 3500 1641 2188 2734 3281 1544 2059 2574 3088 1458 1944 2431 2917 1250 1667 2083 2500 3333

235

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 235 Friday, April 1, 2005 3:29 PM

30.00

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

PIPE ANALYSIS

28.00

4.5

6.625

8.625

10.75

12.75

16.00

Copyright 2005 by Taylor & Francis Group, LLC

0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500 0.250 0.312

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09 42.05 52.27

80000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 6067 8627 11187 13594 4347 4869 7512 9772 3339 3700 4301 5423 2679 3290 3911 5358 2259 2982 3388 3668 4518 2057 2567 3086 3596 4114 1800 2246

5056 7189 9323 11328 3623 4057 6260 8144 2783 3083 3584 4519 2233 2742 3260 4465 1882 2485 2824 3057 3765 1714 2139 2571 2997 3429 1500 1872

4213 5991 7769 9440 3019 3381 5217 6786 2319 2569 2987 3766 1860 2285 2716 3721 1569 2071 2353 2547 3137 1429 1783 2143 2497 2857 1250 1560

3371 4793 6215 7552 2415 2705 4173 5429 1855 2055 2389 3013 1488 1828 2173 2977 1255 1656 1882 2038 2510 1143 1426 1714 1998 2286 1000 1248

7584 10784 13984 16992 5434 6086 9390 12216 4174 4625 5376 6778 3349 4112 4889 6698 2824 3727 4235 4585 5647 2571 3209 3857 4495 5143 2250 2808

8005 11383 14761 17936 5736 6424 9912 12894 4406 4882 5675 7155 3535 4341 5161 7070 2980 3934 4471 4840 5961 2714 3387 4071 4745 5429 2375 2964

8427 11982 15538 18880 6038 6762 10433 13573 4638 5139 5973 7532 3721 4569 5433 7442 3137 4141 4706 5095 6275 2857 3566 4286 4994 5714 2500 3120

GAS PIPELINE HYDRAULICS

14.00

Wall Thickness in.

2785_C006.fm Page 236 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures

Pipe Material API 5L X80 Diameter in.

236

Table 6.12

22.00

24.00

26.00

28.00

2700 3146 3600 1600 1997 2400 2797 3200 1797 2160 2517 2880 3237 1964 2618 3273 3927 1800 2098 2400 2698 3000 3600 1662 2215 2769 3323 1543 2057 2571 3086

2250 2622 3000 1333 1664 2000 2331 2667 1498 1800 2098 2400 2698 1636 2182 2727 3273 1500 1748 2000 2248 2500 3000 1385 1846 2308 2769 1286 1714 2143 2571

1875 2185 2500 1111 1387 1667 1942 2222 1248 1500 1748 2000 2248 1364 1818 2273 2727 1250 1457 1667 1873 2083 2500 1154 1538 1923 2308 1071 1429 1786 2143

1500 1748 2000 889 1109 1333 1554 1778 998 1200 1398 1600 1798 1091 1455 1818 2182 1000 1165 1333 1499 1667 2000 923 1231 1538 1846 857 1143 1429 1714

3375 3933 4500 2000 2496 3000 3496 4000 2246 2700 3146 3600 4046 2455 3273 4091 4909 2250 2622 3000 3372 3750 4500 2077 2769 3462 4154 1929 2571 3214 3857

3563 4152 4750 2111 2635 3167 3690 4222 2371 2850 3321 3800 4271 2591 3455 4318 5182 2375 2768 3167 3559 3958 4750 2192 2923 3654 4385 2036 2714 3393 4071

3750 4370 5000 2222 2773 3333 3884 4444 2496 3000 3496 4000 4496 2727 3636 4545 5455 2500 2913 3333 3747 4167 5000 2308 3077 3846 4615 2143 2857 3571 4286

Copyright 2005 by Taylor & Francis Group, LLC

237

(continued )

2785_C006.fm Page 237 Friday, April 1, 2005 3:29 PM

20.00

62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25 110.64 146.85 182.73 218.27

PIPE ANALYSIS

18.00

0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750

30.00

32.00

34.00

36.00

42.00

Weight lb/ft

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

80000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 1440 1920 2400 2880 1350 1800 2250 2700 1271 1694 2118 2541 1200 1600 2000 2400 1029 1371 1714 2057 2743

1200 1600 2000 2400 1125 1500 1875 2250 1059 1412 1765 2118 1000 1333 1667 2000 857 1143 1429 1714 2286

1000 1333 1667 2000 938 1250 1563 1875 882 1176 1471 1765 833 1111 1389 1667 714 952 1190 1429 1905

800 1067 1333 1600 750 1000 1250 1500 706 941 1176 1412 667 889 1111 1333 571 762 952 1143 1524

1800 2400 3000 3600 1688 2250 2813 3375 1588 2118 2647 3176 1500 2000 2500 3000 1286 1714 2143 2571 3429

1900 2533 3167 3800 1781 2375 2969 3563 1676 2235 2794 3353 1583 2111 2639 3167 1357 1810 2262 2714 3619

2000 2667 3333 4000 1875 2500 3125 3750 1765 2353 2941 3529 1667 2222 2778 3333 1429 1905 2381 2857 3810

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

SMYS

Wall Thickness in.

2785_C006.fm Page 238 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures (Continued)

Pipe Material API 5L X80 Diameter in.

238

Table 6.12

Pipe Material API 5L X90 Diameter in. 4.5

6.625

8.625

10.75

12.75

14.00

Wall Thickness in. 0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09

90000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 6826 9706 12586 15293 4891 5477 8451 10994 3757 4162 4838 6101 3014 3701 4400 6028 2541 3354 3812 4127 5082 2314 2888 3471 4045 4629

5688 8088 10488 12744 4075 4565 7042 9162 3130 3469 4032 5084 2512 3084 3667 5023 2118 2795 3176 3439 4235 1929 2407 2893 3371 3857

4740 6740 8740 10620 3396 3804 5869 7635 2609 2890 3360 4237 2093 2570 3056 4186 1765 2329 2647 2866 3529 1607 2006 2411 2809 3214

3792 5392 6992 8496 2717 3043 4695 6108 2087 2312 2688 3389 1674 2056 2445 3349 1412 1864 2118 2293 2824 1286 1605 1929 2247 2571

8532 12132 15732 19116 6113 6847 10564 13742 4696 5203 6048 7626 3767 4626 5500 7535 3176 4193 4765 5159 6353 2893 3610 4339 5057 5786

9006 12806 16606 20178 6453 7227 11150 14506 4957 5492 6384 8049 3977 4883 5806 7953 3353 4426 5029 5445 6706 3054 3811 4580 5338 6107

9480 13480 17480 21240 6792 7608 11737 15269 5217 5781 6720 8473 4186 5140 6112 8372 3529 4659 5294 5732 7059 3214 4011 4821 5619 6429

Copyright 2005 by Taylor & Francis Group, LLC

239

(continued )

2785_C006.fm Page 239 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures

PIPE ANALYSIS

Table 6.13

16.00

18.00

20.00

22.00

26.00

Copyright 2005 by Taylor & Francis Group, LLC

Weight lb/ft

0.250 0.312 0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750

42.05 52.27 62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25

90000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 2025 2527 3038 3540 4050 1800 2246 2700 3146 3600 2022 2430 2832 3240 3642 2209 2945 3682 4418 2025 2360 2700 3035 3375 4050 1869 2492 3115 3738

1688 2106 2531 2950 3375 1500 1872 2250 2622 3000 1685 2025 2360 2700 3035 1841 2455 3068 3682 1688 1967 2250 2529 2813 3375 1558 2077 2596 3115

1406 1755 2109 2458 2813 1250 1560 1875 2185 2500 1404 1688 1967 2250 2529 1534 2045 2557 3068 1406 1639 1875 2108 2344 2813 1298 1731 2163 2596

1125 1404 1688 1967 2250 1000 1248 1500 1748 2000 1123 1350 1573 1800 2023 1227 1636 2045 2455 1125 1311 1500 1686 1875 2250 1038 1385 1731 2077

2531 3159 3797 4425 5063 2250 2808 3375 3933 4500 2527 3038 3540 4050 4552 2761 3682 4602 5523 2531 2950 3375 3794 4219 5063 2337 3115 3894 4673

2672 3335 4008 4670 5344 2375 2964 3563 4152 4750 2668 3206 3736 4275 4805 2915 3886 4858 5830 2672 3114 3563 4004 4453 5344 2466 3288 4111 4933

2813 3510 4219 4916 5625 2500 3120 3750 4370 5000 2808 3375 3933 4500 5058 3068 4091 5114 6136 2813 3278 3750 4215 4688 5625 2596 3462 4327 5192

GAS PIPELINE HYDRAULICS

24.00

SMYS

Wall Thickness in.

2785_C006.fm Page 240 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures (Continued)

Pipe Material API 5L X90 Diameter in.

240

Table 6.13

32.00

34.00

36.00

42.00

110.64 146.85 182.73 218.27 118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

1736 2314 2893 3471 1620 2160 2700 3240 1519 2025 2531 3038 1429 1906 2382 2859 1350 1800 2250 2700 1157 1543 1929 2314 3086

1446 1929 2411 2893 1350 1800 2250 2700 1266 1688 2109 2531 1191 1588 1985 2382 1125 1500 1875 2250 964 1286 1607 1929 2571

1205 1607 2009 2411 1125 1500 1875 2250 1055 1406 1758 2109 993 1324 1654 1985 938 1250 1563 1875 804 1071 1339 1607 2143

964 1286 1607 1929 900 1200 1500 1800 844 1125 1406 1688 794 1059 1324 1588 750 1000 1250 1500 643 857 1071 1286 1714

2170 2893 3616 4339 2025 2700 3375 4050 1898 2531 3164 3797 1787 2382 2978 3574 1688 2250 2813 3375 1446 1929 2411 2893 3857

2290 3054 3817 4580 2138 2850 3563 4275 2004 2672 3340 4008 1886 2515 3143 3772 1781 2375 2969 3563 1527 2036 2545 3054 4071

2411 3214 4018 4821 2250 3000 3750 4500 2109 2813 3516 4219 1985 2647 3309 3971 1875 2500 3125 3750 1607 2143 2679 3214 4286

241

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 241 Friday, April 1, 2005 3:29 PM

30.00

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

PIPE ANALYSIS

28.00

2785_C006.fm Page 242 Friday, April 1, 2005 3:29 PM

242

GAS PIPELINE HYDRAULICS

For class 1, the range of hydrotest pressures is (1.25 × 1872) psig to (1.3194 × 1872) psig = 2340 psig to 2470 psig where 1.3194 is equal to the factor 1.25 × 95/90, representing the upper limit of the hydrotest envelope. For class 2, the range of hydrotest pressures is (1.25 × 1560) psig to (1.3194 × 1560) psig = 1950 psig to 2058 psig For class 3, the range of hydrotest pressures is (1.25 × 1300) psig to (1.3194 × 1300) psig = 1625 psig to 1715 psig For class 4, the range of hydrotest pressures is (1.25 × 1040) psig to (1.3194 × 1040) psig = 1300 psig to 1372 psig

6.10 BLOWDOWN CALCULATIONS Blowdown valves and piping systems are installed around the mainline valve in a gas transmission piping system in order to evacuate gas from sections of pipeline in the event of an emergency or for maintenance purposes. The objective of the blowdown assembly is to remove gas from the pipeline once the pipe section is isolated by closing the mainline block valves in a reasonable period of time. The pipe size required to blow down a section of pipe will depend on the gas gravity, pipe diameter, length of pipe section, pressure in the pipeline, and blowdown time. AGA recommends the following equation to estimate the blowdown time: 1

T= where T = P1 = G = D = L = d = Fc =

1

0.0588 P1 3G 2 D 2 LFc d2

blowdown time, min initial pressure, psia gas gravity (air = 1.00) pipe inside diameter, in. length of pipe section, mi inside diameter of blowdown pipe, in. choke factor (as follows)

Choke factor list Ideal nozzle = 1.0 Through gate = 1.6 Regular gate = 1.8 Regular lube plug = 2.0 Venturi lube plug = 3.2

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(6.9)

2785_C006.fm Page 243 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

243

In SI units, 1

1

0.0192 P1 3G 2 D 2 LFc T= d2 where P1 = D = L = d =

(SI units)

(6.10)

initial pressure, kPa pipe inside diameter, mm length of pipe section, km pipe inside diameter of blowdown, mm

Other symbols are as defined before. Example 4 Calculate the blowdown time required for an NPS 6, 0.250 in. wall thickness, blowdown assembly on an NPS 24 pipe, 0.500 in. wall thickness, considering a 5 mi pipe section starting at a pressure of 1000 psia. The gas gravity is 0.6 and choke factor = 1.8. Solution Pipe inside diameter = 24 – 2 × 0.500 = 23 in. Blowdown pipe inside diameter = 6.625 – 2 × 0.250 = 6.125 in. Using Equation 6.9, we get 1

1

0.0588 × (1000) 3 (0.6) 2 (23)2 × 5 × 1.8 T= = 58 min, approximately 6.1252

6.11 DETERMINING PIPE TONNAGE Frequently in pipeline design, we are interested in knowing the amount of pipe used so that we can determine the total cost of pipe. A convenient formula for calculating the weight per unit length of pipe used by pipe vendors is given in Equation 6.11. In USCS units, pipe weight in lb/ft is calculated for a given diameter and wall thickness as follows: w = 10.68 × t × (D – t) where w = pipe weight, lb/ft D = pipe outside diameter, in. t = pipe wall thickness, in.

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(6.11)

2785_C006.fm Page 244 Friday, April 1, 2005 3:29 PM

244

GAS PIPELINE HYDRAULICS

The constant 10.68 in Equation 6.11 includes the density of steel and, therefore, the equation is only applicable to steel pipe. For other pipe material, we can ratio the densities to obtain the pipe weight for nonsteel pipe. In SI units, the pipe weight in kg/m is found from w = 0.0246 × t × (D – t)

(SI units)

(6.12)

where w = pipe weight, kg/m D = pipe outside diameter, mm t = pipe wall thickness, mm Example 5 Calculate the total amount of pipe in a 10 mi pipeline, NPS 20, 0.500 in wall thickness. If pipe costs $700 per ton, determine the total pipeline cost. Solution Using Equation 6.11, the weight per foot of pipe is

w = 10.68 × 0.500 × (20 – 0.500) = 101.46 lb/ft Therefore, the total pipe tonnage in 10 miles of pipe is Tonnage = 101.46 × 5280 × 10/2000 = 2679 tons Total pipeline cost = 2679 × 700 = $1,875,300 Example 6 A 60 km pipeline consists of 20 km of DN 500, 12 mm wall thickness pipe connected to a 40 km length of DN 400, 10 mm wall thickness pipe. What are the total metric tons of pipe? Solution Using Equation 6.12, the weight per meter of DN 500 pipe is

w = 0.0246 × 12 × (500 – 12) = 144.06 kg/m and the weight per meter of DN 400 pipe is

w = 0.0246 × 10 × (400 – 10) = 95.94 kg/m Therefore, the total pipe weight for 20 km of DN 500 pipe and 40 km of DN 400 pipe is Weight = (20 × 144.06) + (40 × 95.94) = 6719 tons Total metric tons = 6719

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 245 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

245

Example 7 Calculate the MOP for NPS 16 pipeline, 0.250 in wall thickness, constructed of API 5LX-52 steel. What minimum wall thickness is required for an internal working pressure of 1440 psi? Use class 2 construction with design factor F = 0.60 and for an operating temperature below 250°F. Solution Using Equation 6.8, the internal design pressure is P=

2 × 0.250 × 52, 000 × 0.60 × 1.0 × 1.0 = 975 psig 16

For an internal working pressure of 1440 psi, the wall thickness required is 1440 =

2 × t × 52, 000 × 0.6 × 1.0 16

Solving for t, we get Wall thickness t = 0.369 in. The nearest standard pipe wall thickness is 0.375 in.

Example 8 A natural gas pipeline, 600 km long, is constructed of DN 800 pipe and has a required operating pressure of 9 MPa. Compare the cost of using X-60 and X-70 steel pipe. The material costs of the two grades of pipe are as follows: Pipe Grade

Material Cost—$/tonne

X-60 X-70

800 900

Use a class 1 design factor and temperature deration factor of 1.00. Solution We will first determine the wall thickness of pipe required to withstand the operating pressure of 9 MPa. Using Equation 6.8, the pipe wall thickness required for X-60 pipe (60,000 psi = 414 MPa) is t=

9 × 800 = 12.08 mm. Use 13 mm wall thickness. 2 × 414 × 1.0 × 0.72 × 1.0

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 246 Friday, April 1, 2005 3:29 PM

246

GAS PIPELINE HYDRAULICS

Similarly, the pipe wall thickness required for X-70 pipe (70,000 psi = 483 MPa) is t=

9 × 800 = 10.35 mm. Use 11 mm wall thickness. 2 × 483 × 1.0 × 0.72 × 1.0

The pipe weight in kg/m will be calculated using Equation 6.12. For X-60 pipe, Weight per meter = 0.0246 × 13 × (800 –13) = 251.68 kg/m Therefore, the total cost of 600 km pipeline at $800 per ton of X-60 pipe is Total cost = 600 × 251.68 × 800 = $120.81 million Similarly, the pipe weight in kg/m for X-70 pipe is Weight per meter = 0.0246 × 11 × (800 – 11) = 213.50 kg/m Therefore, the total cost of 600 km pipeline at $900 per ton of X-70 pipe is Total cost = 600 × 213.50 × 900 = $115.29 million Therefore, the X-70 pipe will cost less than the X-60 pipe. The difference in cost is $120.81 – $115.29 = $5.52 million.

6.12 SUMMARY In this chapter we discussed how to calculate the pipe wall thickness required to withstand an internal pressure in a gas pipeline using Barlow’s equation. The influence of the population density in the vicinity of the pipeline on the required pipe wall thickness by reducing the allowable hoop stress in the high population areas was explained by way of class locations. We explored the range of pressures required to hydrotest pipeline sections to ensure safe operation of the pipeline. The effect of pipeline elevations on determining a testing plan by sectioning the pipeline was covered. The need for isolating portions of the pipeline by properly spaced mainline valves and the method of calculating the time required for evacuating gas from the pipeline sections were also discussed. Finally, a simple method of calculating the pipe tonnage was explained. PROBLEMS 1. A gas pipeline is constructed of API 5L X70 steel, NPS 24, 0.375 in. wall thickness. Calculate the MAOP of this pipeline for a class 1 design factor and a temperature deration factor of 1.00. 2. A gas pipeline, DN 500, 12 mm wall thickness, is constructed of API 5L X65 pipe. (a) Calculate the design pressures for class 1 and class 2 locations. (b) What is the range of hydrotest pressures for each of these class locations? Assume the joint factor = 1.00 and temperature deration factor = 1.00.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 247 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

247

3. Calculate the total tonnage of pipe material for a 24 mi pipeline, NPS 16, 0.375 in. wall thickness. If pipe costs $700 per ton, determine the total pipeline cost. 4. A 50 km pipeline consists of 15 km of DN 400, 10 mm wall thickness pipe connected to a 35 km length of DN 300, 8 mm wall thickness pipe. (a) Calculate the total metric tons of pipe. (b) If this pipeline were replaced with a single 50 km long pipeline, DN 400, API 5LX-65 material, what minimum wall thickness would be required for a class 1 design at an MOP of 9 MPa? 5. Calculate the minimum wall thickness for NPS 16 pipeline constructed of API 5LX-60 steel to withstand an internal pressure of 1440 psi. Use a class 1 design and temperature deration factor of 1.00. 6. A natural gas pipeline, 240 mi long, is constructed of NPS 24 pipe and has a required operating pressure of 1200 psig. Compare the cost of using X-70 or X-80 steel pipe. The material cost of X-70 pipe is $850/ton, and for X-80 pipe it is $1000/ton. Use a class 1 design factor and temperature deration factor of 1.00. 7. A natural gas pipeline, NPS 24, traverses a hilly terrain with elevations ranging from 300 ft at Norwalk to 4500 ft at the Fulton summit (milepost 50), followed by an elevation of 500 ft at the pipeline terminus at Danby. The pipeline is 100 mi long and is constructed of API 5L X-65 pipe as follows:

Section Norwalk to mp 30 mp 30 to Fulton Fulton to mp 70 mp 70 to Danby

Wall Thickness (in.)

Class

0.500 0.500 0.375 0.375

2 1 2 3

Determine a hydrostatic test plan for this pipeline, considering a test pressure envelope of 90 to 95% yield. What is the minimum number of test sections required?

REFERENCES 1. Nayyar, M.L., Piping Handbook, 7th ed., McGraw-Hill, New York, 2000. 2. Baumeister, T., Ed., Standard Handbook for Mechanical Engineers, 7th ed., McGrawHill, New York, 1967. 3. Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003. 4. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 5. Department of Transportation—DOT Code of Federal Regulation 49CFR Part 192, Oct. 2000. 6. ASME B31.8: Gas Transmission and Distribution Piping Systems, 2003–2005.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 248 Friday, April 1, 2005 3:29 PM

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 249 Friday, April 1, 2005 3:28 PM

CHAPTER

7

Thermal Hydraulics In this chapter we will further discuss thermal hydraulics, which was briefly reviewed in Chapter 3. The importance of taking into account the gas temperature variation along a gas transmission pipeline and its impact on pressure drop and flow rate will be explained. Isothermal hydraulics, which formed the majority of calculations in Chapter 2 and Chapter 3, will be compared with thermal hydraulics. Since manual calculation of gas pipeline hydraulics, considering thermal effects, is quite laborious and time consuming, we will use examples of pipeline simulation cases using a popular gas pipeline hydraulics software application. 7.1 ISOTHERMAL VERSUS THERMAL HYDRAULICS In the previous chapters the hydraulic analysis of gas flow through pipelines was mainly done based upon isothermal or constant temperature flow. This assumption is fairly good in long-distance pipelines where the gas temperature reaches a constant value equal to or close to the surrounding soil (or ambient) temperature at large distances from the compressor stations. However, upon compressing the gas, depending on the compression ratio, the outlet temperature of the gas from the compressor station can be considerably higher than that of the ambient air or surrounding soil. In Chapter 4, Example 8, we found that when gas is compressed adiabatically from a 60°F suction temperature and a compression ratio of 2.0, the discharge temperature is 278.3°F. Since pipe coating limitations restrict temperatures to about 140 to 150°F, cooling of the compressed gas is necessary at the downstream side of the compressor station. In this example, assuming gas cooling results in a discharge temperature of 140°F as gas enters the pipeline, we find that the temperature difference between the gas at 140°F and the surrounding soil at 70°F will cause heat transfer to take place between the pipeline gas and the surrounding soil. It is found that the gas temperature drops off rapidly for the first few miles and eventually reaches a temperature close to the soil temperature. Additionally, in a long transmission pipeline, the soil temperature can vary along the pipeline as well, causing different heat transfer rates at locations along the pipeline. This is illustrated in Figure 7.1. 249

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 250 Friday, April 1, 2005 3:28 PM

250

GAS PIPELINE HYDRAULICS

Tinlet = 140°F

Tem

pera

ture

grad

ient Toutlet = 70°F

Compressor station Figure 7.1

Terminus

Temperature variation in a gas pipeline.

In some instances, the expansion of gas as it flows along a pipeline can result in gas temperature reaching a slightly lower temperature than the surrounding soil. This is called the Joule-Thompson cooling effect. Thus, if the soil temperature is fairly constant at 70°F, due to the Joule-Thompson effect, the final temperature of the gas at the terminus of the pipeline can drop to 60 or 65°F. This is illustrated in Figure 7.2. This cooling of gas below the surrounding soil temperature depends on the pressure differential and the Joule-Thompson coefficient. Ignoring this cooling effect will result in a more conservative (lower flow rate for a given pressure drop) flow rate calculation, since cooler temperature means less pressure drop in a gas pipeline and, hence, higher flow rate. Thermal hydraulics is the study of gas pipeline pressures, temperatures, and flow rates, taking into account the thermal properties of the soil, pipe, and pipe insulation, if any. Due to such variation in gas temperature, calculation of pressure drop must be made by considering short lengths of pipe that make up the total pipeline. For example, if the pipeline is 50 mi long, we subdivide it into short segments of 1 or 2 mi lengths and apply the General Flow equation for each pipe segment, considering

Gas temperature

Gas inlet temperature = 140°F

Ground temperature = 70°F

Distance Figure 7.2

Joule-Thompson cooling effect in a gas pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

Gas outlet temperature = 60–65°F

2785_C007.fm Page 251 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

251

an average gas temperature and an average ambient soil temperature. Starting with the upstream pressure of segment 1, the downstream pressure will be calculated assuming an average temperature for segment 1. Next, using the calculated downstream pressure as the upstream pressure for segment 2, we calculate the downstream pressure for segment 2. The process is continued until all segments of the pipeline are covered. It must be noted that the variation of temperature from segment to segment must be taken into account to calculate the compressibility factor used in the General Flow equation. The following equation is the General Flow equation that we used frequently in Chapter 2 and Chapter 3: T P 2 − e s P22 Q = 38.77 F b 1 Pb GT f Le Z

0.5 5

D 2.5

(7.1)

After examining Equation 7.1, we see that if the gas flowing temperature Tf is constant (isothermal flow) throughout the length of the pipeline Le, the compressibility factor Z of the gas will depend on the average pressure of the pipe segment. If the gas temperature also varies along the pipeline, based on the preceding discussions, the compressibility factor Z will change in a different manner, since it is a function of both the gas temperature and gas pressure. Therefore, it is seen that calculation of P1 or P2 for a given flow rate Q from Equation 7.1 will yield different results if isothermal conditions do not exist. The calculation of gas temperature at any point along the pipeline, taking into account the heat transfer between the gas and surrounding soil, is quite complicated. It does not lend itself easily to manual calculations. We will discuss the method of calculation for thermal hydraulics in this section. To accurately take into account the temperature variations, a suitable gas pipeline hydraulics simulation program must be used since, as indicated earlier, manual calculation is quite laborious and time consuming. Several commercial simulation programs are available to model steady-state gas pipeline hydraulics. These programs calculate the gas temperatures and pressures by taking into consideration variations of soil temperature, pipe burial depth, and thermal conductivities of pipe, insulation, and soil. One such software program is GASMOD, marketed by SYSTEK Technologies, Inc. (www.systek.us). Appendix D includes a sample simulation of a gas pipeline using the GASMOD software. In this chapter we will use GASMOD to illustrate thermal hydraulics analysis.

7.2 TEMPERATURE VARIATION AND GAS PIPELINE MODELING Consider a buried pipeline transporting gas from point A to point B. We will analyze a short segment of length ∆ L of this pipe, as shown in Figure 7.3, and apply the principles of heat transfer to determine how the gas temperature varies along the pipeline. The upstream end of the pipe segment of length ∆ L is at temperature T1 and the downstream end at temperature T2. The average gas temperature in this segment is represented by T. The outside soil temperature at this location is Ts . Assume steadystate conditions and the mass flow rate of gas to be m. The gas flow from the upstream

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 252 Friday, April 1, 2005 3:28 PM

252

GAS PIPELINE HYDRAULICS

Soil temperature Ts

A

T2

T

T1 Q

B

∆L 1

Figure 7.3

2

Analysis of temperature variation.

end to the downstream end of the segment causes a temperature drop of ∆T. The heat loss from the gas can be represented by ∆ H = − mCp∆T where ∆H m Cp ∆T

= = = =

(7.2)

heat transfer rate, Btu/h mass flow rate of gas, lb/h average specific heat of gas, Btu/lb/ °F temperature difference = T1 – T2, °F

The negative sign in Equation 7.2 indicates loss of heat from upstream temperature T1 to downstream temperature T2. Next, we consider the heat transfer from the gas to the surrounding soil in terms of the overall heat transfer coefficient U and the difference in temperature between the gas and surrounding soil represented by (T – Ts ). Therefore, we can write the following equation for heat transfer: ∆ H = U∆ A (T – Ts) where U ∆A T Ts D

(7.3)

= overall heat transfer coefficient, Btu/h/ft2 / °F = surface area of pipe for heat transfer = p D∆ L = average gas temperature in pipe segment, °F = average soil temperature surrounding pipe segment, °F = pipe inside diameter, ft

Equating the two values of the heat transfer rate ∆ H from Equation 7.2 and Equation 7.3, we get −mCp∆T = U∆A (T – Ts) Simplifying, we get πUD ∆T = − ∆L T − Ts mCp

Copyright 2005 by Taylor & Francis Group, LLC

(7.4)

2785_C007.fm Page 253 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

253

Rewriting Equation 7.4 in differential form and integrating, we get

∫

2

1

dT = T − Ts

πUD

∫ − mCp dL 1

(7.5)

2

Integrating and simplifying, we get T2 − Ts = e −θ T1 − Ts

(7.6)

where e = base of natural logarithms (e = 2.718…) and

θ=

πUD∆ L mCp

(7.7)

Simplifying Equation 7.6 further, we get the downstream temperature of the pipe segment of length ∆ L as T2 = Ts + (T1 − Ts )e −θ

(7.8)

It can be seen from Equation 7.8 that as the pipe length increases, the term e −θ approaches zero and the temperature T2 becomes equal to soil temperature Ts. Therefore, in a long gas pipeline, the gas temperature ultimately equals the surrounding soil temperature. This is illustrated in Figure 7.1. In the preceding analysis, we made several simplifying assumptions. We assumed that the soil temperature and the overall heat transfer coefficient remained constant and ignored the Joule-Thompson effect as gas expands through a pipeline. In a long pipeline, the soil temperature may actually vary along the pipeline and, therefore, must be taken into account in these calculations. One approach would be to subdivide the pipeline into segments that have constant soil temperatures and perform calculations for each segment separately. The Joule-Thompson effect causes the gas to cool slightly due to expansion. Therefore, in a long pipeline, the gas temperature at the delivery point can fall below that of the ground or soil temperature, as indicated in Figure 7.2.

7.3 REVIEW OF SIMULATION MODEL REPORTS To illustrate thermal effects in a gas pipeline, we will analyze a gas transmission pipeline, first using the method outlined in Chapter 3. Next, we will analyze the same pipeline, taking into account the thermal conductivity and soil temperatures. The latter method requires some form of computer simulation models. To do this, we have chosen the commercially available software known as GASMOD. We will compare the results of the isothermal hydraulics of Chapter 3 with the thermal hydraulics using GASMOD. Examples will be used to illustrate the comparison.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 254 Friday, April 1, 2005 3:28 PM

254

GAS PIPELINE HYDRAULICS

Example 1 A natural gas pipeline system is being built from Rockport to Concord, a distance of 240 miles. The pipeline is constructed of NPS 30, 0.500 in. wall thickness, API 5L-X60 pipe. The MOP is 1400 psig. Gas enters the Rockport compressor station at 70°F and 800 psig pressure. The soil temperature can be assumed to be 60°F throughout. The gas flow rate is 420 MMSCFD, and the gas specific gravity and viscosity are 0.6 and 0.000008 lb/ft-s, respectively. The contract delivery pressure required at Concord is 500 psig. Assume an isothermal flow at 70°F and a gas specific heat ratio of 1.29. Use a compressor adiabatic efficiency of 80% and mechanical efficiency of 98%. Use the General Flow equation with a Colebrook friction factor, assuming a pipe internal roughness of 700 µ in. Calculate the pressure profile and the compressor horsepower required at Rockport. Compare these results with thermal hydraulic analysis using GASMOD. Assume a base pressure of 14.7 psia and base temperature of 60°F. The pipeline elevation profile is essentially flat. Solution Inside diameter of pipe D = 30 – 2 × 0.500 = 29 in. First, we calculate the Reynolds number from Equation 2.34: 14.7 0.6 × 420 × 10 6 = 14,671,438 R = 0.0004778 60 + 460 0.000008 × 29 Next, using Colebrook Equation 2.39, we calculate the friction factor as 0.0007 2.51 = −2Log10 + f 3.7 × 29 14, 671, 438 f

1

Solving by trial and error, we get f = 0.0097 Therefore, the transmission factor is, using Equation 2.42, F=

2 0.0097

= 20.33

To calculate the compressibility factor Z, the average pressure is required. Since the inlet pressure is unknown, we will calculate an approximate value of Z using a value of 110% of the delivery pressure for the average pressure. The average pressure is Pavg = 1.1 × (500 + 14.7) = 566.17 psia = 551.47 psig

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 255 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

255

Using CNGA Equation 1.34, we calculate the value of compressibility factor as Z=

1 (566.17−14.7)×344,400(10)1.785× 0.6 1 + 5303.825

= 0.9217

Since there is no elevation difference between the beginning of the pipeline and the end of the pipeline, the elevation component in Equation 2.7 can be neglected and es = 1. The outlet pressure is P2 = 500 + 14.7 = 514.7 psia. From General Flow Equation 2.4, substituting given values, we get 520 P12 − 514.72 420 × 10 6 = 38.77 × 20.33 14.7 0.6 × 530 × 240 × 0.9217

0.5

(29.0)2.5

Solving for the upstream pressure, we get P1 = 1021.34 psia = 1006.64 psig Using this value of P1, we calculate the new average pressure using Equation 2.14: Pavg =

1021.34 × 514.7 2 1021.34 + 514.7 − = 795.87 psia 1021.34 + 514.7 3

This compares with the value of 566.17 psia we assumed initially for calculating Z. Obviously, we were way off. Recalculating Z using the recently calculated value of Pavg, we get Z=

1 (795.87−14.7)×344,400(10)1.785× 0.6 1 + 5303.825

= 0.8926

We will now recalculate the inlet pressure using this value of Z. From General Flow Equation 2.4, we get 520 P12 − 514.72 420 × 10 6 = 38.77 × 20.33 14.7 0.6 × 530 × 240 × 0.8926

0.5

(29.0)2.5

Solving for the upstream pressure, we get P1 = 1009.24 psia = 994.54 psig This compares with 1021.34 psia calculated earlier. This is almost 1% different. We could repeat the process and get a better approximation.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 256 Friday, April 1, 2005 3:28 PM

256

GAS PIPELINE HYDRAULICS

Using this recently calculated value of P1, we calculate the new average pressure using Equation 2.14 as Pavg =

1009.24 × 514.7 2 1009.24 + 514.7 − = 788.72 psia 1009.24 + 514.7 3

This compares with the previous approximation of 795.87 psia. The error is less than 1%. Recalculating Z using this value of Pavg , we get Z=

1 (788.72−14.7)×344,400(10)1.785× 0.6 1 + 5303.825

= 0.8935

We will now recalculate the inlet pressure using this value of Z. From General Flow Equation 2.4, we get 520 P12 − 514.72 420 × 10 = 38.77 × 20.33 14.7 0.6 × 530 × 240 × 0.8935 6

0.5

(29.0)2.5

Solving for the upstream pressure, we get P1 = 1009.62 psia = 994.92 psig This compares with 1009.24 psia calculated earlier. The difference is less than 0.04%; therefore, we can stop iterating any further. The HP required at the Rockport compressor station will be calculated using Equation 4.15 as follows: 0.29 1 + 0.8935 1 1009.62 1.29 1.29 HP = 0.0857 × 420 − 1 = 4962 (70 + 460) 0.8 814.7 2 0.29

Using Equation 4.17, we calculate the driver horsepower required, based on a mechanical efficiency of 0.98. BHP required =

4962 = 5063 0.98

The final results are: Inlet pressure at Rockport = 994.92 psig Delivery pressure at Concord = 500 psig at a flow rate of 420 MMSCFD BHP required at Rockport compressor station = 5063 HP

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 257 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

257

It must be noted that the preceding calculations ignored any elevation changes along the pipeline. If we had considered the pipe elevations at Rockport and Concord, the result would have been different. This pressure of 994.92 psig required at the Rockport compressor station was calculated assuming a constant gas flowing temperature of 70°F and considering the pipeline as one single segment 240 mi long. As explained in earlier chapters, the calculation accuracy is improved if we subdivide the pipeline into short segments. By doing so, we calculate the upstream pressure of each segment starting with the last segment near Concord. If the pipeline is divided into 100 equal pipe segments of 2.4 mi each, the pressure P100 at the upstream end of the last segment is calculated using the General Flow equation, considering a 500 psig downstream pressure. Next, using this calculated pressure, P100, we calculate the upstream pressure P99 of the 99th segment. The process is repeated until all segments are covered and the value of the pressure P1 at Rockport is calculated. This is illustrated in Figure 7.4. By subdividing the pipeline in this fashion, we are improving the accuracy of calculations. Of course, manual calculation in this manner is going to be quite laborious and time consuming, and we should use some form of a computer program to perform this task. Next, we will compare the isothermal calculation results with thermal hydraulics using the GASMOD program. Given next is the output report from the GASMOD program.

Pressure

P1

P99 P100 500 psig Subdivided into 100 segments

Rockport

Concord Distance = 240 mi

Figure 7.4

Subdividing pipe into segments.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 258 Friday, April 1, 2005 3:28 PM

258

GAS PIPELINE HYDRAULICS

******** GASMOD — GAS PIPELINE HYDRAULIC SIMULATION ******* ************ 32-bit Version 5.00.100 ************ DATE:

16-September-2004 TIME: 21:30:37

PROJECT DESCRIPTION: Pipeline from Rockport to Concord Case Number:

1139

Pipeline data file:

C:\GASMOD32\RockportPipeline.TOT

Pressure drop formula: Pipeline efficiency: Compressibility factor method: Inlet gas gravity(air=1.0): Inlet gas viscosity:

Colebrook-White 1.00 CNGA 0.60000 0.0000080(lb/ft-sec)

CALCULATION OPTIONS: Branch pipe calculations: Loop pipe calculations: Compressor fuel calculated: Joule-Thompson effect included: Customized output: Holding delivery pressure at terminus

NO NO NO NO NO

******** Calculations Based on Specified Thermal Conductivities of Pipe, Soil, and Insulation ******** Origin suction temperature: Base temperature: Base pressure: Origin suction pressure: Delivery pressure: Minimum pressure: Gas specific heat ratio: Maximum gas velocity:

70.00(degF) 60.00(degF) 14.700(psig) 800.00(psig) 500.35(psig) 100.00(psig) 1.29 50.00(ft/sec)

Inlet flow rate: Outlet flow rate:

420.0000(MMSCFD) 420.0000(MMSCFD)

**************** PIPELINE PROFILE DATA *********** Distance (mi)

Elevation (ft)

Diameter (in.)

Thickness (in.)

Roughness (in.)

0.00 10.00 20.00 30.00

250.00 250.00 250.00 250.00

30.000 30.000 30.000 30.000

0.500 0.500 0.500 0.500

0.000700 0.000700 0.000700 0.000700

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 259 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

40.00 50.00 60.00 70.00 100.00 120.00 140.00 150.00 170.00 190.00 200.00 220.00 240.00

259

250.00 250.00 250.00 250.00 250.00 250.00 250.00 250.00 250.00 250.00 250.00 250.00 250.00

30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000

0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500

0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700

******** THERMAL CONDUCTIVITY AND INSULATION DATA *********

Distance (mi)

Cover (in)

0.000 10.000 20.000 30.000 40.000 50.000 60.000 70.000 100.000 120.000 140.000 150.000 170.000 190.000 200.000 220.000 240.000

36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000

****************

Thermal Conductivity (Btu/hr/ft/degF) Pipe Soil Insulation 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000

0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800

0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020

Insul.Thk (in)

Soil Temp (degF)

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00

COMPRESSOR STATION DATA **************

FLOW RATES, PRESSURES, AND TEMPERATURES:

Name

Flow Suct. Disch. Rate Press. Press. (MMSCFD) (psig) (psig)

Rockport 420.00

Copyright 2005 by Taylor & Francis Group, LLC

800.00 996.17

Compr. Suct. Disch. Suct. Disch. Ratio Loss. Loss. Temp. Temp MaxPipe (psig) (psig) (degF) (degF) (degF) Temp 1.2408 0.00

0.00

70.00

102.92 140.00

2785_C007.fm Page 260 Friday, April 1, 2005 3:28 PM

260

GAS PIPELINE HYDRAULICS

********* COMPRESSOR EFFICIENCY, HP, AND FUEL USED ********

Name

Compr Mech. Overall Horse Distance Effy. Effy. Effy. Power (mi) (%) (%) (%)

Rockport 0.00

80.00 98.00 78.40

Fuel Factor (MCF/day/ HP)

4,786.94 0.2000

Total Compressor Station Horsepower:

Fuel Used (MMSCFD)

------

4,786.94

**************** LOCATIONS AND FLOW RATES **************** Location Distance Flow in/out Gravity Viscosity Pressure GasTemp. (mi) (MMSCFD) (lb/ft-sec) (psig) (degF) Rockport 0.00 Concord 240.00

420.0000 -420.0000

0.6000 0.6000

0.00000800 0.00000800

996.17 500.35

102.92 60.00

***** REYNOLD'S NUMBER AND HEAT TRANSFER COEFFICIENT *****

Distance (mi) 0.000 10.000 20.000 30.000 40.000 50.000 60.000 70.000 100.000 120.000 140.000 150.000 170.000 190.000 200.000 220.000 240.000

FrictFactor Transmission Reynold'sNum. (Darcy) Factor 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438.

0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099

HeatTransCoeff Compressibility (Btu/hr/ Factor ft2/degF) (CNGA)

20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13

0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361

0.8865 0.8791 0.8747 0.8725 0.8718 0.8722 0.8733 0.8767 0.8821 0.8870 0.8909 0.8952 0.9012 0.9061 0.9114 0.9171 0.9171

******* PIPELINE TEMPERATURE AND PRESSURE PROFILE ****** Distance Diameter Flow Velocity Press. GasTemp. SoilTemp. MAOP Location (mi) (in) (MMSCFD) (ft/sec) (psig) (degF) (degF) (psig) 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 100.00

30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000

Copyright 2005 by Taylor & Francis Group, LLC

420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 420.0000

15.44 15.44 15.44 15.44 15.44 15.44 15.44 15.44 15.44

996.17 979.59 963.25 946.96 930.60 914.07 897.30 880.23 826.86

102.92 87.37 77.22 70.74 66.66 64.11 62.54 61.56 60.36

60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00

1400.00 Rockport 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00

2785_C007.fm Page 261 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

120.00 140.00 150.00 170.00 190.00 200.00 220.00 240.00

30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000

420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 420.0000

261

15.44 15.44 15.44 15.44 15.44 15.44 15.44 15.44

789.12 749.30 728.51 684.85 637.95 613.06 559.72 500.35

60.14 60.05 60.03 60.01 60.00 60.00 60.00 60.00

60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00

1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00

Concord

************* LINE PACK VOLUMES AND PRESSURES ************* Distance (mi)

Pressure (psig)

Line Pack (million std.cu.ft)

0.00 996.17 0.0000 10.00 979.59 17.8138 20.00 963.25 17.9714 30.00 946.96 17.9566 40.00 930.60 17.8193 50.00 914.07 17.5990 60.00 897.30 17.3235 70.00 880.23 17.0114 100.00 826.86 48.9787 120.00 789.12 30.7496 140.00 749.30 29.1424 150.00 728.51 13.9438 170.00 684.85 26.5765 190.00 637.95 24.7442 200.00 613.06 11.6489 220.00 559.72 21.7584 240.00 500.35 19.5576 Total line pack in main pipeline = 350.5950(million std.cu.ft) ************** End of GASMOD Output Report ************* It can be seen from the GASMOD thermal hydraulic analysis report that the inlet pressure at Rockport is 996.17 psig, whereas the manual calculation considering isothermal flow yielded an inlet pressure of 994.92 psig at the Rockport compressor station. Thus, taking into account the temperature variation of the gas along the pipeline, the pressure required at Rockport is approximately 4 psig higher. This does not seem to be very significant. However, in many cases the temperature variation along the pipeline will cause pressures calculated to be significantly different. To recap, the manual calculations were based on an isothermal gas flow temperature of 70°F, whereas the thermal hydraulics shows variation of the gas temperature ranging from 102.92°F at the Rockport compressor discharge to 60°F at Concord. The gas temperature reaches the soil temperature of 60°F at approximately milepost 190, after which it remains constant at 60°F. Figure 7.5 shows the temperature variation in this case.

Next, we will illustrate the calculation of the pressure and temperature profile considering the pipeline elevation difference between Rockport and Concord and considering a branch pipeline bringing in an additional 200 MMSCFD.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 262 Friday, April 1, 2005 3:28 PM

262

GAS PIPELINE HYDRAULICS

Tinlet = 102.92°F T10 = 87.37°F T20 = 77.22°F T30 = 70.74°F Tem

pera

Rockport Figure 7.5

ture

T140 = 60.05°F

grad

ient

Toutlet = 60°F

240 mi

Concord

Gas temperature variation—Rockport to Concord pipeline.

Example 2 Consider a natural gas pipeline system from Rockport (elevation 250 ft) to Concord (elevation 800 ft), a distance of 240 mi. The pipeline is constructed of NPS 30, 0.500 in. wall thickness, API 5L-X60 pipe. The MOP is 1400 psig. The pipeline elevation profile is listed below: Milepost

Elevation

Location

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 100.00 120.00 140.00 150.00 170.00 190.00 200.00 220.00 240.00

250.00 300.00 200.00 320.00 400.00 375.00 410.00 430.00 450.00 500.00 400.00 600.00 700.00 710.00 720.00 750.00 800.00

Rockport

Vale

Concord

Gas enters the Rockport compressor station at 70°F and 800 psig pressure. The soil temperature can be assumed to be 60°F throughout. The gas flow rate is 420 MMSCFD, and the gas specific gravity and viscosity are 0.6 and 0.000008 lb/ft-s, respectively. At Vale (milepost 100, elevation 450 ft), a branch pipeline 80 mi long, NPS 24, 0.375 in. wall thickness, brings in an additional 200 MMSCFD gas from a gathering facility at Drake. The elevation at Drake is 300 ft and that at Vale is 450 ft.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 263 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

263

The pipeline elevation profile for the branch pipe from Drake to Vale is as follows:

Milepost

Elevation

Location

0.00 10.00 20.00 40.00 50.00 70.00 80.00

300.00 100.00 125.00 200.00 250.00 300.00 450.00

Drake

Vale

The inlet temperature at the beginning of the branch is 70°F in Figure 7.6. The contract delivery pressure required at Concord is 500 psig. Assume an isothermal flow at 70°F and gas specific heat ratio of 1.29. Use a compressor adiabatic efficiency of 80% and mechanical efficiency of 98%. Use the General Flow equation with a Colebrook friction factor, assuming a pipe internal roughness of 700 µ in. Calculate the pressure profile and the compressor horsepower required at Rockport. Compare these results with thermal hydraulic analysis using GASMOD. Assume a base pressure of 14.7 psia and base temperature of 60°F. Solution Inside diameter of pipe D = 30 – 2 × 0.500 = 29 in. The calculation of the pressure at milepost 100 will be done first. This is because we know the delivery pressure at Concord and the 140 mi segment from Vale (at milepost 100 to the pipeline terminus at Concord) flows 620 MMSCFD. In comparison, the first 100 mi from Rockport to Vale, carries only 420 MMSCFD, and both upstream and downstream pressures are unknown. After finding the pressure at Vale, we can

20

0

M

80

M

m

SC

iN

FD

PS

24

Drake elevation 300 ft

420 MMSCFD 240 mi NPS 30 Rockport m.p. 0 elevation 250 ft Figure 7.6

Vale m.p. 100 elevation 450 ft

Rockport to Concord pipeline with branch from Drake.

Copyright 2005 by Taylor & Francis Group, LLC

Concord m.p. 240 elevation 800 ft

2785_C007.fm Page 264 Friday, April 1, 2005 3:28 PM

264

GAS PIPELINE HYDRAULICS

calculate the upstream pressure at Rockport, considering the 100 mi segment at the lower flow rate. First, we calculate the Reynolds number from Equation 2.34 for a 620 MMSCFD flow rate: 14.7 0.6 × 620 × 10 6 = 21,657,837 R = 0.0004778 60 + 460 0.000008 × 29 Next, using Colebrook Equation 2.39, we calculate the friction factor as 0.0007 2.51 = −2Log10 + . 3 7 29 × 21, 657, 837 f f

1

Solving by trial and error, we get f = 0.0096 Therefore, the transmission factor is, using Equation 2.42, F=

2 0.0096

= 20.41

To calculate the compressibility factor Z, the average pressure is required. Since the inlet pressure at Vale is unknown, we will calculate an approximate value of Z using a value of 110% of the delivery pressure for the average pressure. The average pressure is Pavg = 1.1 × (500 + 14.7) = 566.17 psia = 551.47 psig Using CNGA Equation 1.34, we calculate the value of the compressibility factor as Z=

1 = 0.9217 (566.17−14.7)×344,400(10)1.785× 0.6 + 1 5303.825

Since there is an elevation difference of 350 (800 − 450) ft between Vale and Concord, we must apply the elevation correction according to Equation 2.7. Using Equation 2.10, the elevation adjustment parameter is 800 − 450 s = 0.0375 × 0.6 = 0.0161 530 × 0.9217 The equivalent length Le from Equation 2.9 is Le =

Copyright 2005 by Taylor & Francis Group, LLC

140(1.0163 − 1) = 141.74 mi 0.0161

2785_C007.fm Page 265 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

265

Next, using Equation 2.7, we calculate the pressure at Vale, milepost 100, as follows: 520 P12 − 1.0163 × 514.72 620 × 10 = 38.77 × 20.41 14.7 0.6 × 530 × 141.74 × 0.9217

0.5

292.5

6

Solving for the upstream pressure at Vale, we get P1 = 1122.49 psia = 1107.79 psig Using this calculated value of P1, we calculate the new average pressure using Equation 2.14 as Pavg =

2 1122.49 × 514.7 = 856.2 psia = 841.5 psig 1122.49 + 514.7 − 3 1122.49 + 514.7

This compares with the previous approximation of 551.47 psig. Obviously, the assumed value was way off. Recalculating the compressibility factor, using the recently calculated average pressure, Z=

1 (856.2−14.7)×344,400(10)1.785× 0.6 1 + 5303.8825

= 0.8852

Recalculating the pressure at Vale, we get 520 P12 − 1.0163 × 514.72 620 × 10 6 = 38.77 × 20.41 14.7 0.6 × 530 × 141.74 × 0.8852

0.5

292.5

Solving for the upstream pressure at Vale, we get P1 = 1104.88 psia = 1090.18 psig Compared to the last calculated value, the difference is: 1090.18 – 1107.79 = −17.61 psig or 1.6%. One more iteration would get us closer to the correct value. We recalculate the new average pressure using Equation 2.14 as Pavg =

1104.88 × 514.7 2 1104.88 + 514.7 − = 845.63 psia = 830.93 psig 1104.88 + 514.7 3

We recalculate the compressibility factor, using the recently calculated average pressure, as Z=

1 (845.63−14.7)×344,400(10)1.785× 0.6 1 + 5303.825

= 0.8865

Recalculating the pressure at Vale, we get 520 P12 − 1.0163 × 514.72 620 × 10 = 38.77 × 20.41 14.7 0.6 × 530 × 141.74 × 0.8865 6

Copyright 2005 by Taylor & Francis Group, LLC

0.5

292.5

2785_C007.fm Page 266 Friday, April 1, 2005 3:28 PM

266

GAS PIPELINE HYDRAULICS

Solving for the upstream pressure at Vale, we get P1 = 1106.15 psia = 1091.45 psig Compared to the last calculated value, the difference is: 1090.18 – 1091.45 = −1.27 psig or 0.12%. This is close enough, and no further iteration is needed. Therefore, the pressure at Vale = 1106.15 psia = 1091.45 psig. Next, using this as the downstream pressure for the pipe segment from Rockport to Vale, we calculate the upstream pressure at Rockport as follows, for a flow rate of 420 MMSCFD. From previous calculations at 420 MMSCFD (Example 1), the Reynolds number is R = 14,671,438 The friction factor was calculated as f = 0.0097 and the transmission factor was F=

2 0.0097

= 20.33

To calculate the compressibility factor Z, the average pressure is required. Since the inlet pressure at Rockport is unknown, we will calculate an approximate value of Z using a value of 110% of the downstream pressure at Vale for the average pressure. The average pressure is Pavg = 1.1 × 1106.15 = 1216.77 psia = 1202.07 psig Using this, we calculate the compressibility factor as Z=

1 (1216.77−14.7)×344,400(10)1.785× 0.6 1 + 5303.825

= 0.8437

Since there is an elevation difference of 200 (450 − 250) ft between Rockport and Vale, the elevation correction according to Equation 2.7 must be applied. Using Equation 2.10, the elevation adjustment parameter is 450 − 250 = 0.0101 s = 0.0375 × 0.6 530 × 0.8437 The equivalent length Le from Equation 2.9 is Le =

Copyright 2005 by Taylor & Francis Group, LLC

100(1.0101 − 1) = 100 mi 0.0101

2785_C007.fm Page 267 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

267

Calculating the pressure at Rockport, we get 520 P12 − 1.0101 × 1106.152 420 × 10 6 = 38.77 × 20.33 14.7 0.6 × 530 × 100 × 0.8437

0.5

292.5

Solving for the upstream pressure at Rockport, we get P1 = 1238.04 psia = 1223.34 psig Next, we will compare the isothermal calculation results with thermal hydraulics using the GASMOD program. Given below is the output report from the GASMOD program. ******** GASMOD — GAS PIPELINE HYDRAULIC SIMULATION ******* ************ 32-bit Version 5.00.100 ************ DATE: PROJECT DESCRIPTION: Pipeline from Rockport to Concord Case number: Pipeline data file:

17-September-2004 TIME: 07:16:19

1143 C:\GASMOD32\RockportPipeline.TOT

Pressure drop formula: Pipeline efficiency: Compressibility factor method: Inlet gas gravity (air=1.0): Inlet gas viscosity:

Colebrook-White 1.00 CNGA 0.60000 0.0000080(lb/ft-sec)

CALCULATION OPTIONS: Branch pipe calculations: Loop pipe calculations: Compressor fuel calculated: Joule-Thompson effect included: Customized output: Holding delivery pressure at terminus

YES NO NO NO NO

**** Calculations Based on Specified Thermal Conductivities of Pipe, Soil, and Insulation **** Origin suction temperature: Base temperature: Base pressure: Origin suction pressure: Delivery pressure: Minimum pressure: Gas specific heat ratio: Maximum gas velocity:

70.00(degF) 60.00(degF) 14.700(psig) 800.00(psig) 499.66(psig) 100.00(psig) 1.29 50.00(ft/sec)

Inlet flow rate: Outlet flow rate:

420.0000(MMSCFD) 620.0000(MMSCFD)

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 268 Friday, April 1, 2005 3:28 PM

268

GAS PIPELINE HYDRAULICS

**************** PIPELINE PROFILE DATA *********** Distance (mi)

Elevation (ft)

Diameter (in)

Thickness (in)

Roughness (in)

250.00 300.00 200.00 320.00 400.00 375.00 410.00 430.00 450.00 500.00 400.00 600.00 700.00 710.00 720.00 750.00 800.00

30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000

0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500

0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 100.00 120.00 140.00 150.00 170.00 190.00 200.00 220.00 240.00

********* THERMAL CONDUCTIVITY AND INSULATION DATA ********

Distance (mi)

Cover (in)

0.000 10.000 20.000 30.000 40.000 50.000 60.000 70.000 100.000 120.000 140.000 150.000 170.000 190.000 200.000 220.000 240.000

36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000

Copyright 2005 by Taylor & Francis Group, LLC

Thermal Conductivity (Btu/hr/ft/degF) Insul.Thk Pipe Soil Insulation (in) 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000

0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800

0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

Soil Temp (degF) 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00

2785_C007.fm Page 269 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

269

**************** COMPRESSOR STATION DATA ************** FLOW RATES, PRESSURES, AND TEMPERATURES: Flow Suct. Disch. Suct. Disch. Suct. Disch. MaxPipe Rate Press. Press. Compr. Loss. Loss. Temp. Temp Temp Name (MMSCFD) (psig) (psig) Ratio (psig) (psig) (degF) (degF) (degF) Rockport 420.00

800.00 1223.97 1.5204 0.00

0.00

70.00

132.24 140.00

****** COMPRESSOR EFFICIENCY, HP, AND FUEL USED *********

Name

Compr Mech. Overall Distance Effy. Effy. Effy. Horse (mi) (%) (%) (%) Power

Rockport 0.00

80.00 98.00 78.40

Fuel Fuel Factor Used (MCF/day/HP) (MMSCFD)

9,513.40 0.2000

------

Total Compressor Station Horsepower: 9,513.40

**************** LOCATIONS AND FLOW RATES **************** Location Distance Flow in/out Gravity Viscosity Pressure GasTemp. (mi) (MMSCFD) (lb/ft-sec) (psig) (degF) Rockport 0.00 100.00 Concord 240.00

420.0000 200.0000 -620.0000

0.6000 0.6000 0.6000

0.00000800 0.00000800 0.00000800

1223.97 1085.71 499.66

132.24 60.65 60.01

****** REYNOLD'S NUMBER AND HEAT TRANSFER COEFFICIENT *****

Distance (mi) Reynold'sNum. 0.000 10.000 20.000 30.000 40.000 50.000 60.000 70.000 100.000 120.000 140.000 150.000 170.000 190.000 200.000 220.000 240.000

14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 21,657,838. 21,657,838. 21,657,838. 21,657,838. 21,657,838. 21,657,838. 21,657,838. 21,657,838. 21,657,838.

Copyright 2005 by Taylor & Francis Group, LLC

FrictFactor Transmission Factor (Darcy) 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0098 0.0098 0.0098 0.0098 0.0098 0.0098 0.0098 0.0098 0.0098

20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.25 20.25 20.25 20.25 20.25 20.25 20.25 20.25 20.25

HeatTransCoeff Compressibility (Btu/hr/ Factor ft2/degF) (CNGA) 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3365 0.3365 0.3365 0.3365 0.3365 0.3365 0.3365 0.3365 0.3365

0.8820 0.8678 0.8578 0.8516 0.8479 0.8460 0.8454 0.8465 0.8516 0.8589 0.8650 0.8721 0.8821 0.8903 0.8998 0.9148 0.9148

2785_C007.fm Page 270 Friday, April 1, 2005 3:28 PM

270

GAS PIPELINE HYDRAULICS

******** PIPELINE TEMPERATURE AND PRESSURE PROFILE ******** Distance (mi)

Diameter Flow Velocity Press. GasTemp. SoilTemp. MAOP Location (in) (MMSCFD) (ft/sec) (psig) (degF) (degF) (psig)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 100.00 120.00 140.00 150.00 170.00 190.00 200.00 220.00 240.00

30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000

420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 620.0000 620.0000 620.0000 620.0000 620.0000 620.0000 620.0000 620.0000 620.0000

12.60 12.60 12.60 12.60 12.60 12.60 12.60 12.60 18.60 18.60 18.60 18.60 18.60 18.60 18.60 18.60 18.60

1223.97 1208.59 1197.97 1181.38 1166.06 1153.87 1139.88 1126.25 1085.71 1023.96 961.88 922.84 847.15 765.54 720.84 620.90 499.66

132.24 107.14 90.16 79.02 71.88 67.38 64.56 62.81 60.65 60.34 60.17 60.13 60.06 60.03 60.02 60.01 60.01

60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00

1400.00 Rockport 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 Concord

************ LINE PACK VOLUMES AND PRESSURES ************

Distance (mi)

Pressure (psig)

Line Pack (million std.cu.ft)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 100.00 120.00 140.00 150.00 170.00 190.00 200.00 220.00 240.00

1223.97 1208.59 1197.97 1181.38 1166.06 1153.87 1139.88 1126.25 1085.71 1023.96 961.88 922.84 847.15 765.54 720.84 620.90 499.66

0.0000 21.2972 22.0280 22.4450 22.5706 22.5714 22.4716 22.2810 65.4581 41.3973 38.6813 18.2370 34.0291 30.7118 14.0358 25.1637 20.7997

Total line pack in main pipeline = 444.1785(million std.cu.ft)

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 271 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

271

NUMBER OF PIPE BRANCHES = 1 BRANCH TEMPERATURE AND PRESSURE PROFILE: Incoming Branch File: VALEBRANCH.TOT Branch Location: at 100 (mi) Distance Elevation Diameter Flow Velocity Press. Gas Temp. Amb Temp. Location (mi) (ft) (in) (MMSCFD) (ft/sec) (psig) (degF) (degF) 0.00 10.00 20.00 40.00 50.00 70.00 80.00

150.00 100.00 125.00 200.00 200.00 200.00 250.00

24.000 24.000 24.000 24.000 24.000 24.000 24.000

200.000 200.000 200.000 200.000 200.000 200.000 200.000

9.81 9.81 9.96 10.14 10.23 10.40 10.51

1163.79 1156.06 1146.13 1125.37 1115.97 1096.89 1085.79

70.00 63.97 61.56 60.24 60.09 60.01 60.01

60.00 60.00 60.00 60.00 60.00 60.00 60.00

Drake

Vale

*************** End of GASMOD Output Report ************* It can be seen from the GASMOD thermal hydraulic analysis report that the inlet pressure at Rockport is approximately 1224 psig, whereas the manual calculation considering isothermal flow yielded an inlet pressure of approximately 1223 psig at the Rockport compressor station. This difference is not very significant. However, in many cases the temperature variation along the pipeline will cause pressures calculated to be significantly different, especially in short pipelines. As an example, if we had a pipeline 100 mi long similar to the pipe section between Rockport and Vale, the thermal hydraulics will show a drastic temperature variation, from 132.24 to 60.65°F. Therefore, an isothermal analysis at 70°F for the entire 100 mi length will show considerable discrepancy in pressures. This is left as an exercise for the reader.

To recap, the manual calculations were based on an isothermal gas flow temperature of 70°F, whereas the thermal hydraulics shows variation of the gas temperature ranging from 132.24°F at the Rockport compressor discharge to 60.01°F at Concord, which is very close to the surrounding soil temperature of 60°F. The compression ratio is 1.52 at the Rockport compressor station, where the 800 psig inlet pressure of the gas is increased to the discharge pressure of 1224 psig. This, in accordance with our previous analysis in Chapter 4 under compressors, causes the discharge temperature of the gas to increase to 132.24°F. If the compression ratio were higher, the discharge temperature of the gas due to compression would have been still higher. The pipeline coating temperature limitation is 140°F and would then require gas cooling in order to avoid damage to the pipe coating. It can be seen from the GASMOD report that the gas flow temperature starts off at 132.24°F at milepost 0 (Rockport) and quickly drops to 67.38°F at milepost 50. This is the exponential temperature decay we discussed in an earlier section. Beginning at milepost 50, the gas temperature starts dropping off more gradually until it almost attains the soil temperature at milepost 240 (Concord). Also, the section of pipe between milepost 170 and Concord is at a fairly constant temperature, close to the soil temperature of 60°F. Therefore, the 70 mi pipeline section between milepost 170 and Concord can be considered to be in isothermal flow, for all practical purposes. A manual calculation of this last 70 mi pipe segment flowing

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 272 Friday, April 1, 2005 3:28 PM

272

GAS PIPELINE HYDRAULICS

at 620 MMSCFD will yield a pressure profile very close to the pressures shown in the GASMOD report. This will be demonstrated next. First, we calculate the Reynolds number from Equation 2.34: 14.7 0.6 × 620 × 10 6 = 21,657,837 R = 0.0004778 60 + 460 0.000008 × 29 Next, using Colebrook Equation 2.39, we calculate the friction factor as 0.0007 2.51 = −2Log10 + f 3.7 × 29 21, 657, 837 f

1

Solving by trial and error, we get f = 0.0096 Therefore, the transmission factor is, using Equation 2.42, F=

2 0.0096

= 20.41

To calculate the compressibility factor Z, the average pressure is required. Since the pressure at milepost 170 is unknown, we will calculate an approximate value of Z using a value of 850 psig for the average pressure. Using CNGA Equation 1.34, we calculate the value of the compressibility factor as Z=

1 = 0.8765 850×344 ,400 (10 )1.785× 0.6 1 + 5203.825

Since there is an elevation difference of 100 (800 − 700) ft between milepost 170 and Concord, we must apply the elevation correction according to Equation 2.7. Using Equation 2.10, the elevation adjustment parameter is 800 − 700 s = 0.0375 × 0.6 = 0.0049 520 × 0.8765 The equivalent length Le from Equation 2.9 is Le =

70(1.0049 − 1) = 70.17 mi 0.0049

Next, using Equation 2.7, we calculate the pressure at milepost 70 as follows: 520 P12 − 1.0049 × 514.72 620 × 10 = 38.77 × 20.41 14.7 0.6 × 520 × 70.17 × 0.8765 6

P1 = 851.59 psia = 836.89 psig

Copyright 2005 by Taylor & Francis Group, LLC

0.5

292.5

2785_C007.fm Page 273 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

273

This compares with 850 psig we assumed earlier. Recalculating the average pressure based on P1 = 851.59 psia and P2 = 514.7 psia, we get, using Equation 2.14, Pavg =

2 851.59 × 514.7 = 696.99 psia 851.59 + 514.7 − 3 851.59 + 514.7

Next, using CNGA Equation 1.34, we recalculate the compressibility factor as Z=

1 682.29×344 ,400 (10 )1.785× 0.6 1 + 5203.825

= 0.8984

Next, using Equation 2.7, we calculate the pressure at milepost 70 as follows: 520 P12 − 1.0049 × 514.72 620 × 10 = 38.77 × 20.41 14.7 0.6 × 520 × 70.17 × 0.8984 6

0.5

292.5

P1 = 858.30 psia = 843.6 psig This value is less than 1% different from the previously calculated value of 836.39 psig. Therefore, we do not have to iterate any further. Comparing the value of 843.6 psig with the pressure of approximately 847 psig from the GASMOD report, we see that we are less than 0.5% apart. Thus, the assumption of isothermal flow in the last 70 mi section of the pipeline is a valid one. We would have been closer still if the 70 mi section had been subdivided into two or more segments and the upstream pressures calculated as discussed in an earlier section. In conclusion, we can state that calculating the pressures and HP in a gas pipeline based on the assumption of constant temperature throughout the pipeline will yield satisfactory answers if the pipeline is long. For shorter pipelines, calculations must be performed by subdividing the pipeline into short segments and taking into account heat transfer between the pipeline gas and the surrounding soil.

7.4 SUMMARY In this chapter we reviewed the thermal effects of pressure drop and horsepower required in a natural gas pipeline system. We pointed out the differences between the results obtained from isothermal and thermal hydraulic analysis. This was illustrated with example problems using an isothermal analysis compared to a more rigorous approach considering heat transfer between the pipeline gas and the surrounding soil. A popular gas pipeline hydraulic simulation software application was used to illustrate the calculation methodology.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 274 Friday, April 1, 2005 3:28 PM

274

GAS PIPELINE HYDRAULICS

PROBLEMS 1. Apply the technique discussed in the temperature variation calculation section to calculate the temperature profile of a gas pipeline 4 mi long, NPS 20, with 0.375 in wall thickness, at a flow rate of 200 MMSCFD. 2. A 200 mile, NPS 24, 0.500 inch wall thickness pipeline from Mobile to Savannah is used for transporting 300 MMSCFD of natural gas (gravity = 0.65 and viscosity = 0.000008 lb/ft-s). The MOP is 1400 psig. The gas inlet temperature and pressure at Mobile are 80°F and 1200 psig, respectively. The soil temperature can be assumed to be 60°F throughout. The delivery pressure required at Savannah is 900 psig. Assume isothermal flow at 70°F. Using the Panhandle B equation with an efficiency of 0.95, calculate the free flow volume with no compressor stations. Compare these results with thermal hydraulic analysis using subdivided pipe segments and heat transfer calculations. The base pressure is 14.7 psia and base temperature is 60°F.

REFERENCES 1. Nayyar, M.L., Piping Handbook, 7th ed., McGraw-Hill, New York, 2000. 2. Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003. 3. McCain, W.D. Jr., The Properties of Petroleum Fluids, Petroleum Publishing Company, Tulsa, OK, 1973. 4. Holman, J.P., Thermodynamics, 2nd ed., McGraw-Hill, New York, 1974. 5. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 6. Engineering Data Book, 10th ed., Gas Processors Suppliers Association, Tulsa, OK, 1994. 7. Pipeline Design for Hydrocarbon Gases and Liquids, American Society of Civil Engineers, New York, 1975. 8. Steady Flow in Gas Pipelines, Contract Report No. 10, July 1965, Pipeline Research Council International, Houston, TX, 1965. 9. McAdams, W.H., Heat Transmission, McGraw-Hill, New York, 1954. 10. Katz, D.L. et al., Handbook of Natural Gas Engineering, McGraw-Hill, New York, 1959.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 275 Friday, April 1, 2005 3:27 PM

CHAPTER

8

Transient Analysis and Case Studies This chapter is devoted to transient pressure analysis and case studies of typical longdistance gas transmission pipelines. The subject of transient pressure analysis is quite complex, and understanding the theory behind it requires delving into differential equations and solution by the method of characteristics. Further, these calculations require some form of computer simulation to arrive at meaningful results. Nevertheless, we will discuss several scenarios that are typical of unsteady flow in gas pipelines that cause transient conditions. The objective is to determine how the pressure varies along the pipeline due to disturbances caused by transient conditions, such as a mainline valve closure and compressor station shutdown. If these transient conditions cause the pipeline pressures at some points to exceed the MOP, measures must be provided to ensure that the pressures do not violate the limits allowed by design codes. For detailed analysis of transient pressures in gas pipelines, the reader should refer to the texts listed in the Reference section of this chapter. We will also be reviewing several real-world pipeline transportation scenarios that encompass the concepts covered in the preceding chapters.

8.1 UNSTEADY FLOW In the preceding chapters, we analyzed pipelines that were in steady-state flow. This means that, at any point in time, the pipelines were operating at constant flow rates with a constant pressure and temperature profile. In other words, if the pressures, temperatures, and flow rates were measured at some point in time, say at 10 a.m. on a certain day, these parameters persisted in values throughout the period under investigation. Therefore, at some other time, such as 12 noon or 5 p.m., the pressure profile, the temperature profile, and the gas flow rates were all the same as that at 10 a.m. In reality, this does not happen. Due to one reason or another, the flow rates and pressures tend to change. This might be due to changes in delivery conditions, such as variation in the amount of gas being received at delivery stations due to changes in operation of facilities that require the gas. Further environmental conditions, such as

275

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 276 Friday, April 1, 2005 3:27 PM

276

GAS PIPELINE HYDRAULICS

atmospheric temperatures, can cause variation in compressor performance, resulting in flow and pressure changes at the compressor discharge. The latter will, in turn, cause changes in pressure and temperature of the pipeline gas. Another reason for unsteady flow condition can be a result of switching or closing valves to divert flow to different customers or shutdown or startup of compressor stations. Such unsteady operation causes transient pressures in the pipeline. Under unsteady or transient flow, the pressures, temperatures, and flow rates become time dependent. This means that it will no longer be correct to use results based upon steady-state flow calculations. 8.1.1

Transient Due to Mainline Valve Closure

To illustrate a transient condition, let us review a simple pipeline system with a head compressor station and a mainline valve at the end of the pipeline, as shown in Figure 8.1. The pipeline has been operating in steady-state condition for a long time. The valve at the end of the pipeline is suddenly closed due to malfunction or human error. Immediately, the pressure at the valve and at points upstream of it starts to rise, as shown by the dashed lines in Figure 8.1. Since gas is compressible, the compressor at the upstream end continues to pump gas without sensing the pressure rise downstream. This will result in an increase in line pack in the downstream section of the pipe, which will progress toward the upstream end. The transient pressure waves moving upstream will eventually reach the discharge of the compressor, causing the discharge pressure to rise. If the increased pressure attains the discharge shutdown setting, the compressor will trip and shut down, producing no further pumping pressure. The blocked-in gas in the pipeline will continue to undergo pressure variation from upstream to downstream as the pressure waves go back and forth at the speed of sound in gas. Eventually, the pressure surge dies out because of friction and loss of inertia resulting from reduction in gas velocity. This is illustrated in another

1400 psig

Hydr

Pressure

aulic

pressure rise with time after valve closure

pres

sure

grad

ient 600 psig

Pipe elevation profile

A Compressor station Figure 8.1

Distance

Transient due to valve closure.

Copyright 2005 by Taylor & Francis Group, LLC

B Terminus

2785_C008.fm Page 277 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

277

G C Head

B

K

H

E

he

s

Sy

ad he or

A

tem

s es pr om

C

ad

F

D

Q Gas flow rate, ACFM Figure 8.2

Compressor performance curve vs. pipeline system head curve.

way in Figure 8.2, using the compressor performance curve and the pipeline system curve. The steady-state system head curve is represented by AB and the compressor performance curve by CD. The steady-state operating point is therefore at E, where the compressor head H matches the pipeline system head required at the flow rate Q. As the valve at the end of the pipeline is closed, the system head curve shifts to the left, indicating reduction in gas flow due to increased pipeline resistance caused by the constriction in the valve. 8.1.2

Transient Due to Compressor Shutdown

Another transient condition that can occur in a simple pipeline described in Figure 8.2 is that of a compressor station shutdown from a steady-state operating condition. Suppose the compressor shuts down in 30 seconds after a long period of steady-state flow. Since there is no pressure being generated at the upstream end of the pipeline, but gas continues to be delivered at the downstream end, the line pack in the pipeline starts reducing starting at the downstream end. The pressures continue to fall along the pipeline and eventually stabilize at some blocked-in pressure. Another slightly complicated compressor station shutdown scenario that causes transient pressures is illustrated in Figure 8.3. In this case, a pipeline with two compressor stations is shown with a hydraulic pressure gradient under steady-state conditions. If the intermediate compressor station shuts down and the gas continues to be pumped from the first compressor station bypassing the second compressor station, the hydraulic pressure gradient will eventually be as indicated in Figure 8.3. However, before steady-state conditions are achieved with only one compressor operating, transient pressures are developed from the point of shutdown of the intermediate compressor station. Suppose the initial flow rate with both compressors operating under steady-state conditions is indicated by a flow rate of Q. The compressor

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 278 Friday, April 1, 2005 3:27 PM

278

GAS PIPELINE HYDRAULICS

Hydraulic pressure gradient with one compressor station 1400 psig

1400 psig

900 psig 800 psig

Flow

Sheridan

Compton

Beaumont

NPS 16 pipeline 200 mi long Figure 8.3

Transient due to compressor station shutdown.

Head

ead

K

H

st em

C

sor h

he

G

pres

B

Sy

Com

ad

performance curve superimposed on the pipeline system head curve for the pipe segment between the two compressor stations is shown in Figure 8.4. Initially, the system head curve AB for the pipe segment 1 between the two compressor stations results in a flow rate of Q with the operating point at E. At this point, the compressor head H of the first compressor station matches the pipeline system head required at the flow rate Q. When the second compressor station shuts down, it no longer provides the discharge pressure to boost the gas in pipe segment 2. Therefore, the first compressor station has to push the gas all the way to the end of the pipeline. It therefore has to contend with a longer pipe segment, which has a system head curve FG as shown in Figure 8.4. It can be seen that the new operating point K is at

E

00

,0

15 R

F

PM

D

00

,0

12

A

Q1

Q

Gas flow rate, ACFM Figure 8.4

Transient due to intermediate compressor shutdown.

Copyright 2005 by Taylor & Francis Group, LLC

PM

R

Q2

2785_C008.fm Page 279 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

279

a reduced flow rate Q1. If the point K on the compressor head curve is at too high a discharge pressure, the control mechanism will signal the compressor to slow down in speed. Thus, if the original compressor curve was based on 15,000 rpm, the compressor would slow down to a speed such as 12,000 rpm. This results in a new operating point, L, corresponding to a flow rate Q2, as shown in Figure 8.4. In summary, shutting down the second compressor station causes the operating point to move from point E on the compressor head curve at 15,000 rpm down to point L on the compressor head curve at 12,000 rpm. Correspondingly, the flow rate will decrease from Q at point E to Q2 at point L. 8.2 CASE STUDIES In the next few pages of this chapter, we are going to look at some real-life gas transmission pipeline systems. We will be applying the concepts learned in the previous chapters to determine the pressures and flow rates required in various scenarios. 8.2.1

Offshore Pipeline Case

Consider a gas production facility located offshore. The gas is compressed from the offshore platform through submarine pipelines that go ashore and subsequently connect to onshore pipelines for transportation of gas to industrial consumers. We will look at sizing such an offshore and onshore piping system for transporting a given quantity of gas. Calculations will be performed considering different options such as the AGA equation and Panhandle equations. We will illustrate this using an example. Case Study 1—Offshore/Onshore Pipeline A natural gas pipeline system originates at an offshore facility that compresses the gas through 200 mi of NPS 30, 0.625 in. wall thickness submarine pipelines to an onshore location, as depicted in Figure 8.5. A compressor station located onshore is used to compress the gas through a 120 mi, NPS 24, 0.500 inch wall thickness onshore buried pipeline for eventual delivery to a power plant. Determine the maximum flow rate possible under the following conditions. Neglect elevation effects. The compression ratio is 1.5. Use the Weymouth equation with 95% efficiency. Assume a base pressure of 14.7 psia and base temperature of 60°F. The gas flowing temperature is 60 °F and the compressibility factor is 0.88. The gas gravity is 0.65. 1480 psig

Compressor Station 500 psig 120 mNPS 24

200 mNPS 30

Figure 8.5

Offshore/onshore pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 280 Friday, April 1, 2005 3:27 PM

280

GAS PIPELINE HYDRAULICS

(a) Gas pressure at the platform equals 1480 psig and free flow occurs without use of any compression offshore or onshore. The delivery pressure at the power plant is 500 psig. (b) Considering the MOP at the platform and onshore equal to 1480 psig, determine the maximum throughput possible with compression offshore and onshore. Solution (a) Free flow with 1480 psig at the offshore platform. Using Weymouth Equation 2.52, neglecting elevation effects, calculate the pressure at the beginning of the NPS 24 onshore pipeline. The pipe inside diameter = 24 – 2 × 0.500 = 23 in. P12 − 514.72 60 + 460 Q = 433.5 × 0.95 14.7 0.65 × 520 × 120 × 0.88

0.5

232.667

(8.1)

28.752.667

(8.2)

Similarly, considering the NPS 30 pipeline 200 mi long, Pipe inside diameter = 30 − 2 × 0.625 = 28.75 in. 60 + 460 1494.72 − P12 Q = 433.5 × 0.95 14.7 0.65 × 520 × 200 × 0.88

0.5

Eliminating Q from both Equation 8.1 and Equation 8.2 by division, we get P 2 − 514.72 1= 1 2 2 1494.7 − P1

0.5

200 120

0.5

23.0 28.75

2.667

(8.3)

Solving for the pressure P1 at the junction of the two pipes onshore, P1 = 1253.7 psia = 1239 psig Next, substituting this value of P1 in Equation 8.1, we calculate the free flow volume flow rate as 520 1253.72 − 514.72 Q = 433.5 × 0.95 14.7 0.65 × 520 × 120 × 0.88

0.5

232.667

or Q = 377.53 MMSCFD Therefore, without any compression, the free flow possible is 377.53 MMSCFD.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 281 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

281

(b) With compressors installed at onshore and offshore locations, each location will be delivering at an MOP of 1480 psig. With a compression ratio of 1.5, the suction pressure at the onshore compressor is Ps =

1480 + 14.7 = 996.47 psia = 981.77 psig 1.5

First, calculate the capacity of the NPS 24 onshore pipeline, considering 1480 psig at the upstream end and 500 psig at the downstream end 120 mi away. Using Weymouth Equation 2.52, 520 1494.72 − 514.72 Q = 433.5 × 0.95 × 14.7 0.65 × 520 × 120 × 0.88

0.5

232.667

Q = 463.43 MMSCFD Next, we must determine if the offshore NPS 30 pipeline can transmit this flow starting at 1480 psig at the offshore platform and with a downstream pressure of 981.77 psig calculated earlier. 520 1494.72 − 996.472 Q = 433.5 × 0.95 × 14.7 0.65 × 520 × 200 × 0.88

0.5

(28.75)2.667

Solving for Q, we get Q = 516.76 MMSCFD Thus, the NPS 30 submarine pipeline has a capacity of 516.76 MMSCFD, whereas the onshore NPS 24 pipeline has a capacity of only 463.43 MMSCFD. Picking the lower of the two flow rates, the maximum throughput possible with the onshore compressor is 463.43 MMSCFD.

Case Study 2—Gas Gathering System and Trunk Line to Power Plant Natural gas gathered from the San Juan gas fields is collected at Chico and transported through a DN 800, 15 mm wall thickness pipeline system, 420 km long, that ties into another DN 800, 15 mm wall thickness gas transmission pipeline at Rio for eventual delivery to a power plant at Madera, as shown in Figure 8.6. Chico is at an elevation of 2100 m, whereas Rio and Madera are at 1650 m and 3100 m, respectively. The length of the pipeline from Rio to Madera is 280 km. The required delivery pressure at Madera is 35 Bar gauge. The gas gravity and viscosity are 0.65 and 0.012 cP, respectively. The gas inlet temperature at Chico is 20 °C, and the pressure is 40 Bar gauge. Assume a constant gas flow temperature of 20 °C. The pipeline MOP is 100 Bar gauge. The base temperature and base pressure are 15 °C and 1 Bar absolute, respectively. Use the Panhandle B equation with a pipeline efficiency of 95%. Assume a gas compressibility factor of 0.85 throughout.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 282 Friday, April 1, 2005 3:27 PM

282

GAS PIPELINE HYDRAULICS

1.5 Mm3/day Rio Elev: 1650 m 0 km

0B

DN

Flo

°C 20

ay

w

San Juan Chico

: mp

/d

40 Bar gauge 20°C

0

3

m

80

M

10 P:

MO

4

28

DN

g ar

6

m

k 20

ge

au

0

80

te

35 Bar gauge Madera Elev: 3100 m

Elev: 2100 m Figure 8.6

Chico–Rio pipeline to Madera power plant.

(a) Determine the compressor station power required to deliver 6 Mm3/day at Madera. Use 80% isentropic efficiency and a specific heat ratio of 1.4. (b) What modifications are required to provide gas volumes of 1.5 Mm3/day for an industrial consumer at Rio in addition to that required at Madera? (c) What pipeline capacity can be expected if all compressor stations are shut down and free flow occurs from Chico to Rio and Madera? Ignore deliveries at Rio and assume all gas flows to Madera. Solution Assume initially that one compressor station at Chico will be able to transport 6 Mm3/day to Madera. 1 Bar = 100 kPa Pipe inside diameter = 800 – 2 × 15 = 770 mm The elevation adjustment parameter from Equation 2.11 for Rio to Madera is 3100 − 1650 = 0.2589 s = 0.0684 × 0.65 (20 + 273)0.85 e s = e 0.2589 = 1.2954 The equivalent length from Equation 2.9 is Le = 280 ×

Copyright 2005 by Taylor & Francis Group, LLC

(1.2954 − 1) = 319.52 km 0.2589

2785_C008.fm Page 283 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

283

Using Panhandle B Equation 2.60, considering elevation difference, first for the Rio to Madera pipe segment, we get 15 + 273 6 × 10 6 = 1.002 × 10 −2 × 0.95 × 100

1.02

P12 − 1.2954(3600)2 × 0.961 × 293 × 319.52 × 0.85 0.65

0.51

770 2.53

Solving for the pressure at Rio, P1 = 4818 kPa Next, using this pressure as the downstream pressure for the 420 km pipe segment from Chico to Rio, we get 1650 − 2100 Elevation adjustment parameter s = 0.0684 × 0.65 = −0.0803 (20 + 273)0.85 e s = e −0.0803 = 0.9228 The equivalent length is Le = 420 ×

(0.9228 − 1) = 403.74 km −0.0803

15 + 273 6 × 10 6 = 1.002 × 10 −2 × 0.95 × 100

1.02

P12 − 0.9228 × 48182 × 0.961 × 293 × 403.74 × 0.85 0.65

0.51

770 2.53

Solving for the upstream pressure at Chico, we get P1 = 5435 kPa = 54.35 Bar absolute Since the inlet pressure at Chico is 40 Bar gauge, the compression ratio required at Chico is Compression ratio =

54.35 = 1.33 40 + 1

The compressor station power required is calculated from Equation 4.16 as follows: 0.4 1.40 1 + 0.85 1 Power = 4.0639 × 6 (293) (1.33) 1.4 − 1 0.4 2 0.8

or Power = 2455 KW

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 284 Friday, April 1, 2005 3:27 PM

284

GAS PIPELINE HYDRAULICS

When the Rio delivery of 1.5 Mm3/day is included, we calculate the upstream pressure at Chico for the 420 km segment as follows: 288 7.5 × 10 6 = 1.002 × 10 −2 × 0.95 × 100

1.02

P12 − 0.9228 × 48182 × 0.961 × 293 × 403.74 × 0.85 0.65

0.51

770 2.53

By proportion, we get 7.5 P 2 − 0.9228 × 48182 = 1 2 6.0 5435 − 0.9228 × 48182 Solving for P1, we get P1 = 5831 kPa = 58.31 Bar absolute The new compression ratio becomes Compression ratio =

58.31 = 1.42 40 + 1

The new power required at the Chico compressor station is 0.4 1.40 1 + 0.85 1 (293) Power = 4.0639 × 7.5 (1.42) 1.4 − 1 = 3808 KW 0.4 2 0.8

(c) When the compressor station at Chico is shut down, the pressure available is only 40 Bar or 4000 kPa. Using this upstream pressure and considering the entire (420 + 280) km = 700 km pipeline from Chico to Madera, the free flow capability is calculated using the Panhandle B equation by considering the elevation changes in two steps. From Equation 2.12 and Equation 2.13, for the 420 km segment the elevation falls from 2100 m to 1650 m and s = − 0.0803 and es = 0.9228 (as calculated earlier) From Equation 2.12, j1 =

0.9228 − 1 = 0.9614 − 0.0803

L1 = 420 km Similarly, for the 280 km second segment of the pipeline, the elevation rises from 2100 m to 3100 m, measured from Chico. s = 0.1785

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 285 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

285

and es = 1.1954 j2 =

1.1954 − 1 = 1.095 0.1785

L2 = 280 km From Equation 2.13, the equivalent length is Le = 0.9614 × 420 + 1.095 × 280 × 0.9228 = 686.72 km For the entire line, 3100 − 2100 = 0.1785 s = 0.0684 × 0.65 293 × 0.85 e s = e 0.1785 = 1.1954 Applying the Panhandle B equation for the entire pipeline, we get 288 Q = 1.002 × 10 −2 × 0.95 × 100

1.02

4100 2 − 1.1954 × 3600 2 0.961 × 293 × 686.72 × 0.85 0.65

51 0.5

770 2.53

Q = 1,926,314 m3/day = 1.93 Mm3/day Thus, with the Chico compressor station shut down, the free flow throughput is Q = 1.93 Mm3/day Obviously, this is inadequate to feed the Madera power plant that requires 6 Mm3/day.

Case Study 3—Fairfield to Beaumont and Travis Pipeline A natural gas pipeline, NPS 24, is being built from the gas fields at Fairfield (elevation 610 ft) to transport gas to a 400 MW power plant at Beaumont (elevation 350 ft) 280 mi away, as illustrated in Figure 8.7. Along the way at Mavis (milepost 50, elevation 1200 ft), an industrial consumer requires 10 MMSCFD, and a small community at Mayberry (milepost 110, elevation 1800 ft) requires natural gas for a municipal gas distribution system with a city gate pressure of 600 psig and 20 MMSCFD. During the first 2 years of operation, the gas flow requirements are as follows: Mavis: 10 MMSCFD at 300 psig Mayberry: 20 MMSCFD at 600 psig Beaumont: 100 MMSCFD at 400 psig Total: 130 MMSCFD

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 286 Friday, April 1, 2005 3:27 PM

286

GAS PIPELINE HYDRAULICS

400 psig D

F SC

M 0M

10

20 MMSCFD

20 m

i NPS

10 MMSCFD 0m

Mayberry 0 Elev: 1800 ft 3 S m.p. 110 NP

Beaumont Elev: 350 ft

60 MMSCFD 350 psig 16 Travis Elev: 420 ft

28

Mavis Elev: 1200 ft m.p. 50

500 psig 70°F Fairfield Elev: 610 ft Figure 8.7

Fairfield to Beaumont and Travis pipeline.

At the end of the second year, a 240 MW power plant at Travis (elevation 420 ft) will come on stream and require a gas delivery of 60 MMSCFD at 350 psig. This requires a total pipeline capacity of 190 MMSCFD out of Fairfield. The gas pressure and temperature at the inlet to the pipeline are 500 psig and 70°F. The soil temperatures can be assumed to be as follows: Fairfield to Mavis: 60°F Mavis to Mayberry: 50°F Mayberry to Beaumont: 70°F The branch pipe to Travis starts at the Travis junction (milepost 200, elevation 750 ft) and extends 20 mi to the Travis power plant. It is an NPS 16, 0.250 in. wall thickness pipe. It is anticipated that API 5LX-70 material will be used for the pipe. The cost of pipe material is $1200 per ton for pipe coated, wrapped, and delivered to the field. Construction cost of the pipeline can be estimated at $20,000 per in.-diameter mi. Compressor stations cost is $2000 per installed HP. Mainline valves are to be installed at 20 mi intervals and cost $100,000 per site. Receipt and delivery meters at Fairfield, Mavis, Mayberry, Travis, and Beaumont are expected to cost as follows: Fairfield meter: $500,000 Mavis meter: $200,000 Mayberry meter: $250,000 Travis meter: $300,000 Beaumont meter: $350,000 Fuel consumption can be estimated at 0.2 MCF per day per HP. Fuel gas cost is $4 per MCF. Assume base pressure = 14.7 psia and base temperature = 60°F. The MOP of the pipeline is 1440 psig. Use the General Flow equation with a Colebrook friction factor and the CNGA equation for the compressibility factor.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 287 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

(a) Determine the pipe wall thickness required for the specified MOP. (b) Determine the locations and HP of the compressor stations necessary for the first 2 years (phase 1) and after that (phase 2). (c) Estimate the total capital cost of pipeline, compressor stations, and other facilities for phase 2. Solution During the phase 1 operation, we will calculate the pressures and HP required, considering 100 MMSCFD delivery to the Beaumont power plant at 400 psig. Since the MOP is 1440 psig, the minimum wall thickness needed for the class 1 location is calculated from Equation 6.8: 1440 =

2t × 70, 000 × 0.72 24

Solving for t, we get Wall thickness t = 0.343 in. or Use 0.375 in. standard size pipe. Inside diameter D = 24 – 2 × 0.375 = 23.25 in. First, calculate the upstream pressure at milepost 110, assuming a gas flow temperature at 70°F and compressibility factor of 0.85. The Reynolds number from Equation 2.34 is 14.7 0.6 × 100 × 10 6 Re = 0.0004778 = 4, 357,109 520 8 × 10 −6 × 23.25 Assuming an internal roughness e = 0.0007 in. and using Equation 2.45, we get the transmission factor F as follows: 0.0007 1.255F F = −4 log10 + 3.7 × 23.25 4.3571 × 10 6 Solving by successive iteration, F = 19.45 The upstream pressure at milepost 110 is calculated from General Flow Equation 2.4, considering the elevation difference. 350 − 1800 Elevation adjustment parameter s = 0.0375 × 0.6 = −0.0724 530 × 0.85 e s = e −0.0724 = 0.9301

Copyright 2005 by Taylor & Francis Group, LLC

287

2785_C008.fm Page 288 Friday, April 1, 2005 3:27 PM

288

GAS PIPELINE HYDRAULICS

and Le = 170 ×

(0.9301 − 1) = 164.03 mi −0.0724

520 P12 − 0.9301 × 414.72 100 × 10 = 38.77 × 19.45 14.7 0.6 × 530 × 164.03 × 0.85

0.5

6

23.252.5

Solving for P1, we get P1 = 501.7 psia = 487.0 psig The average pressure in the pipe segment is, by Equation 2.14, Pavg =

501.7 × 414.7 2 501.7 + 414.7 − = 459.6 psia = 444.9 psig 501.7 + 414.7 3

We will confirm the value of the compressibility factor Z we used earlier, using CNGA Equation 1.34: Z=

1 (10)1.785× 0.6 1 + 444.9×344,4003×.825 (530)

or Z = 0.9359 This value of Z is way off compared to the 0.85 value we used in our calculations. Recalculating P1 using the recent value of Z, we obtain 520 P12 − 0.9301 × 414.72 100 × 10 6 = 38.77 × 19.45 14.7 0.6 × 530 × 164.03 × 0.9359

0.5

23.252.5

Solving for P1 by proportion, we get 501.72 − 0.9301 × 414.72 P12 − 0.9301 × 414.72 = 0.85 0.9359 or P1 = 510.86 psia Recalculating the average pressure and the new compressibility factor Z, we find Pavg =

510.86 × 414.7 2 510.86 + 414.7 − = 464.45 psia = 449.75 psig 510.86 + 414.7 3

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 289 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

289

and Z=

1

1+

449.75×344,400×(10)1.785× 0.6 (530)3.825

or Z = 0.9352 The percentage difference between this value of Z compared to the previously calculated value is 0.9352 − 0.9359 = −0.07 % 0.9359 This is good enough, and we won’t iterate any further. Therefore, Pressure at Mayberry takeoff (milepost 110) = 510.86 psia = 496.16 psig Next, calculate the pressure at milepost 50, considering the pipe segment between Mavis and Mayberry at 50°F flowing temperature and a flow rate of 120 MMSCFD. The Reynolds number, by proportion, is R = 4, 357,109 ×

120 = 5, 228, 531 100

The transmission factor F is calculated from 0.0007 1.255F F = −4 log + 3.7 × 23.25 5.2285 × 10 6 Solving by iteration, F = 19.57 The upstream pressure at milepost 50 is found from General Flow Equation 2.4, considering the elevation difference, as follows: 1800 − 1200 Elevation adjustment parameter s = 0.0375 × 0.6 = 0.0311 510 × 0.85 e s = e 0.0311 = 1.0316 and Le = 60 ×

(1.0316 − 1) = 60.96 mi 0.0311

520 P12 − 1.0316 × 510.862 120 × 10 = 38.77 × 19.57 14.7 0.6 × 510 × 60.96 × 0.85 6

Copyright 2005 by Taylor & Francis Group, LLC

0.5

23.252.5

2785_C008.fm Page 290 Friday, April 1, 2005 3:27 PM

290

GAS PIPELINE HYDRAULICS

Solving for P1, we get P1 = 562.03 psia = 547.33 psig Calculating the average pressure, Pavg =

562.03 × 510.86 2 562.03 + 510.86 − = 536.85 psia = 522.15 psig 562.03 + 510.86 3

Recalculating the compressibility factor Z using the new average pressure, we get Z=

1 1.785× 0.6

(10) 1 + 522.15×344,4003×.825 (510)

or Z = 0.9147 Next, we recalculate the pressure at milepost 50: s = 0.0311 ×

0.85 = 0.0289 0.9147

e s = 1.0293 and Le = 60 ×

0.0293 = 60.88 mi 0.0289

520 P12 − 1.0293 × 510.862 120 × 10 6 = 38.77 × 19.57 14.7 0.6 × 510 × 60.88 × 0.9147

0.5

23.252.5

Solving for P1 by proportion, we get 562.032 − 1.0316 × 510.862 P12 − 1.0293 × 510.862 = 0.85 × 60.96 0.9147 × 60.88 P1 = 564.59 psia = 549.9 psig The average pressure and Z are calculated next: Pavg =

564.59 × 510.86 2 564.59 + 510.86 − = 538.33 psia = 523.63 psig 564.59 + 510.86 3

and Z=

Copyright 2005 by Taylor & Francis Group, LLC

1 (10)1.785× 0.6 1 + 523.63×344,4003×.825 (510)

2785_C008.fm Page 291 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

291

or Z = 0.9145 This is not too far from the previously calculated Z value of 0.914. Therefore, we will not iterate any further. The pressure at milepost 50 is P1 = 564.59 psia = 549.9 psig Next, calculate the upstream pressure at Fairfield, considering the 50 mi pipe segment flowing 130 MMSCFD at 60°F. The Reynolds number from Equation 2.34 is 14.7 0.6 × 130 × 10 6 Re = 0.0004778 = 5, 664, 242 520 8 × 10 −6 × 23.25 Using Equation 2.45, we get the transmission factor F as follows: 0.0007 1.255F F = −4 log10 + 3.7 × 23.25 5.6642 × 10 6 Solving for F, we get F = 19.61 The elevation adjustment is s = 0.0375 × 0.6

1200 − 610 = 0.0284, where Z = 0.9 is assumed. 520 × 0.90 e s = 1.0288

and Le = 50 ×

0.0288 = 50.65 mi 0.0284

Using the General Flow equation, we calculate the upstream pressure P1 as 520 P12 − 1.0288 × 564.592 130 × 10 = 38.77 × 19.61 14.7 0.6 × 520 × 50.65 × 0.9 6

0.5

23.252.5

Solving for the pressure at Fairfield, we get P1 = 613.89 psia = 599.2 psig The average pressure and Z are calculated next: Pavg =

613.89 × 564.59 2 613.89 + 564.59 − = 589.58 psia = 574.88 psig 613.89 + 564.59 3

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 292 Friday, April 1, 2005 3:27 PM

292

GAS PIPELINE HYDRAULICS

and Z=

1 (10)1.785× 0.6 1 + 574.88×344,4003×.825 (520)

or Z = 0.913 Recalculating, the elevation adjustment is s=

0.0284 × 0.9 = 0.028 0.913

e s = 1.0284 and Le = 50 ×

1.0284 − 1 = 50.70 mi 0.028

Recalculating the pressure at Fairfield, using the General Flow equation, we get 520 P12 − 1.0284 × 564.592 130 × 10 6 = 38.77 × 19.61 14.7 0.6 × 520 × 50.7 × 0.913

0.5

23.252.5

By proportions, P12 − 1.0284 × 564.592 613.892 − 1.0288 × 564.592 = 50.7 × 0.913 50.65 × 0.90 or P1 = 614.40 psia = 599.7 psig Recalculating, the average pressure and Z are Pavg =

614.4 × 564.59 2 614.4 + 564.59 − = 589.85 psia = 575.15 psig 614.4 + 564.59 3

and Z=

1 (10)1.785× 0.6 1 + 575.15×344,4003×.825 (520)

or Z = 0.913, which is the same as before. Therefore, the pressure at Fairfield is P1 = 614.40 psia = 599.7 psig

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 293 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

293

The HP required is calculated from Equation 4.15: 0.4 1.4 1 + 0.913 1 614.4 1.4 1 HP = 0.0857 × 130 520 − 0.8 514.7 2 0.4

or HP = 1258 for phase 1 For phase 2, the inlet volume at Fairfield increases to 190 MMSCFD. The pressure at milepost 200 will be calculated considering the pipe segment from milepost 200 to Beaumont at 100 MMSCFD. From earlier calculations, F = 19.45 350 − 750 s = 0.0375 × 0.6 = −0.0189 530 × 0.9 e s = 0.9813 and Le = 80 ×

0.9813 − 1 = 79.15 mi −0.0189

We will assume Z = 0.9 throughout for simplicity. The pressure at milepost 200 is found from the General Flow equation 520 P12 − 0.9813 × 414.72 100 × 10 6 = 38.77 × 19.45 14.7 0.6 × 530 × 79.15 × 0.9

0.5

23.252.5

Solving for P1, we get P1 = 464.35 psia Next, calculate the pressure at milepost 110, considering 90 mi of pipe at 70°F with a flow of 160 MMSCFD. We will assume F = 19.45 and Z = 0.9 throughout for simplicity. 750 − 1800 s = 0.0375 × 0.6 = −0.0495 530 × 0.9 e s = 0.9517 and Le = 90 ×

Copyright 2005 by Taylor & Francis Group, LLC

0.9517 − 1 = 87.86 mi −0.0495

2785_C008.fm Page 294 Friday, April 1, 2005 3:27 PM

294

GAS PIPELINE HYDRAULICS

The pressure at milepost 110 is found from the General Flow equation 520 P12 − 0.9517 × 464.352 160 × 10 = 38.77 × 19.45 14.7 0.6 × 530 × 87.86 × 0.9

0.5

23.252.5

6

Solving for P1, we get P1 = 581.7 psia Next, calculate the pressure at milepost 50, considering 60 mi of pipe at 50°F with a flow of 180 MMSCFD. 1800 − 1200 s = 0.0375 × 0.6 = 0.0294 510 × 0.9 e s = 1.0298 and Le = 60 ×

1.0298 − 1 = 60.92 mi 0.0294

Therefore, the pressure at milepost 50 is found from the General Flow equation 520 P12 − 1.0298 × 581.72 180 × 10 6 = 38.77 × 19.45 14.7 0.6 × 510 × 60.92 × 0.9

0.5

23.252.5

Solving for P1, we get P1 = 678.04 psia Finally, calculate the pressure at Fairfield considering 50 mi of pipe at 60°F with a flow rate of 190 MMSCFD. 1200 − 610 s = 0.0375 × 0.6 = 0.0284 520 × 0.9 e s = 1.0288 and Le = 50 ×

Copyright 2005 by Taylor & Francis Group, LLC

1.0288 − 1 = 50.65 mi 0.0284

2785_C008.fm Page 295 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

295

Therefore, the pressure at Fairfield is found from the General Flow equation 520 P12 − 1.0288 × 678.04 2 190 × 10 6 = 38.77 × 19.45 14.7 0.6 × 520 × 50.65 × 0.9

0.5

23.252.5

Solving for P1, we get P1 = 761.04 psia The HP required at Fairfield for phase 2 is calculated from Equation 4.15: 0.4 1.4 1 + 0.9 1 761.04 1.4 1 HP = 0.0857 × 190 520 − 2 0.88 514.7 0.4

or HP = 4161 for phase 2 The capital cost is calculated next. The weight per foot of NPS 24 pipe is calculated using Equation 6.11: w = 10.68 × 0.375 × (24 – 0.375) = 94.62 lb/ft Similarly, for the Travis branch, the weight per foot of NPS 16 pipe is calculated using Equation 6.11: w = 10.68 × 0.25 × (16 – 0.25) = 42.05 lb/ft The tonnage for 280 mi of NPS 24 pipe is Tons =

94.62 × 5280 × 280 = 69, 943 2000

The tonnage for 20 mi of NPS 16 pipe is Tons =

42.05 × 5280 × 20 = 2220 2000

Total pipe cost = $1200 × (69,943 + 2220) = $86.61 million The installation cost of the pipe is calculated next: Installation cost = $20,000 × (24 × 280 + 16 × 20) = $140.8 million The installation cost of the compressor station for phase 2 is Compressor cost = $2000 × 4161 = $8.33 million

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 296 Friday, April 1, 2005 3:27 PM

296

GAS PIPELINE HYDRAULICS

Considering mainline valves at 20 mi intervals, the total number of valves required for both the main line and the Travis branch is Number of valves =

280 20 + 1+ + 1 = 17 20 20

Total cost of valves = $100,000 × 17 = $1.7 million Total cost of all meter stations = ($500 + $200 + $250 + $300 + $350) thousand = $1.6 million Therefore, Total capital cost of all facilities = ($140.8 + $8.33 + $1.7 + $1.6 + $86.61) million = $239.04 million To account for other items and indirect costs, increase the above by 30%: Total capital cost = $239.04 × 1.3 = $310.75 million.

8.3 SUMMARY This chapter reviewed some elementary concepts of transient pressures caused by valve closures and compressor station shutdown. Since the calculation methodology of transient pressures and flow rates requires the solution of partial differential equations and manual calculation is quite laborious, we refer the reader to an advanced text that specializes in this area of hydraulics. We also covered several real-world pipeline case studies. PROBLEMS 1. A natural gas pipeline system from an offshore facility is used to compress natural gas through 120 mi of NPS 24, 0.375 inch wall thickness pipe to an onshore location, similar to Figure 8.5. The compressor station located onshore is used to pump the gas through an 80 mi, NPS 20, 0.375 inch wall thickness pipe to a power plant. Determine the maximum flow rate possible under the following conditions. Neglect elevation effects. The compression ratio is 1.5. Use the Weymouth equation with 95% efficiency. Assume a base pressure of 14.7 psia and base temperature of 60°F. The gas flowing temperature is 60°F, and the compressibility factor is 0.9. The gas gravity is 0.6. a. The gas pressure at the platform equals 1440 psig and free flow occurs without use of any compression offshore or onshore. The delivery pressure at the power plant is 500 psig. b. Considering the MOP at the platform and onshore equal to 1440 psig, determine the maximum throughput possible with compression offshore and onshore.

2. Natural gas gathered from the Blanco fields is collected at Tapas and transported through a DN 600, 10 mm wall thickness pipeline system, 240 km long, that ties

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 297 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

297

into another DN 600, 10 mm wall thickness gas transmission pipeline at Rojas for eventual delivery to a power plant at Montecito. Tapas is at an elevation of 1500 m, whereas Rojas and Montecito are at 650 m and 300 m, respectively. The length of the pipeline from Rojas to Montecito is 140 km. The required delivery pressure at Montecito is 40 Bar gauge. The gas gravity and viscosity are 0.60 and 0.012 cP, respectively. The gas inlet temperature at Tapas is 20°C and the pressure is 40 Bar gauge. Assume a constant gas flow temperature of 20°C. The pipeline MOP is 100 Bar gauge. The base temperature and base pressure are 15°C and 1 Bar absolute, respectively. Use the Panhandle B equation with a pipeline efficiency of 95%. Assume a gas compressibility factor of 0.90 throughout. a. Determine the compressor station power required to deliver 4 Mm3/day at Montecito. Use 80% isentropic efficiency and a specific heat ratio of 1.4. b. What pipeline capacity can be expected if all compressor stations are shut down and free flow occurs from Tapas to Rojas and Montecito?

REFERENCES 1. Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003. 2. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 3. Engineering Data Book, 10th ed., Gas Processors Suppliers Association, Tulsa, OK, 1994. 4. Pipeline Design for Hydrocarbon Gases and Liquids, American Society of Civil Engineers, New York, 1975. 5. Steady Flow in Gas Pipelines, Contract Report No. 10, July 1965, Pipeline Research Council International, Houston, TX, 1965. 6. Katz, D.L. et al., Handbook of Natural Gas Engineering, McGraw-Hill, New York, 1959. 7. Wylie, E. B. and Streeter, V.L., Fluid Transients in Systems, Prentice Hall, New York, 1993.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 298 Friday, April 1, 2005 3:27 PM

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 299 Friday, April 1, 2005 3:12 PM

CHAPTER

9

Valves and Flow Measurements

In this chapter we will discuss the various types of valves and flow measurements used on gas pipelines. The design and construction codes for valves, materials of construction, and application of the different types of valves and their performance characteristics will be explained. The importance of flow measurement in a gas pipeline and the accuracy of available instruments, codes, and standards used will be discussed. Various American National Standards Institute (ANSI), American Petroleum Institute (API), and American Gas Association (AGA) formulas used in connection with orifice meters will be reviewed. Since a small error in measurement in gas flow in a pipeline can translate to several thousand dollars of loss of revenue, it is important that industry strives to improve upon measurement methods. Accordingly, gas transportation companies and related industries have been researching better ways to improve flow measurement accuracy. For a detailed discussion of gas flow measurement, the reader is referred to the publications listed in the Reference section.

9.1 PURPOSE OF VALVES Valves are installed on pipelines and piping systems to isolate sections of piping for maintenance, to direct the fluid from one location to another, to shut down flow through pipe sections, and to protect pipe and prevent loss of fluid in the event of a rupture. On long-distance pipelines transporting natural gas and other compressible fluids, design codes and regulatory requirements dictate that sections of pipeline be isolated by installing mainline block valves at certain fixed spacing. For example, DOT 49 CFR, Part 192 requires that in class 1 locations, mainline valves be installed 20 mi apart. Class locations were discussed in Chapter 6.

299

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 300 Friday, April 1, 2005 3:12 PM

300

GAS PIPELINE HYDRAULICS

Grade Steel sleeve size as required

Pad or saddle

Sectionalizing or lateral control valve Figure 9.1

3 concrete cube

Mainline valve installation. (Reproduced from Katz et al., Handbook of Natural Gas Engineering, McGraw-Hill, New York, 1959. With permission.)

A typical mainline block valve installation on a gas transmission pipeline is illustrated in Figure 9.1.

9.2 TYPES OF VALVES The various types of valves used in the gas pipeline industry include the following: • • • • • • • • •

Gate valve Ball valve Plug valve Butterfly valve Globe valve Check valve Control valve Relief valve Pressure regulating valve

Each of these valves listed will be discussed in detail in the following sections. Valves can be of screwed design, welded ends, or flanged ends. In the gas industry, large valves are generally of the welded type, in which the valve is attached to the pipe on either side by a welded joint to prevent gas leakage to the atmosphere.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 301 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

Figure 9.2

301

Mainline block valve.

In smaller sizes, screwed valves are used. A typical welded end mainline valve, along with smaller valves on either side, is shown in Figure 9.2. Valves may be operated manually using a hand wheel or using an electric, pneumatic, or gas operator, as shown in Figure 9.3.

Bypass Operator Bypass ball valve

Grade

Riser Extension

Mainline valve

Figure 9.3

Valve with motor operator. (Reproduced from Nayyar, M.L., Piping Handbook, McGraw-Hill, New York, 2000. With permission.)

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 302 Friday, April 1, 2005 3:12 PM

302

GAS PIPELINE HYDRAULICS

Table 9.1 ANSI Pressure Ratings for Valves Class

Allowable Pressure, psi

150 300 400 600 900 1500

275 720 960 1440 2160 3600

9.3 MATERIAL OF CONSTRUCTION Most valves used in gas pipelines are constructed of steel and conform to specifications such as API, ASME, and ANSI standards. For certain gases that are corrosive and require certain special properties, some exotic materials can be used. The next section lists applicable standards and codes used in the design and construction of valves and fittings on gas pipelines. The valve trim material, which refers to the various working parts of a valve such as the stem, wedge, and disc, are constructed of many different materials depending upon the pressure rating and service. Valve manufacturers designate their products using some form of a proprietary numbering system. However, the purchaser of the valve must specify the type of material and operating conditions required. A typical gate valve specification might be as follows: NPS 12, ANSI 600 gate valve, cast steel flanged ends rising stem 13% CR, single wedge CS, stellite faced, seat rings SS 304, ABC company #2308. Valve operators may consist of a hand wheel or lever that is attached to the stem of the valve. Gear systems are used for larger valves. Electric motor operated valves are quite commonly used in gas pipeline systems, as are gas and pneumatic operators. Many valves can be buried, resulting in a portion of the valve and the operator above ground. The pressure rating of a valve represents the internal pressure that the valve can be subject to under normal operating conditions. For example, an ANSI 600 rating refers to a valve that can be safely operated at pressures up to 1440 psig. Most gas pipelines are operated around this pressure rating. Table 9.1 shows the ANSI pressure ratings for valves and pipes. If a valve is designated as an ANSI 600 rated valve, the manufacturer of the valve must hydrostatically test the valve at a higher pressure for a specified period of time, as required by the design code. Generally, the hydrotest pressure is 150% of the valve rating. This compares with a hydrotest pressure of 125% of MOP for pipelines, as discussed in Chapter 6. 9.4 CODES FOR DESIGN AND CONSTRUCTION The following is a list of applicable standards and codes used in the design and construction of valves and fittings on gas pipelines. ASME B31.8: Gas Transmission and Distribution Piping Systems ASME B16.3: Malleable Iron Threaded Fittings

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 303 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

303

ASME B16.5: Pipe Flanges and Flanged Fittings ASME B16.9: Factory Made Wrought Steel Butt Welding Fittings ASME B16.10: Face to Face and End to End Dimensions of Valves ASME B16.11: Forged Steel Fittings, Socket Welding and Threaded Fittings ASME B16.14: Ferrous Pipe Plugs, Bushing, etc. ASME B16.20: Metallic Gaskets ASME B16.21: Nonmetallic Gaskets ASME B16.25: Butt Welding Ends ASME B16.28: Wrought Steel, Butt Welding, Short Radius Elbows and Returns ASME B16.36: Orifice Flanges ANSI/ASTM A182: Forged or rolled alloy-steel pipe flanges, forged fittings, and valves and parts for high temperature service API 593: Ductile iron plug valves API 594: Wafer type check valves API 595: Cast iron gate valves API 597: Steel venturi gate valves API 599: Steel plug valves API 600: Steel gate valves API 602: Compact cast-steel gate valves API 603: Class 150 corrosion-resistant gate valves API 604: Ductile iron gate valves API 606: Compact carbon-steel gate valves (extended bodies) API 609: Butterfly valves to 150 psig and 150°F API 6D: Pipeline valves MSS DS-13: Corrosion resistant cast flanged valves MSS SP-25: Standard marking system for valves, fittings, and flanges

9.5 GATE VALVE A gate valve is generally used to completely shut off fluid flow or, in the fully open position, provide full flow in a pipeline. Thus, it is used either in the fully closed or fully open position. A gate valve consists of a valve body, seat, and disc; a spindle; a gland; and a wheel for operating the valve. The seat and the gate together perform the function of shutting off the flow of fluid. A typical gate valve is shown in Figure 9.4. Gate valves are generally not suitable for regulating flow or pressure or operating in a partially open condition. For this service, a plug valve or control valve should be used. It must be noted that, due to the type of construction, a gate valve requires many turns of the hand wheel to completely open or close the valve. When fully opened, gate valves offer little resistance to flow, and their equivalent length to diameter ratio (L/D) is approximately 8. The equivalent L/Ds for commonly used valves and fittings are listed in Table 9.2. This ratio represents the resistance of the valve. The gate valves used in the main lines carrying oil or gas must be of full bore or through conduit design to enable smooth passage of scrapers or pigs used for cleaning or monitoring pipelines. Such gate valves are referred to as full bore or through conduit gate valves.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 304 Friday, April 1, 2005 3:12 PM

304

Figure 9.4

GAS PIPELINE HYDRAULICS

Typical gate valve.

Table 9.2 Equivalent Lengths of Valves and Fittings Description

L/D

Gate valve Globe valve Angle valve Ball valve Plug valve straightway Plug valve 3-way thru-flo Plug valve branch flo Swing check valve Lift check valve Standard elbow–90° Standard elbow–45° Standard elbow long radius 90° Standard tee thru-flo Standard tee thru-branch Mitre bends—a = 0 Mitre bends—a = 30 Mitre bends—a = 60 Mitre bends—a = 90

8 340 55 3 18 30 90 50 600 30 16 16 20 60 2 8 25 60

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 305 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

305

Stem Seat ring Spherical plug

Body Figure 9.5

Typical ball valve. (Reproduced from Nayyar, M.L., Piping Handbook, McGrawHill, New York, 2000. With permission.)

9.6 BALL VALVE A ball valve consists of a valve body in which a large sphere with a central hole equal to the inside diameter of the pipe is mounted. As the ball is rotated, in the fully open position the valve provides the through conduit or full bore required for unrestricted flow of the fluid and scrapers or pigs. Compared to a gate valve, a ball valve has very little resistance to flow in the fully open position. When fully open, the L/D ratio for a ball valve is approximately 3.0. The ball valve, like the gate valve, is generally used in the fully open or fully closed position. A typical ball valve is shown in Figure 9.5. Unlike a gate valve, a ball valve requires one-quarter turn of the hand wheel to go from the fully open to the fully closed position. Such quick opening and closing of a ball valve can be important in some installations where isolating pipe sections quickly is needed in the event of emergency. 9.7 PLUG VALVE The plug valve traces its origin to the beginnings of the valve industry. It is a simple device for shutting off or allowing the flow of a fluid in a pipe by a simple quarter turn of the handle. In this sense, it is similar to the ball valve. Plug valves are generally used in screwed piping and in small pipe sizes. Plug valves can be hand wheel operated or operated using a wrench or gearing mechanism. The L/D ratio for this type of valve ranges from 18 to 90, depending upon the design. A typical plug valve is shown in Figure 9.6. 9.8 BUTTERFLY VALVE The butterfly valve was originally used where a tight closure was not absolutely necessary. However, over the years, this valve has been manufactured with fairly tight seals made of rubber or elastomeric materials that provide good shutoff similar

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 306 Friday, April 1, 2005 3:12 PM

306

GAS PIPELINE HYDRAULICS

Plug Body Washer Nut

Figure 9.6

Typical plug valve.

to other types of valves. Butterfly valves are used where space is limited. Unlike gate valves, butterfly valves can be used for throttling or regulating flow, as well as in the full open and fully closed positions. The pressure loss through a butterfly valve is higher in comparison with the gate valve. The L/D ratio for this type of valve is approximately 45. Butterfly valves are used in large and small sizes. They can be hand wheel operated or operated using a wrench or gearing mechanism. A typical butterfly valve is shown in Figure 9.7. 9.9 GLOBE VALVE Globe valves, so called because of their outside shape, are widely used in plant piping. They are suitable for manual and automatic operation. Unlike the gate valve, globe valves can be used for regulating flow or pressures as well as complete

Figure 9.7

Typical butterfly valve.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 307 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

Figure 9.8

307

Typical globe valve.

shutoff of flow. They can also be used as pressure relief valves or as check valves. Compared to a gate valve or ball valve, the globe valve has considerably higher pressure loss in the fully open position. This is due to the fact that the flow of fluid changes direction as it goes through the valve. The L/ D ratio for this type of valve is approximately 340. Globe valves are manufactured in sizes up to NPS 16. They are generally hand wheel operated. A typical globe valve is shown in Figure 9.8.

9.10 CHECK VALVE Check valves are normally in the closed position and are open when the fluid flows through them. They also have the capability of shutting off the flow in the event the pressure downstream exceeds the upstream pressure. In this respect, they are used for flow in one direction only. Thus, they prevent back flow through the valve. Since flow of the fluid through these valves is allowed to be in one direction only, check valves must be installed properly by noting the normal direction of flow. An arrow stamped on the outside of the valve body indicates the direction of flow. Check valves can be classified as swing check valves and lift check valves. The L/D ratios for check valves range from 50 for the swing check valve to as high as 600 for lift check valves. Examples of typical check valves are shown in Figure 9.9.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 308 Friday, April 1, 2005 3:12 PM

308

GAS PIPELINE HYDRAULICS

Flow

Flow

Swing check

Tilting disc check

Lift check Figure 9.9

Typical check valves. (Reproduced from Tullis, J.P., Hydraulics of Pipelines, John Wiley & Sons, New York, 1989. With permission.)

9.11 PRESSURE CONTROL VALVE A pressure control valve is used to automatically control the pressure at a certain point in a pipeline. In this respect it is similar to a pressure regulator discussed next. Whereas the pressure regulator is generally used to maintain a constant downstream pressure, a pressure control valve is used to control the upstream pressure. The upstream and downstream are relative to the location of the valve on the pipeline. Generally, a bypass piping system around the control valve is installed to isolate the control valve in the event of an emergency or for maintenance work on the control valve. This is illustrated in Figure 9.10. Upstream pressure

P1

Pressure drop ∆P

Downstream pressure P2

Flow Figure 9.10

Pressure control valve.

Copyright 2005 by Taylor & Francis Group, LLC

Q

2785_C009.fm Page 309 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

Figure 9.11

309

Pressure regulator. (Reproduced from Katz et al., Handbook of Natural Gas Engineering, McGraw-Hill, New York, 1959. With permission.)

9.12 PRESSURE REGULATOR A pressure regulator is a valve that is similar to a pressure control valve. Its function is to control or regulate the pressure in a certain section of a pipeline system. For example, on a lateral piping that comes off a main pipeline, used for delivering gas to a customer, a lower pressure might be required on the customer side. If the main pipeline pressure at the point of connection to the lateral pipeline is 800 psig, but the customer’s piping is limited to 600 psig, a pressure regulator is used to reduce the pressure by 200 psig, as shown in Figure 9.11. 9.13 PRESSURE RELIEF VALVE The pressure relief valve is used to protect a section of piping by relieving the pipeline pressure when it reaches a certain value. For example, if the MOP of a pipeline system is 1400 psig, a pressure relief valve may be set at 1450 psig. Any upset conditions

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 310 Friday, April 1, 2005 3:12 PM

310

GAS PIPELINE HYDRAULICS

that cause the pipeline pressure to exceed the normal 1400 psig will cause the relief valve to open at the set point of 1450 psig and expel the gas to the atmosphere or to a relief vessel, thereby protecting the pipeline from overpressure and, eventually, rupture. The difference between the relief valve set point (1450 psig) and the pipeline MOP (1400 psig) will depend on the actual application, the valve type, and expected fluctuations in pressure. Generally, the difference will range between 20 and 50 psig. Too close a difference will result in frequent operation of the relief valve, which will be a nuisance and, in many cases, a waste of valuable gas. A large difference between a relief valve set point and the pipeline MOP may render the valve ineffective.

9.14 FLOW MEASUREMENT Gas flow measurement in a pipeline is necessary for properly accounting for the amount of gas transported from one point to another along a gas pipeline. The owner of the gas and the customer who purchases the gas both require that the correct amount of gas be delivered for the agreed-upon price. Even a very small error in flow measurement on large-capacity pipelines can result in huge losses to either the owner or customer of the gas. For example, consider a gas pipeline transporting 300 MMSCFD at a tariff of 50 cents per MCF. An error of 1% in the gas flow measurement can translate to a loss of more than $500,000 per year to either the seller or the buyer. Hence, it is easy to appreciate the importance of good, accurate flow measurement in gas pipelines. Over the years, gas flow measurement technology have improved considerably. Many organizations have jointly developed standards and procedures for measurement of natural gas through orifice meters installed in pipelines. AGA, API, ANSI, and ASME have together endorsed standards for orifice metering of natural gas. The AGA Measurement Committee Report No. 3 is considered to be the leading publication in this regard. This standard is also endorsed by ANSI and API and is referred to as the ANSI/API 2530 standard. We will refer to sections of this standard when discussing orifice meters.

9.15 FLOW METERS Since the orifice meter is the main flow measurement instrument used in the gas industry, we will discuss this first. 9.15.1

Orifice Meter

The orifice meter is a flat steel plate that has a concentric machined hole with a sharp edge and is positioned inside the pipe, as shown in Figure 9.12. As the gas flows through the pipeline and then through the orifice plate, due to the reduction in cross-sectional area as the gas approaches the orifice, the velocity of flow increases and, correspondingly, the pressure drops. After the orifice, the crosssectional area increases again back to the full pipe diameter, which results in expansion of gas and decrease in flow velocity. This process of accelerating flow through

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 311 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

Pressure tap Section 1 D

D/2

311

Pressure tap Section 2 Orifice plate P2

P1

Flow

d

D

Figure 9.12

d

D

Orifice meter.

the orifice and subsequent expansion forms a vena contracta, or a throat, immediately past the orifice, as shown in the figure. Three different types of orifice meters are illustrated in Figure 9.13. The different types of orifice meters shown have different crest shapes, which affect the extent of contraction of the jet of gas as it flows through the orifice. The contraction coefficient Cc is defined in terms of the area of cross section of the vena contracta compared to the cross-sectional area of the orifice, as defined below: Cc =

Ac Ao

(9.1)

where Cc = contraction coefficient, dimensionless Ac = cross-sectional area of the vena contracta, in2 Ao = cross-sectional area of the orifice, in2

Flange P1

Pipe

Orifice P2

V

(a) Sharp-Crested Figure 9.13

Flange P1

Pipe

Orifice P2

Flange P1

Pipe

Orifice P2

V

V

(b) Round-Crested

(c) Nozzle-Crested

Different types of orifice meters. (Reproduced from Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003.)

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 312 Friday, April 1, 2005 3:12 PM

312

GAS PIPELINE HYDRAULICS

The discharge through the orifice meter is represented by the following basic equation: Q = CcCv Ao where Q = Cc = Cv = Ao = A = p1 = p2 = r = z1 = z2 = g =

2 ( p1 − p2 )/ρ + g( z1 − z2 ) 1 − Cc 2 ( Ao / A)2

(9.2)

flow rate, ft3/s contraction coefficient, dimensionless discharge coefficient, dimensionless cross-sectional area of the orifice, in2 cross-sectional area of pipe containing the orifice, in2 upstream pressure, psig downstream pressure, psig density of gas, lb/ft3 upstream elevation, ft downstream elevation, ft acceleration due to gravity

When the elevation difference between the upstream and downstream pressure taps is negligible, the discharge equation for the orifice meter can be simplified to Q = CcCv Ao

2( p1 − p2)/ρ 1 − Cc

2

( ) Ao A

2

(9.3)

where all symbols are as defined earlier. For round-crested and nozzle-crested orifice meters, shown in Figure 9.13, the value of Cc can be taken as 1.0. This indicates an absence of vena contracta for these types of orifices. For the sharp-crested orifice at high Reynolds numbers or for turbulent flow, Cc is calculated from the equation 5

A 2 Cc = 0.595 + 0.29 o A

(9.4)

where all symbols are as defined earlier. There are basically two types of pressure measurements in orifice meters. These are called flange taps and pipe taps. They relate to the locations where the pressure measurements are taken. A flange tap requires that the upstream tap be located at a distance of 1 in. upstream of the nearest plate face and that the downstream tap be located 1 in. downstream of the nearest plate face. Pipe taps are such that the upstream tap be located at a distance of 2.5 times the inside diameter of the pipe, upstream of the nearest plate face and that the downstream tap be located at a distance of 8 times the inside diameter of the pipe, downstream of the nearest plate face. Figure 9.14 illustrates the location of flange taps and pipe taps.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 313 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

313

Flange taps 1 in. (+/– 0.04 in for beta ratio < 0.6) (+/– 0.02 in for beta ratio > 0.6)

Same as downstream

Flow

8D +/– 0.1D

2 1/2D +/– 0.1D

2 1/2 D and 8D Taps (pipe taps) Figure 9.14

Flange taps and pipe taps.

Several terms used in the calculation of the orifice flow must be explained first. The differential pressure for an orifice is the pressure difference between the upstream and downstream taps. The orifice diameter is defined as the arithmetic average of four or more inside diameter measurements evenly spaced. Strict tolerances for the orifice diameters are specified in the AGA3/ANSI 2530 standard. Table 9.3 shows these tolerances taken from the standard. 9.15.1.1

Meter Tube

The meter tube is the piece of pipe in which the orifice plate is installed, along with straightening vanes as needed. A typical meter tube consisting of the orifice plate and straightening vanes is illustrated in Figure 9.15. The dimensions of the meter tube, such as A, B, C, and C′, depend upon the orifice to pipe diameter ratio, also known as beta ratio b, and are specified in AGA Report No. 3. For example, for beta = 0.5, A = 25

A′ = 10

B=4

C=5

C′ = 5.5

Table 9.3 Orifice Plate Diameter Tolerances

Copyright 2005 by Taylor & Francis Group, LLC

Orifice Diameter, in.

Tolerance /, in.

0.250 0.375 0.500 0.625 0.750 0.875 1.000 above 1.000

0.0003 0.0004 0.0005 0.0005 0.0005 0.0005 0.0005 0.0005 (per inch dia)

2785_C009.fm Page 314 Friday, April 1, 2005 3:12 PM

314

GAS PIPELINE HYDRAULICS

Meter tube A′ B C

C′

Straightening vanes Figure 9.15

Orifice

Meter tube installation.

These numbers are actually multiples of the pipe or meter tube diameter. The requirements of straightening vanes before the orifice plate depend on the specific installation. The main reason for straightening vanes is to reduce flow disturbance at the orifice plate from upstream fittings. Refer to AGA Report No. 3 for various meter tube configurations. The orifice flow rate is the mass flow rate or volume flow rate of gas per unit of time. The density is the mass per unit volume of gas at a specific temperature and pressure. 9.15.1.2

Expansion Factor

The expansion factor is a dimensionless factor used to correct the calculated flow rate to take into account the reduction in gas density as it flows through an orifice, which is caused by the increased velocity and corresponding reduced static pressure. Methods of calculating the expansion factor Y will be discussed in subsequent sections. The beta ratio is defined as the ratio of the orifice diameter to the meter tube diameter, as follows:

β=

d D

(9.5)

For orifice meters with flange taps, the beta ratio ranges between 0.15 and 0.70. For orifice meters with pipe taps, the beta ratio ranges between 0.20 and 0.67, where b = beta ratio, dimensionless d = orifice diameter, in. D = meter tube diameter, in. The fundamental orifice meter flow equation described in ANSI 2530/AGA Report No. 3 is as follows: qm =

Copyright 2005 by Taylor & Francis Group, LLC

π C Y d 2 (2 gρ f ∆P)0.5 (1 − β 4 )0.5 4

(9.6)

2785_C009.fm Page 315 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

315

or qm = KY

where qm rf C b d D Y g ∆P K

= = = = = = = = = =

π 2 d (2 gρ f ∆P)0.5 4

(9.7)

β=

d D

(9.8)

K=

C CD 2 = (1 − β 4 )0.5 ( D 4 − d 4 )0.5

(9.9)

mass flow rate of gas, lb/s density of gas, lb/ ft3 discharge coefficient beta ratio, dimensionless orifice diameter, in. meter tube diameter, in. expansion factor, dimensionless acceleration due to gravity, ft/s2 pressure drop across the orifice, psi flow coefficient, dimensionless

These equations were arrived at using the conservation of energy and mass equations with thermodynamics and the equation of state for the gas in question. It can be seen that, essentially, these formulas give the mass flow rate of gas. We need to convert these to the volume flow rate using the density. The coefficient of discharge C in the preceding equation is approximately 0.6, and the flow coefficient K is a value that is between 0.6 and 0.7. Both the flow coefficient K and the expansion factor Y are determined using test data. The volume flow rate at standard (base) conditions is calculated from the mass flow rate as follows: qv =

qm ρb

(9.10)

where q v = volume flow rate, ft3/s qm = mass flow rate, lb/s r b = gas density at base temperature, lb/ft3 The expansion factor Y for low-compressibility fluids, such as water at 60°F and 1 atmosphere pressure, is taken as 1.0. For gases, Y can be calculated as explained in the next section. The flow coefficient K is found to vary with the diameter of the meter tube D, orifice diameter d, mass flow rate qm, and fluid density and viscosity at the flowing temperature. For gases, K also varies with the ratio of

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 316 Friday, April 1, 2005 3:12 PM

316

GAS PIPELINE HYDRAULICS

the differential pressure to the static pressure and k, the ratio of specific heat of the gas. In many cases, the flow coefficient K is considered to be a function of the Reynolds number, acoustic ratio, meter tube diameter, and beta ratio. Rearranging Equation 9.7, we get KY =

4 qm

(9.11)

π d [2 gc ρ f ∆P]0.5 2

Several empirical equations are available to calculate the flow coefficient K. The following equation by Buckingham and Bean is endorsed by the National Bureau of Standards (NBS) and is listed in AGA Report No.3. For flange taps, 1 0.5 0.007 0.076 Ke = 0.5993 + + 0.364 + 0.5 β 4 + 0.4 1.6 − 0.07 + − β D D D D 5

65 0.034 − 0.009 + (0.5 − β )1.5 + 2 + 3 (β − 0.7)2.5 D D where Ke = D = d = b =

2.5

(9.12)

flow coefficient for Reynolds number Rd = d(106/15), dimensionless meter tube diameter, in. orifice diameter, in. beta ratio, dimensionless

For pipe taps, Ke = 0.5925 +

0.0182 0.06 2 0.225 5 + 0.440 − β + 0.935 + β D D D

+ 1.35β 14 +

1.43 (0.25 − β )2.5 D 0.5

(9.13)

where all symbols are as defined before. For flange taps and pipe taps, the value of Ko is calculated from Ko =

Ke

1+

15×10−6 E d

(9.14)

where the parameter E in Equation 9.14 is found from E = d (830 − 5000 β + 9000 β 2 − 4200 β 3 + B)

Copyright 2005 by Taylor & Francis Group, LLC

(9.15)

2785_C009.fm Page 317 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

317

The value of parameter B in Equation 9.15 is defined as follows: For flange taps, B = For pipe taps, B =

530 D 0.5

(9.16)

875 + 75 D

(9.17)

Finally, the flow coefficient K is calculated from E K = Ko 1 + Rd

(9.18)

where Ko = flow coefficient for infinitely large orifice Reynolds numbers, dimensionless Rd = Reynolds number at the inlet of orifice, dimensionless The Reynolds number used in the preceding equations is calculated from Rd = where Rd = Vf = d = rf = m =

Vf d ρ f µ

(9.19)

Reynolds number at the inlet of orifice, dimensionless velocity of fluid at inlet of orifice, ft/s orifice diameter, ft fluid density at flowing conditions, lb/ft3 dynamic viscosity of fluid, lb/ft.s

The values of flow coefficient K calculated using the preceding equations apply to orifice meters manufactured and installed in accordance with AGA Report No. 3, as long as the meter tube is greater than 1.6 in. inside diameter and the beta ratio is between 0.10 and 0.75. The uncertainties in flow coefficient K, in accordance with AGA Report No. 3, follow: For For For For

flange taps, the uncertainty is +/−0.5% for 0.15 < b < 0.70 flange taps, the uncertainty is greater than +/−1.0% for 0.15 > b > 0.70 pipe taps, the uncertainty is +/−0.75% for 0.20 < b < 0.67 pipe taps, the uncertainty is greater than +/−1.5% for 0.20 > b > 0.67

The expansion factor Y is calculated in two ways. In the first method, it is calculated using the upstream pressure, and in the second method, it is calculated using the downstream pressure. The following equation is used for the expansion factor Y1 with reference to upstream pressure.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 318 Friday, April 1, 2005 3:12 PM

318

GAS PIPELINE HYDRAULICS

For flange taps, Y1 = 1 − (0.41 + 0.35β 4 )

x1 k

(9.20)

For pipe taps, Y1 = 1 − [0.333 + 1.145(β 2 + 0.7β 5 + 12β 13 )]

x1 k

(9.21)

and the pressure ratio x1 is x1 =

Pf 1 − Pf 2 hw = 27.707 Pf 1 Pf 1

(9.22)

where Y1 = expansion factor based on upstream pressure x1 = ratio of differential pressure to absolute upstream static pressure hw = differential pressure between upstream and downstream taps in in. of water at 60°F Pf1 = static pressure at upstream tap, psia Pf2 = static pressure at downstream tap, psia x1/k = acoustic ratio, dimensionless k = ratio of specific heats of gas, dimensionless The value of Y 1 calculated using these equations is subject to a tolerance from 0 to +/− 0.5% for the range of x = 0 to 0.20. For larger values of x, the uncertainty is larger. For flange taps, the values of Y1 are valid for a beta ratio range of 0.10 to 0.80. For pipe taps, the beta ratio range is 0.10 to 0.70. With reference to the downstream pressure, the expansion factor Y2 is calculated using the following equations. For flange taps, 1 Y2 = Y1 1 − x1

0.5

Y2 = (1 + x 2 )0.5 − (0.41 + 0.35β 4 )

(9.23) x2 k (1 + x 2 )0.5

(9.24)

For pipe taps, Y2 = (1 + x 2 )0.5 − [0.333 + 1.145(β 2 + 0.7β 5 + 12β 13 )]

Copyright 2005 by Taylor & Francis Group, LLC

x2 k (1 + x 2 )0.5

(9.25)

2785_C009.fm Page 319 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

319

and the pressure ratio x2 is x2 =

Pf 1 − Pf 2 hw = Pf 2 27.707 Pf 2

(9.26)

where all symbols are as defined before. The density of the flowing gas used in Equation 9.6 must be obtained from the equation of state or from tables. It is important to use the correct density in the flow equations. Otherwise, the uncertainty in flow measurement could be as great as 10%. Generally, the density of the gas can be calculated from the perfect gas law discussed in Chapter 1, with the modification using the compressibility factor. The following equation is obtained by rearranging the real gas equation and using the gravity of gas (see Chapter 1 for details):

ρf =

m Gi MPf = V Z f RT f

(9.27)

ρf1 =

Gi MPf 1 Z f 1RT f

(9.28)

ρb =

MGi Pb RZ bTb

(9.29)

where m = mass of gas V = volume of gas Gi = gravity of gas (air = 1.00) M = molecular weight of gas Pf = absolute gas pressure Zf = compressibility factor at flowing temperature R = gas constant Tf = absolute flowing temperature subscript f 1 refers to upstream tap flowing conditions subscript f 2 refers to downstream tap flowing conditions subscript b refers to base conditions Two other equations, based on real gas specific gravity and taking the base conditions of 14.73 psia and 60°F, result in the gas densities at the upstream tap and at the base conditions as follows:

ρf1 =

MZ bGPf 1 0.99949 RZ f 1T f

(9.30)

MGPb 0.99949 RTb

(9.31)

ρb =

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 320 Friday, April 1, 2005 3:12 PM

320

GAS PIPELINE HYDRAULICS

Knowing the densities at the upstream tap and at the base condition, the following equation is used for the volume flow rate. This equation is derived from the equations listed in the preceding sections. qv =

π 4

(

2 g KY1d 2

)

( ρ f 1∆P)

ρb

(9.32)

Combining all equations we have reviewed so far, AGA Report No. 3 shows a compact equation for the flow of gas through an orifice meter as follows: Qv = C hw Pf

(9.33)

where Qv = gas flow rate at base conditions, ft3/h hw = differential pressure between upstream and downstream taps in in. of water at 60°F Pf = absolute static pressure, psia C = orifice flow constant For Pf , subscript 1 is used for upstream and subscript 2 for downstream pressure. The orifice flow constant C consists of the product of several factors that depend on the Reynolds number, expansion factor, base pressure, base temperature, flowing temperature, gas gravity, and supercompressibility factor of gas. It is given by the following equation: C = Fb Fr Fpb Ftb Ftf Fgr FpvY

(9.34)

where the dimensionless factors are Fb = basic orifice factor Fr = Reynolds number factor Fpb = pressure base factor Ftb = temperature base factor Ftf = flowing temperature factor Fgr = gas relative density factor Fpv = supercompressibility factor Y = expansion factor These values of the factors that constitute the orifice flow constant C are defined in AGA Report No. 3 and are listed in Appendix B of that publication. However, each of these factors can be calculated as follows. The basic orifice factor is Fb = 338.178d 2 K o where Ko is calculated using Equation 9.14.

Copyright 2005 by Taylor & Francis Group, LLC

(9.35)

2785_C009.fm Page 321 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

321

The Reynolds number factor is Fr = 1 +

E Rd

(9.36)

K = K o Fr

(9.37)

The pressure base factor is 14.73 Pb

(9.38)

Tb 519.67

(9.39)

Fpb = The temperature base factor is Ftb = The flowing temperature factor is

519.67 Ftf = Tf

0.5

(9.40)

The gas relative density factor is 1 Fgr = Gr

0.5

(9.41)

where all symbols in the preceding equations are as defined before.

9.16 VENTURI METER The venturi meter, shown in Figure 9.16, is based upon Bernoulli’s equation. It consists of a smooth gradual contraction from the main pipe size to a reduced section known as the throat, finally expanding back gradually to the original pipe diameter. This type of venturi meter is called the Herschel type. The angle of contraction from the main pipe to the throat section is in the range of 21° +/− 2°. The gradual expansion from the throat to the main pipe section is in the range of 5 to 15°. This design causes the least energy loss such that the discharge coefficient can be assumed at 1.0. Venturi meters range in size from 4.0 in. to 48.0 in. The beta ratio, equal to d/D, generally ranges between 0.30 and 0.75. The gas pressure in the main pipe section is represented by P1 and that at the throat is represented by P2. As gas flows through a venturi meter, it increases in flow velocity in the narrow throat section. Correspondingly, the pressure reduces in the throat section according to Bernoulli’s equation. After gas leaves the throat section,

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 322 Friday, April 1, 2005 3:12 PM

322

GAS PIPELINE HYDRAULICS

Pipe section 1 Throat section 2 P1

Pipe section 3

P2

D

Flow

d

D

h Manometer

Figure 9.16

Venturi meter.

it reduces in flow velocity due to the increase in pipe cross-sectional area, and it reaches the original flow velocity. The flow velocity in the main pipe section before the throat is calculated from the known pressures P1 and P2:

V1 =

2 g ( P1− P2 ) + ( Z1 − Z 2 ) − hL ρ

( ) −1 2

A1 A2

(9.42)

Neglecting the elevation difference Z1 – Z2 and the friction loss hL, this equation reduces to the following:

V1 = C

where V1 = r = A1 = A2 = C =

2 g ( P1− P2 ) ρ

( ) −1 A1 A2

2

(9.43)

velocity of gas in the main pipe section before the throat the average gas density cross-sectional area of the pipe cross-sectional area of the throat discharge coefficient, dimensionless

The volume flow rate is then calculated by multiplying the velocity by the crosssectional area, resulting in the following equation:

Q = CA1

Copyright 2005 by Taylor & Francis Group, LLC

2 g ( P1− P2 ) ρ

( ) −1 A1 A2

2

(9.44)

2785_C009.fm Page 323 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

323

Using the beta ratio, we simplify the above equation as follows:

Q = CA1

2 g ( P1− P2 ) ρ

(9.45)

( ) −1 1 β

4

The discharge coefficient C is a number less than 1.0, and it depends on the Reynolds number in the main pipe section. For a Reynolds number greater than 2 × 105, the value of C remains constant at 0.984. In smaller pipe sizes, such as 2 to 10 in., venturi meters are machined and, therefore, have a better surface finish than the larger rough cast meters. Smaller venturi meters have a C value of 0.995 for Reynolds numbers larger than 2 × 105. 9.17 FLOW NOZZLE The flow nozzle shown in Figure 9.17 is another device for measuring flow rate. It consists of a main pipe section, followed by a gradual reduction in cross-section area and a short cylindrical section, ending in a gradual expansion to the original pipe size. The discharge coefficient C for a flow nozzle is approximately 0.99 for Reynolds numbers greater than 106. At lower Reynolds numbers, due to greater energy loss subsequent to the nozzle throat, C values are lower. The discharge coefficient C depends on the beta ratio and Reynolds number. It is calculated using the following equation: C = 0.9975 − 6.53

Pressure tap Section 1 D

D/2

P1

P2

To manometer Flow nozzle.

Copyright 2005 by Taylor & Francis Group, LLC

(9.46)

Pressure tap Section 2

d Flow

D

Figure 9.17

β R

2785_C009.fm Page 324 Friday, April 1, 2005 3:12 PM

324

GAS PIPELINE HYDRAULICS

where b = d/D R = Reynolds number based on the pipe diameter D Example 1 An orifice meter with 4 in. diameter is installed in a pipe with an inside diameter of 12.09 in. The differential pressure is measured at 30 in. of water, and the static pressure upstream is 600 psig. The gas gravity = 0.6 and the gas flowing temperature = 70°F. The base temperature and the base pressure are 60°F and 14.7 psia, respectively. Assuming flange taps, calculate the flow rate in standard ft3/h. The barometric pressure is 14.5 psia. Solution The basic orifice factor F b is calculated from the AGA 3 appendix as follows:

Fb = 3258.5 (hP)0.5 = [30 × (600 + 14.5)]0.5 = 135.78 Fr = 1 +

0.0207 = 1.0002 135.78

Fpb =

14.73 = 1.002 14.7

Ftb =

60 + 460 = 1.006 519.67

519.67 Ftf = 70 + 460 1 Fgr = 0.6

0.5

= 0.9902

0.5

= 1.291

Fpv = 1.0463 30 h = = 0.0488 P 614.5

β=

4 = 0.3309 12.09

Y = 0.9995 C = 3258.5 × 1.0002 × 1.002 × 1.006 × 0.9902 × 1.291 × 1.00463 × 0.9995 = 4391.96 Using Equation 9.33, the flow rate is Qv = 4391.96 × 135.78 = 596, 340 ft 3 /h

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 325 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

325

9.18 SUMMARY In this chapter we covered the topics of valves and flow measurement as they relate to gas pipeline transportation. The various types of valves used and their functions were reviewed. The importance of flow measurement in natural gas pipeline transaction was explained. The predominantly used measuring device known as an orifice meter was discussed in detail. The calculation methodology based on AGA Report No. 3 was reviewed. The venturi meter and the flow nozzle were also discussed.

PROBLEMS 1. An orifice meter 2 in. in diameter is installed in a pipe with an inside diameter of 12.09 in. The differential pressure is measured at 20 in. of water, and the static pressure upstream is 500 psig. The gas gravity = 0.65 and gas flowing temperature = 75°F. The base temperature and the base pressure are 60°F and 14.7 psia, respectively. Assuming pipe taps, calculate the flow rate in standard ft3/h. The barometric pressure is 14.6 psia. 2. An orifice meter has a bore size of 1 in. diameter and is installed in a pipe with an inside diameter of 6.125 in. The differential pressure is measured at 10 in. of water, and the static pressure upstream is 300 psig. The gas gravity = 0.6 and gas flowing temperature = 70°F. The base temperature and the base pressure are 60°F and 14.7 psia, respectively. Assuming flange taps, calculate the flow rate in standard ft3/h. The barometric pressure is 14.6 psia.

REFERENCES 1. Miller, R.W., Flow Measurement Engineering Handbook, McGraw-Hill, New York, 1983. 2. Upp, E.L., Fluid Flow Measurement, Gulf Publishing Company, Houston, TX, 1993. 3. Flow of Fluids through Valves, Fittings and Pipe, Crane Company, New York, 1976. 4. Cheremisinoff, N., Applied Fluid Flow Measurement, Marcel Dekker, Inc., New York, 1979. 5. Nayyar, M.L., Piping Handbook, 7th ed., McGraw-Hill, New York, 2000. 6. Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003. 7. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 8. Engineering Data Book, 10th ed., Gas Processors Suppliers Association, Tulsa, OK, 1994. 9. Steady Flow in Gas Pipelines, Contract Report No. 10, July 1965, Pipeline Research Council International, Houston, TX, 1965. 10. Katz, D.L. et al., Handbook of Natural Gas Engineering, McGraw-Hill, New York, 1959. 11. Orifice Metering of Natural Gas, AGA Report No. 3, ANSI/API 2530, American Gas Association, Arlington, VA, June 1987. 12. Mendel, O., Practical Piping Handbook, PennWell Books, Tulsa, OK, 1981.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 326 Friday, April 1, 2005 3:12 PM

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 327 Friday, April 1, 2005 3:32 PM

CHAPTER

10

Pipeline Economics In the previous chapters we explored different scenarios of pipe sizes and pressures to transport natural gas through pipelines from one location to another. Various pressure drop formulas, compression requirements, and HP required were calculated without delving too much into costs of facilities. In this chapter the economic aspects of pipelines will be reviewed. The economic pipe size required for a particular throughput will be arrived at considering the various costs that make up a pipeline system. The initial capital cost of pipeline and ancillary facilities will be discussed, along with the annual operating and maintenance costs. Since pipelines are generally designed to transport gas belonging to one company by another company, a methodology for determining transportation cost or tariff will be analyzed. A pipeline can be constructed to transport natural gas for the owner of the pipeline, to sell gas to another company, or to transport some other company’s gas. These three scenarios represent three major uses of pipeline transportation of natural gas. The economics involved in the selection of pipe diameter, compressor station, and related facilities will vary slightly for each scenario. As an owner company transporting its own gas, minimal facilities will probably be built. However, Department of Transportation (DOT) codes and other regulatory requirements will still have to be met to ensure a safe pipeline operation that will not endanger humans or the environment. In the second scenario, in which a company builds a pipeline to transport its gas and sells the gas at the end of the pipeline to a customer, minimal facilities will be constructed without too much regulatory control. In the third scenario, a pipeline company constructs and operates a pipeline for the purpose of transporting gas belonging to other companies. This will be under the jurisdiction of the Federal Energy Regulatory Commission (FERC) or a state agency such as the Public Utilities Commission (PUC) in California or the Texas Railroad Commission in Texas. An interstate pipeline in which the pipeline crosses one or more state boundaries will be regulated by the FERC. A pipeline that is intrastate, such as wholly within California, will be subject to PUC rules and not FERC. Such regulatory requirements impose strict guidelines on the type and number of facilities and costs that may be passed on to the customer requesting gas transportation. These regulatory requirements will dictate that excessive capital facilities not be built and the amortized cost passed on 327

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 328 Friday, April 1, 2005 3:32 PM

328

GAS PIPELINE HYDRAULICS

to the customers. Whereas a private pipeline company transporting its own gas may build in extra compressor units as spares to ensure uninterrupted operation in the event of equipment failure, FERC-regulated pipelines may not be able to do so. Thus, pipeline economics will differ slightly from case to case. In this chapter we will not discuss other modes of transportation of gas, such as truck transport of pressurized gas containers. The general economic principles discussed here are applicable to private unregulated pipelines as well as FERC-regulated pipelines used for interstate transportation of natural gas. 10.1 COMPONENTS OF COST In a gas pipeline system the major components that contribute to the initial capital cost are the pipeline, compressor stations, mainline valve stations and metering facilities, telecommunications, and supervisory control and data acquisition (SCADA). Other costs include environmental and permitting costs, right of way (ROW), acquisition cost, engineering and construction management, legal and regulatory costs, contingency, and allowance for funds used during construction (AFUDC). The recurring annual costs will include operating and maintenance (O&M) costs, fuel, energy and utility costs, rental, permitting, and annual right of way costs. The O&M costs will include payroll and general and administrative (G&A) costs. In any pipeline system constructed to provide transportation of gas, there will be capital costs and annual operating costs. If we decide on a useful life of the pipeline (say, 30 or 40 years) we can annualize all costs and also determine the revenue stream necessary to amortize the total investment in the pipeline project. The revenue earned after expenses and taxes plus a percentage for proﬁt divided by the volume transported will give the transportation tariff necessary. The calculation of capital cost, operating cost, and transportation tariff will be illustrated using an example. Throughout this chapter we will need to convert annual cash ﬂows or expenses into present value and vice versa. A useful equation relating the present value of a series of annual payments over a number of years at a speciﬁed interest rate is as follows: PV = where PV R i n

= = = =

R 1 1− i (1 + i)n

(10.1)

present value, $ series of cash ﬂows, $ interest rate, decimal value number of periods, years

For example, $10,000 in annual payments for 20 years at an annual interest rate of 10% results in a present value of PV =

Copyright 2005 by Taylor & Francis Group, LLC

10, 000 1 = $85,136 1− 20 0.10 (1 + 0.10)

2785_C010.fm Page 329 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

329

Similarly, we can convert a present value of $10 million into an annualized cost based on 8% interest for 30 years as follows. From Equation 10.1, 10, 000, 000 =

R 1 1− 0.08 (1.08)30

Solving for the annual cost R, we get R = $888,274 Next, we will calculate the cost of service and transportation tariff using a simple example. Example 1 A natural gas pipeline transports 100 MMSCFD at a load factor of 95%. The capital cost is estimated at $60 million and the annual operating cost is $5 million. Amortizing the capital at 10% for a project life of 25 years, calculate the cost of service and transportation tariff for this pipeline. Solution All costs will be converted to annualized values for a 25-year project life and 10% interest rate. This will be the cost of service on an annual basis. When this cost is divided by the annual pipeline throughput, we obtain the transportation tariff. The capital cost of $60 million is ﬁrst converted to annual cash ﬂow at a 10% interest rate for a period of 25 years. Using Equation 10.1, Annualized capital cost =

60 × 0.10 = $6.61 million 1 − 1 25 (1.10)

This assumes zero salvage value at the end of the 25-year useful life of the pipeline. Therefore, for a project life of 25 years and a discount rate of 10%, the capital cost of $60 million is equivalent to annual cost of $6.61 million. Adding the annual operating cost of $5 million, the total annual cost is $6.61 + $5 = $11.61 million per year. This annual cost is deﬁned as the cost of service incurred each year. Actually, to be accurate, we should take into account several other factors such as the tax rate, depreciation of assets, and proﬁt margin to arrive at a true cost of service. The transportation tariff is deﬁned as the cost of service divided by the annual volume transported. At a 95% load factor and ﬂow rate of 100 MMSCFD, the transportation tariff is Tariff =

Copyright 2005 by Taylor & Francis Group, LLC

$11.61 × 10 6 × 10 3 = $0.3348 per MCF 100 × 10 6 × 365 × 0.95

2785_C010.fm Page 330 Friday, April 1, 2005 3:32 PM

330

GAS PIPELINE HYDRAULICS

In other words, for this pipeline, every MCF of gas transported requires a payment of approximately 33.5 cents to the pipeline owner that provides the transportation. This is a very rough and simplistic calculation of an example of tariff. In reality, we must take into account many other factors to arrive at an accurate cost of service. For example, the annual operating cost will vary from year to year over the life of the pipeline, due to inﬂation and other reasons. Taxes, depreciation of assets, and salvage value at the end of the life of the pipeline must also be considered. Nevertheless, the preceding analysis gives a quick overview of the approach used to calculate a rough value of the transportation cost.

10.2 CAPITAL COSTS The capital cost of a pipeline project consists of the following major components: • • • • • • • • •

Pipeline Compressor stations Mainline valve stations Meter stations Pressure regulator stations SCADA and telecommunication Environmental and permitting Right of way acquisitions Engineering and construction management

In addition, there are other costs such as allowance for funds used during construction (AFUDC) and contingency. Each of the preceding major categories of capital cost will be discussed next. 10.2.1

Pipeline

The pipeline cost consists of those costs associated with the pipe material, coating, pipe ﬁttings, and the actual installation or labor cost. In Chapter 6, we introduced a simple formula to calculate the weight of pipe per unit length. From this and the pipe length, the total tonnage of pipe can be calculated. Given the cost per ton of pipe material, the total pipe material cost can be calculated. Knowing the construction cost per unit length of pipe, we can also calculate the labor cost for installing the pipeline. The sum of these two costs is the pipeline capital cost. Using Equation 6.11 for pipe weight, the cost of pipe required for a given pipeline length is found from PMC =

10.68( D − T )TLC × 5280 2000

where PMC = pipe material cost, $ L = length of pipe, mi D = pipe outside diameter, in.

Copyright 2005 by Taylor & Francis Group, LLC

(10.2)

2785_C010.fm Page 331 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

331

= pipe wall thickness, in. = pipe material cost, $/ton

T C

In SI units, PMC = 0.0246( D − T )TLC where PMC L D T C

= = = = =

(10.3)

pipe material cost, $ length of pipe, km pipe outside diameter, mm pipe wall thickness, mm pipe material cost, $/metric ton

Generally, pipe will be supplied externally coated and wrapped. Therefore, we must add this cost or a percentage to the bare pipe cost to account for the extra cost and the delivery cost to the construction site. In the absence of actual cost, we may increase the bare pipe cost by a small percentage, such as 5%. For example, using Equation 10.2 for a 100 mi pipeline, NPS 20 with 0.500 in. wall thickness, the total pipe cost, based on $800 per ton, is PMC =

10.68(20 − 0.5)0.5 × 100 × 800 × 5280 = $21.99 million 2000

If the pipe is externally coated and wrapped and delivered to the ﬁeld at an extra cost of $5 per ft, this cost can be added to the bare pipe cost as follows: Pipe coating and wrapping cost = $5 × 5280 × 100 = $2.64 million Therefore, the total pipe cost becomes $21.99 + $2.64 = $24.63 million The labor cost to install the pipeline can be represented in dollars per unit length of pipe. For example, the labor cost might be $60 per ft or $316,800 per mi of pipe for a particular size pipe in a certain construction environment. This number will depend on whether the pipeline is installed in open country, ﬁelds, or city streets. Such numbers are generally obtained from contractors who will take into consideration the difﬁculty of trenching, installing pipe, and backﬁlling in the area of construction. For estimation purposes, there is a wealth of historical data available for construction cost for various pipe sizes. Sometimes the pipe installation cost is expressed in terms of dollars per in. diameter per mi of pipe. For example, an NPS 16 pipe might have an installation cost of $15,000 per in.-diameter-mile. Thus, if 20 mi of NPS 16 pipe are to be installed, we estimate the labor cost as follows: Pipe installation cost = $15,000 × 16 × 20 = $4.8 million

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 332 Friday, April 1, 2005 3:32 PM

332

GAS PIPELINE HYDRAULICS

Table 10.1

Typical Pipeline Installation Costs

Pipe Diameter, in.

Average Cost, $/in.-dia/mi

8 10 12 16 20 24 30 36

18,000 20,000 22,000 14,900 20,100 33,950 34,600 40,750

If we convert this cost on a unit length basis, we get Pipe installation cost =

4.8 × 10 6 = $45.45 per ft 20 × 5280

Table 10.1 shows typical installation costs for pipelines. These numbers must be veriﬁed by discussions with construction contractors who are familiar with the construction location. Several other construction costs must be added to the installation costs for straight pipe. These expeditures include costs for road, highway, and railroad crossings and stream and river crossings. These costs can be provided as lump sum numbers, which can be added to the pipeline installation costs to come up with a total pipeline construction cost. For example, a pipeline might include two road and highway crossings that total $300,000 in addition to a couple of river crossings costing $1 million. Compared to the installation cost of a long-distance pipeline, the road and river crossings total might be a small percentage. 10.2.2

Compressor Stations

In order to provide transportation of gas through a pipeline, we have to install one or more compressor stations to provide the necessary gas pressure. Once we decide on the details of the compressor station equipment and piping, a detailed bill of materials can be developed from the engineering drawings. Based upon quotations from equipment vendors, a detailed cost estimate of the compressor stations can be developed. In the absence of vendor data and in situations where a rough order of magnitude costs for compressor stations is desired, we can use an all-inclusive price of dollars per installed HP. For example, using an installed cost of $2000 per HP, for a 5000 HP compressor station, the capital cost will be estimated as follows: Compressor station cost = 2000 × 5000 = $10 million In the above calculations, the all-inclusive number of $2000 per installed HP is expected to include material and equipment cost and the labor cost for installing the compressor equipment, piping, valves, instrumentation, and controls within the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 333 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

333

compressor stations. Generally, the $/HP number decreases as the size of the compressor HP increases. Thus, a 5000 HP compressor station might be estimated on the basis of $2000 per HP, whereas a 20,000 HP compressor station will be estimated at an installed cost of $1500 per HP. These numbers are mentioned for illustration purposes only. Actual $/HP values must be obtained from historical pipeline cost data and in consultation with compressor station construction contractors and compressor station equipment vendors. Generally, the pipeline and compressor station costs constitute the bulk of the total pipeline project cost. 10.2.3

Mainline Valve Stations

Mainline block valves are installed to isolate sections of a pipeline for safety reasons and maintenance and repair. In the event of a pipeline rupture, the damaged pipeline section can be isolated by closing off the mainline valves on either side of the rupture location. For mainline valve stations installed at speciﬁed intervals along the pipeline, the cost of facilities can be speciﬁed as a lump sum ﬁgure that includes the mainline valve and operator, blowdown valves and piping, and other pipe and ﬁttings that constitute the entire block valve installation. Generally, a lump sum ﬁgure can be obtained for a typical mainline block valve installation from a construction contractor. For example, an NPS 16 mainline valve installation might be estimated at $100,000 per site. In a 100 mi, NPS 16 pipeline, DOT code requirements might dictate that a mainline valve be installed every 20 mi. Therefore, in this case there would be six mainline valves for a 100 mi pipeline. At $100,000 per site, the total installed cost of all mainline valve stations will be $600,000. This will be added to the capital cost of the pipeline facilities. 10.2.4

Meter Stations and Regulators

Meter stations are installed for measuring the gas ﬂow rate through the pipeline. These meter stations will consist of meters, valves, ﬁttings, instrumentation, and controls. Meter stations can also be estimated as a ﬁxed price, including material and labor for a particular site. For example, a 10 in. meter station might cost $300,000 lump sum. If there are four such meter stations on a 100 mi gas pipeline, the total meter station cost will be $1.2 million. The meter station costs, like the mainline valve station costs, will be added to the pipeline cost. Pressure regulating stations are installed at some locations on a gas pipeline to reduce the pressure for delivery to a customer or to protect a section of a pipeline with a lower MOP. Such pressure regulating stations can also be estimated as a lump sum per site and added to the capital cost of the pipeline. 10.2.5

SCADA and Telecommunication System

Typically on a gas pipeline, the pressures, ﬂow rates, and temperatures are monitored along the pipeline by means of electronic signals sent from remote terminal units (RTU) on various valves and meters to a central control center via telephone lines or

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 334 Friday, April 1, 2005 3:32 PM

334

GAS PIPELINE HYDRAULICS

microwave or satellite communication systems. The term supervisory control and data acquisition (SCADA) is used to refer to these facilities. SCADA is used to remotely monitor, operate, and control a gas pipeline system from a central control center. In addition to monitoring valve status, ﬂows, temperatures, and pressures along a pipeline, SCADA also monitors the compressor stations. In many cases, starting and stopping of compressor units are performed remotely using SCADA. The cost of SCADA facilities range from $2 million to $5 million or more, depending on the pipeline length, number of compressor stations, and the number of mainline valves and meter stations. Sometimes this category is estimated as a percentage of the total project cost, such as 2 to 5%. 10.2.6

Environmental and Permitting

The environmental and permitting costs are those costs that are associated with the modiﬁcations to pipeline, compressor stations, and valve and meter stations to ensure that these facilities do not pollute the atmosphere, streams, and rivers or damage ecosystems including the ﬂora and fauna, ﬁsh and game, and endangered species. Many sensitive areas, such as Native American religious and burial sites, must be considered and allowances must be made for mitigation of habitat in certain areas. Permitting costs can include those costs associated with changes needed to compression equipment, pipeline alignment such that toxic emissions from pipeline facilities do not endanger the environment, humans, and plant and animal life. In many cases, these costs include acquisition of land to compensate for the areas that were disturbed due to pipeline construction. Such lands acquired will be allocated for public use, such as parks and wildlife preserves. Permitting costs will also include an environmental study, the preparation of an environmental impact report, and permits for road crossings, railroad crossings, and stream and river crossings. These environmental and permitting costs on a gas pipeline project may range between 10 and 15% of total project costs. 10.2.7

Right of Way Acquisitions

The right of way (ROW) for a pipeline is acquired from private parties and state and local government and federal agencies for a fee. This fee might be a lump sum payment at the time of acquisition with additional annual fees to be paid for a certain duration. For example, the right of way can be acquired from private farms, cooperatives, the Bureau of Land Management (BLM), and railroads. The initial cost for acquiring the ROW will be included in the capital cost of the pipeline. The annual rent or lease payment for land will be considered an expense. The latter will be included in the annual costs, such as operating costs. As an example, the ROW acquisition costs for a gas pipeline might be $30 million. This cost would be added to the total capital costs of the gas pipeline. Also, there might be annual ROW lease payments of $300,000 a year, which would be added to other annual costs such as operating and maintenance costs and administrative costs. For most gas pipelines, the initial ROW costs will be in the range of 6 to 10% of the total project costs.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 335 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

10.2.8

335

Engineering and Construction Management

Engineering costs are those costs that pertain to the design and preparation of drawings for the pipeline, compressor stations, and other facilities. This will include both preliminary and detailed engineering design costs, including development of speciﬁcations, manuals, purchase documents, equipment inspection, and other costs associated with materials and equipment acquisition for the project. The construction management costs include ﬁeld personnel cost, rental facilities, ofﬁce equipment, transportation, and other costs associated with overseeing and managing the construction effort for the pipeline and facilities. On a typical pipeline project, engineering and construction management costs range from 15 to 20% of the total pipeline project cost. 10.2.9

Other Project Costs

In addition to the major cost categories discussed in the preceding sections, there are other costs that should be included in the total pipeline project cost. These include legal and regulatory costs necessary for ﬁling an application with the FERC and state agencies that have jurisdiction over interstate and intrastate transportation of natural gas, as well as a contingency costs intended to cover categories not considered or not envisioned when the project was conceptualized. As the project is engineered, new issues and problems might surface that require additional funds. These are generally included in the category of contingency cost. The ﬁnal category of cost, referred to as allowance for funds used during construction (AFUDC), is intended to cover the cost associated with ﬁnancing the project during various stages of construction. Contingency and AFUDC costs can range between 15 and 20% of the total project cost. Table 10.2 shows a cost breakdown for a typical natural gas pipeline project.

Table 10.2

Cost Breakdown for a Typical Natural Gas Pipeline Project

Description 1 2 3 4 5 6 7 8 9 10

Pipeline Compressor stations Mainline valve stations Meter stations Pressure regulator stations SCADA and telecommunications Environmental and permitting Right of way acquisiton Engineering and construction management Contingency

11 12

Working capital AFUDC

Million $

2 to 5% 10 to 15% 6 to 10% 15 to 20% 10%

Sub-Total

Total

Copyright 2005 by Taylor & Francis Group, LLC

160.00 20.00 1.20 1.20 0.10 5.48 21.90 14.60 36.50 26.10 287.08

5%

5.00 14.35 306.43

2785_C010.fm Page 336 Friday, April 1, 2005 3:32 PM

336

GAS PIPELINE HYDRAULICS

10.3 OPERATING COSTS Once the pipeline, compressor stations, and ancillary facilities are constructed and the pipeline is put into operation, there will be annual operating costs over the useful life of the pipeline, which might be 30 to 40 years or more. These annual costs consist of the following major categories: • Compressor station fuel or electrical energy cost • Compressor station equipment maintenance and repair costs • Pipeline maintenance costs, such as pipe repair, relocation, aerial patrol, and monitoring • SCADA and telecommunication • Valve, regulator, and meter station maintenance • Utility costs, such as water and natural gas • Annual or periodic environmental and permitting costs • Lease, rental, and other recurring right of way costs • Administrative and payroll costs

Compressor station costs include periodic equipment maintenance and overhaul costs. For example, a gas turbine–driven compressor unit may have to be overhauled every 18 to 24 months. Table 10.3 shows the annual operating cost of a typical gas pipeline. Example 2 A new pipeline is being constructed to transport natural gas from a gas processing plant to a power plant 100 mi away Two project phases are envisioned. During the ﬁrst phase lasting 10 years, the amount of gas shipped is expected to be a constant

Table 10.3

Annual Operating Cost of a Typical Gas Pipeline

Description 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Salaries Payroll overhead (20%) Admin and general (50%) Vehicle expense Office expenses (6%) Misc materials and tools Compressor station maintenance Consumable materials Periodic maintenance ROW payments Utilities Gas control SCADA contract install and maintenance Internal corrosion inspection ($750,000/3 years) Cathodic protection survey Total O&M

Copyright 2005 by Taylor & Francis Group, LLC

$ per year 860,000 172,000 516,000 72,800 92,880 100,000 50,000 150,000 350,000 150,000 100,000 200,000 250,000 100,000 3,163,680

2785_C010.fm Page 337 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

337

volume of 120 MMSCFD at a 95% load factor. A pipe size of NPS 16, 0.250 in. wall thickness is required to handle the volumes with two compressor stations with a total of 5000 HP. The total pipeline cost can be estimated at $800,000 per mi and compressor station cost at $2000 per HP installed. The annual operating costs are estimated at $8 million. The pipeline construction project will be ﬁnanced by borrowing 80% of the required capital at an interest rate of 6%. The regulatory rate of return allowed on equity capital is 14%. Consider a project life of 20 years and an overall tax rate of 40%. (a) Calculate the annual cost of service for this pipeline and the transportation tariff in $/MCF. (b) The second phase, lasting the next 10 years, is projected to increase throughput to 150 MMSCFD. Calculate the transportation tariff for the second phase, considering the capital cost to increase by $20 million and the annual cost to increase to $10 million, with the same load factor as phase 1. Solution (a) First, calculate the total capital cost of facilities of phase 1. Pipeline cost = $800,000 × 100 = $80 million Compressor station cost = $2000 × 5000 = $10 million Total capital cost = $80 + $10 = $90 million 80% of this capital of $90 million will be borrowed at 6% interest for 20 years. From Equation 10.1, the annual cost to amortize the loan is Loan amortization cost =

90 × 0.8 × 0.06 1−

( 1.106 )

20

= $6.28 million

Therefore, we need to build into the cost of service the annual payment of $6.28 million to retire the debt of $72 million (80% of $90 million) over the project life of 20 years. On the remaining capital (equity) of ($90 – $72) million or $18 million, a 14% rate of return per year is allowed. Therefore, 14% of $18 million can be included in the cost of service to account for the equity capital. Annual revenue on equity capital = 0.14 × $18 million = $2.52 million Since the tax rate is 40%, the adjusted annual revenue on equity capital is $2.52 million = $4.2 million 1 − 0. 4 Next, add the operating cost of $8 million per year to the annual costs for debt and equity just calculated to arrive at the annual cost of service as follows: Annual payment to retire debt = $6.28 million Annual revenue on equity capital = $4.2 million Annual operating cost = $8 million

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 338 Friday, April 1, 2005 3:32 PM

338

GAS PIPELINE HYDRAULICS

Therefore, Annual cost of service = 6.28 + 4.2 + 8 = $18.48 million The transportation tariff at 120 MMSCFD and 95% load factor is Tariff =

18.48 × 10 6 × 10 3 = $0.4441 per MCF. 120 × 10 6 × 365 × 0.95

(b) In the second phase, which lasts 10 years, the capital cost increases by $20 million. The extra $20 million will be assumed to be ﬁnanced by 80% debt and 20% equity as before. The annual cost to amortize the debt is Loan amortization cost =

20 × 0.8 × 0.06 1−

( 1.106 )

10

= $2.17 million

The remaining capital of ($20 – $16) or $4 million is equity that, according to regulatory guidelines, can earn 14% interest. It must be noted that the interest rate and ROR used in this example are approximate and only for the purpose of illustration. The actual ROR allowed on a particular pipeline will depend on various factors such as the state of the economy, current FERC regulations, or state laws, and can range from as low as 8% to as high as 16% or more. Similarly, the interest rate of 6% used for debt amortization is also an illustrative number. The actual interest rate on debt will depend on various factors such as the state of the economy, money supply, and the federal interest rate charged by Federal Reserve (prime rate). This rate will vary with the country where the pipeline is built and the multinational bank that might ﬁnance the pipeline project. For phase 2, the annual revenue on equity capital is 4 × 0.14 = $0.56 million Accounting for a 40% tax rate, the adjusted annual revenue on equity capital is $0.56 = $0.93 million 1 − 0. 4 Therefore, for phase 2, the increase in capital of $20 million and operating cost of $2 million will result in an increase in cost of service of Annual cost of service = $2.17 + 0.93 + 2 = $5.1 million In summary, for phase 2, the total cost of services is $18.48 + $5.1 = $23.58 million At a ﬂow rate of 150 MMSCFD and 95% load factor, the tariff for phase 2 is 23.58 × 10 6 × 10 3 = $0.4534 per MCF 150 × 10 6 × 365 × 0.95

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 339 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

339

10.4 DETERMINING ECONOMIC PIPE SIZE For a particular pipeline transportation application, there is an economic or optimum pipe diameter that will result in the lowest cost of facilities. For example, a pipeline that requires 100 MMSCFD gas to be transported from a source location to a destination location may be constructed of a wide range of pipe materials and diameters. We may choose to use NPS 14, NPS 16, or NPS 18 pipe or any other pipe size for this application. Using the smallest-diameter pipe will cause the greatest pressure drop and the highest HP requirement for a given volume ﬂow rate. The largest pipe size will result in the lowest pressure drop and, hence, require the least HP. Therefore, the NPS 14 system will be the lowest in pipe material cost and highest in HP required. On the other hand, the NPS 18 system will require the least HP but considerably more pipe material cost due to the difference in pipe weight per unit length. Determining the optimum pipe size for an application will be illustrated in the next example. Example 3 A gas pipeline is to be constructed to transport 150 MMSCFD of natural gas from Dixie to Florence, 120 mi away. Consider three pipe sizes—NPS 14, NPS 16, and NPS 18—all having 0.250 in. wall thickness. Determine the most economical pipe diameter, taking into account the pipe material cost, cost of compressor stations, and fuel costs. The selection of pipe size may be based on a 20-year project life and a present value (PV ) of discounted cash ﬂow at 8% per year. Use $800 per ton for pipe material and $2000 per installed HP for compressor station cost. Fuel gas can be estimated at $3 per MCF. The following information from hydraulic analysis is available: NPS 14 pipeline: Two compressor stations, 8196 HP total. Fuel consumption is 1.64 MMSCFD. NPS 16 pipeline: One compressor station, 3875 HP. Fuel consumption is 0.78 MMSCFD. NPS 18 pipeline: One compressor station, 2060 HP. Fuel consumption is 0.41 MMSCFD. Solution First, calculate the capital cost of 120 mi of pipe for each case. From Equation 10.2, the cost of NPS 14 pipe is PMC =

10.68(14 − 0.250) × 0.250 × 120 × 800 × 5280 = $9.3 million 2000

Similarly, the cost of NPS 16 pipe is PMC =

10.68(16 − 0.250) × 0.250 × 120 × 800 × 5280 = $10.66 million 2000

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 340 Friday, April 1, 2005 3:32 PM

340

GAS PIPELINE HYDRAULICS

and the cost of NPS 18 pipe is PMC =

10.68(18 − 0.250) × 0.250 × 120 × 800 × 5280 = $12.01 million 2000

Next, calculate the installed cost of compressor stations for each pipe size. For NPS 14 pipe, the compressor station cost is 8196 × 2000 = $16.39 million For NPS 16 pipe, the compressor station cost is 3875 × 2000 = $7.75 million For NPS 18 pipe, the compressor station cost is 2060 × 2000 = $4.12 million The operating fuel cost for each case will be calculated next, considering fuel gas at $3 per MCF and 24-hour-a-day operation for 350 days a year. A shutdown for 15 days per year is allowed for maintenance and any operational upset conditions. For NPS 14 pipe, the fuel cost is 1.64 × 10 3 × 350 × 3 = $1.72 million per year For NPS 16 pipe, the fuel cost is 0.78 × 10 3 × 350 × 3 = $0.82 million per year For NPS 18 pipe, the fuel cost is 0.41 × 10 3 × 350 × 3 = $0.43 million per year The actual operating cost includes many other items besides the fuel cost. For simplicity, in this example we will only consider the fuel cost. The annual fuel cost for the project life of 20 years will be discounted at 8% in each case. This will then be added to the sum of the pipeline and compressor station capital cost to arrive at a present value (PV). The present value of a series of cash ﬂows, each equal to R for a period of n years at an interest rate of i%, is given by Equation 10.1. The PV of NPS 14 fuel cost is, from Equation 10.1, PV =

1 1.72 1− = 1.72 × 9.8181 = $16.89 million 20 0.08 (1 + 0.08)

The PV of NPS 16 fuel cost is PV = 0.82 × 9.8181 = $8.05 million

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 341 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

341

The PV of NPS 18 fuel cost is PV = 0.43 × 9.8181 = $4.22 million Therefore, adding up all costs, the PV for NPS 14 is PV14 = 9.3 + 16.39 + 16.89 = $42.58 million Adding up all costs, the PV for NPS 16 is PV16 = 10.66 + 7.75 + 8.05 = $26.46 million and adding up all costs, the PV for NPS 18 is PV18 = 12.01 + 4.12 + 4.22 = $20.35 million Therefore, we see that the lowest cost option is NPS 18 pipeline with a PV of $20.35 million.

In the preceding example, if the ﬂow rate had been lower or higher, the result may be different. For each pipe size, if we were to calculate the HP required at various ﬂow rates and the corresponding fuel consumption, we could generate a graph showing the variation of total cost with ﬂow rate. Obviously, as ﬂow rate is increased, the HP required and fuel consumption also increase. Performing these calculations for different pipe sizes will yield a graph similar to that shown in Figure 10.1. In the next example, we will consider three pipe sizes (NPS 16, NPS 18, and NPS 20) and calculate the capital cost and O&M cost for a range of ﬂow rates to develop curves similar to those shown in Figure 10.1.

NPS 20

Cost

NPS 18 NPS 16

Minimum cost Flow rate Figure 10.1

Pipeline cost vs. flow rate for various pipe sizes.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 342 Friday, April 1, 2005 3:32 PM

342

GAS PIPELINE HYDRAULICS

Example 4 For a natural gas pipeline 120 mi long, three pipe sizes were analyzed for ﬂow rate ranges of 50 to 500 MMSCFD using a hydraulic simulation software application, GASMOD (www.systek.us). The following are the pipe sizes and ﬂow rates studied: NPS 16 pipe: ﬂow rates—50 to 200 MMSCFD NPS 18 pipe: ﬂow rates—50 to 300 MMSCFD NPS 20 pipe: ﬂow rates—100 to 500 MMSCFD The wall thickness was 0.250 in. for NPS 16 and NPS 18 and 0.500 in. for NPS 20. From the hydraulic simulation, the number of compressor stations required, HP, and fuel consumption were obtained as shown in Table 10.4.

Table 10.4 Flow Rate, MMSCFD 50 100 150 175 200

Flow Rate, MMSCFD 50 100 150 175 200 250 300

Flow Rate, MMSCFD 100 150 175 200 250 300 400 500

Copyright 2005 by Taylor & Francis Group, LLC

Hydraulic Simulation Results for Three Pipe Sizes NPS 16 Compressor Stations 1 1 1 2 2 NPS 18 Compressor Stations 1 1 1 1 1 2 2 NPS 20 Compressor Stations 1 1 1 1 1 2 2 2

Total HP

Fuel, MMSCFD

49 1072 3875 5705 9203

0.01 0.21 0.78 1.14 1.84

Total HP

Fuel, MMSCFD

49 209 2060 3394 4954 9348 17902

0.01 0.04 0.41 0.68 1 1.87 3.58

Total HP

Fuel, MMSCFD

98 1053 2057 3281 6312 10519 31401 73207

0.02 0.21 0.41 0.66 1.26 2.1 6.28 14.64

2785_C010.fm Page 343 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

343

Develop annualized costs for each pipe size and ﬂow rate using the following assumptions: The capital cost of the pipe material is based on $800 per ton. For pipe installation cost, use the following: NPS 16: $50 per foot NPS 18: $60 per foot NPS 20: $80 per foot For compressor station capital cost, use $2000 per installed HP. Fuel gas can be assumed to be $3 per MCF. The project life is 20 years, and the interest rate for discounting cash ﬂow is 8%. Add 40% to the pipe and compressor capital costs to account for miscellaneous costs such as meter stations; valves; ROW; environmental, engineering, and construction management; and contingency. The pipeline is assumed to be operational 350 days a year. Solution From the given hydraulic simulation data, using the assumptions listed, we develop the total capital cost of pipe, compressor station, and miscellaneous costs. The pipe material cost is calculated from Equation 10.2, using $800 per ton for pipe material cost, as follows: For NPS 16 pipe, Pipe material cost =

10.68(16 − 0.25)0.25 × 120 × $800 × 5280 = $10.66 million 2000

Similarly, for NPS 18 pipe, Pipe material cost =

10.68(18 − 0.25)0.25 × 120 × $800 × 5280 = $12.01 million 2000

and NPS 20 pipe material cost is Pipe material cost =

10.68(20 − 0.5)0.5 × 120 × $800 × 5280 = $26.39 million 2000

These costs are shown in Table 10.5 through Table 10.7.

Copyright 2005 by Taylor & Francis Group, LLC

NPS 16 Pipe Cost Summary

Number of Flow Compressor Rate Stations

Total HP

Fuel, MMSCFD

50 100 150 175 200

49 1072 3875 5705 9203

0.01 0.21 0.78 1.14 1.84

1 1 1 2 2

Fuel, $/yr

Pipe Material, $

Pipe Labor, $

0.01 0.22 0.82 1.20 1.93

10.66 10.66 10.66 10.66 10.66

31.68 31.68 31.68 31.68 31.68

Copyright 2005 by Taylor & Francis Group, LLC

42.34 42.34 42.34 42.34 42.34

0.098 2.144 7.75 11.41 18.406

16.97 17.79 20.04 21.50 24.30

Total Capital, $

O&M Cost, $/yr

Annualized Capital, $/yr

Total Annual Cost, $

Annual Cost, $/MCF

59.41 62.27 70.12 75.25 85.04

2.00 2.00 2.00 3.00 3.00

6.05 6.34 7.14 7.66 8.66

8.06 8.56 9.96 11.86 13.59

0.4607 0.2447 0.1897 0.1937 0.1942

GAS PIPELINE HYDRAULICS

Notes: Pipe material cost = $800/ton Pipe labor cost = $50/ft for NPS 16 Compressor station cost = $2000 per installed HP Miscellaneous cost is 40% of pipe and compressor station cost Operating cost based on 350 days per year Fuel cost is $3 per MCF Capital cost is annualized at 8% interest for 20-year project life Table values in millions of dollars

NPS 16 Total Compressor Misc Pipe Station Cost, Cost, $ Cost, $ $

2785_C010.fm Page 344 Friday, April 1, 2005 3:32 PM

344

Table 10.5

Number of Flow Compressor Rate Stations

Total HP

Fuel, MMSCFD

50 100 150 175 200 250 300

49 209 2060 3394 4954 9348 17902

0.01 0.04 0.41 0.68 1.00 1.87 3.58

1 1 1 1 1 2 2

Fuel, $/yr

Pipe Material, $

Pipe Labor, $

0.01 0.04 0.43 0.71 1.05 1.96 3.76

12.01 12.01 12.01 12.01 12.01 12.01 12.01

38.02 38.02 38.02 38.02 38.02 38.02 38.02

NPS 18 Total Compressor Misc Pipe Station Cost, Cost, $ Cost, $ $ 50.03 50.03 50.03 50.03 50.03 50.03 50.03

0.098 0.418 4.12 6.788 9.908 18.696 35.804

20.05 20.18 21.66 22.73 23.97 27.49 34.33

Total Capital, $

O&M Cost, $/yr

Annualized Capital, $/yr

Total Annual Cost, $

Annual Cost, $/MCF

70.18 70.62 75.81 79.54 83.91 96.21 120.16

2.00 2.00 2.00 2.00 2.00 3.00 3.00

7.15 7.19 7.72 8.10 8.55 9.80 12.24

9.16 9.24 10.15 10.82 11.60 14.76 19.00

0.5233 0.2639 0.1934 0.1766 0.1657 0.1687 0.1809

Notes: Pipe material cost = $800/ton Pipe labor cost = $60/ft for NPS 18 Compressor station cost = $2000 per installed HP Miscellaneous cost is 40% of pipe and compressor station cost Operating cost based on 350 days per year Fuel cost is $3 per MCF Capital cost is annualized at 8% interest for 20-year project life Table values in millions of dollars

345

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 345 Friday, April 1, 2005 3:32 PM

NPS 18 Pipe Cost Summary

PIPELINE ECONOMICS

Table 10.6

NPS 20 Pipe Cost Summary

Number of Flow Compressor Rate Stations

Total HP

Fuel, MMSCFD

Fuel, $/yr

Pipe Material, $

100 150 175 200 250 300 400 500

98 1053 2057 3281 6312 10519 31401 73207

0.02 0.21 0.41 0.66 1.26 2.1 6.28 14.64

0.02 0.22 0.43 0.69 1.32 2.21 6.59 15.37

26.39 26.39 26.39 26.39 26.39 26.39 26.39 26.39

1 1 1 1 1 2 2 2

NPS 20 Pipe Total Compressor Misc Labor, Pipe Station Cost, $ Cost, $ Cost, $ $

Total Capital, $

O&M Cost, $/yr

50.69 50.69 50.69 50.69 50.69 50.69 50.69 50.69

108.18 110.86 113.67 117.10 125.58 137.36 195.83 312.89

2.00 2.00 2.00 2.00 2.00 3.00 3.00 3.00

Copyright 2005 by Taylor & Francis Group, LLC

0.196 2.106 4.114 6.562 12.624 21.038 62.802 146.414

30.91 31.67 32.48 33.46 35.88 39.25 55.95 89.40

11.02 11.29 11.58 11.93 12.79 13.99 19.95 31.87

13.04 13.51 14.01 14.62 16.11 19.20 29.54 50.24

0.3726 0.2574 0.2287 0.2089 0.1842 0.1828 0.2110 0.2871

GAS PIPELINE HYDRAULICS

Notes: Pipe material cost = $800/ton Pipe labor cost = $80/ft for NPS 20 Compressor station cost = $2000 per installed HP Miscellaneous cost is 40% of pipe and compressor station cost Operating cost based on 350 days per year Fuel cost is $3 per MCF Capital cost is annualized at 8% interest for 20-year project life Table values in millions of dollars

77.08 77.08 77.08 77.08 77.08 77.08 77.08 77.08

Total Annual Annualized Annual Cost, Capital, $/yr Cost, $ $/MCF

2785_C010.fm Page 346 Friday, April 1, 2005 3:32 PM

346

Table 10.7

2785_C010.fm Page 347 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

347

The labor cost for installing pipe is calculated as follows: For NPS 16 pipe, Pipe installation cost = $50 × 5280 × 120 = $31.68 million Similarly, for NPS 18 pipe, Pipe installation cost = $60 × 5280 × 120 = $38.02 million and for NPS 20 pipe, Pipe installation cost = $80 × 5280 × 120 = $50.69 million Next, we calculate the installation cost of compressor stations using $2000 per installed HP. For the NPS 16 pipe at 100 MMSCFD ﬂow rate, the HP required is 1072 and the installation cost is $2000 × 1072 = $2.14 million Similarly, the installation costs of each compressor station for all cases are calculated and tabulated as shown in Table 10.5 through Table 10.7. The miscellaneous cost is 40% of the sum of the pipe cost and compressor station cost, as follows: Pipe material cost = $10.66 million Pipe installation cost = $31.68 million Compressor station cost = $2.14 million Thus, for NPS 16 pipe at 100 MMSCFD, Miscellaneous cost = 0.40 × (10.66 + 31.68 + 2.14) = $17.79 million The operation and maintenance costs are added to the annual fuel cost to obtain the total annual cost. The total capital cost is annualized at 8% interest for 20 years and added to the O&M and fuel costs. For example, for NPS 16 pipe at 100 MMSCFD ﬂow rate, the total capital cost of $62.27 million is annualized at $6.34 million and added to the O&M and fuel costs to obtain the total annual cost of $8.56 million. Dividing this annual cost by the gas transported per year, we obtain the annual cost per MCF as follows: Annual cost per MCF =

8.56 × 10 6 × 10 3 = $0.2447 100 × 10 6 × 350

Similarly, for each pipe size and ﬂow rate, the values are tabulated as shown in Table 10.5 through Table 10.7. Upon reviewing Table 10.5 for NPS 16 pipe, we see that the annual cost per MCF decreases from $0.4607 to $0.1897 as the ﬂow rate increases from 50 to 150 MMSCFD.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 348 Friday, April 1, 2005 3:32 PM

348

GAS PIPELINE HYDRAULICS

After that, it increases with ﬂow rate and reaches a value of $0.1942 at 200 MMSCFD. Therefore, for NPS 16 pipe, 150 MMSCFD is the optimum ﬂow rate that results in the lowest transportation cost. Similarly, from Table 10.6, for NPS 18 pipe, the annual cost per MCF decreases from $0.5233 to $0.1657 as the ﬂow rate increases from 50 to 200 MMSCFD. After that, it increases with ﬂow rate and reaches a value of $0.1809 at 300 MMSCFD. Therefore, for NPS 18 pipe, 200 MMSCFD is the optimum ﬂow rate that results in the lowest transportation cost. Finally, from Table 10.7, for NPS 20 pipe, the annual cost per MCF decreases from $0.3726 to $0.1828 as the ﬂow rate increases from 100 to 300 MMSCFD. After that, it increases with ﬂow rate and reaches a value of $0.2871 at 500 MMSCFD. Therefore, for NPS 20 pipe, 300 MMSCFD is the optimum ﬂow rate that results in the lowest transportation cost. A plot of the annual cost per MCF vs. ﬂow rate for the three pipe sizes is shown in Figure 10.2.

In the preceding calculations, to simplify matters, several assumptions were made. Miscellaneous costs were estimated as a percentage of the pipeline and compressor station costs. Also, we considered the annual costs to be constant from year to year. A more nearly accurate calculation would be to escalate the annual costs by a percentage each year to account for inﬂation, using the Consumer Price Index (CPI). Nevertheless, the preceding calculations illustrate a methodology of economic analysis to determine the most optimum pipe size. Example 5 In Chapter 5, we compared expanding the capacity of the gas pipeline from Windsor to Cardiff using two options—installing intermediate compressor stations or installing pipe loops. Using the results of Example 1 in Chapter 5, compare the two options, 0.6000 0.5000

$/MCF

0.4000 NPS 20 NPS 18 NPS 16

0.3000 0.2000 0.1000 0.0000

50

100

150

175

200

250

300

Flow rate, MMSCFD Figure 10.2

Annualized cost vs. flow rate for three pipe sizes.

Copyright 2005 by Taylor & Francis Group, LLC

400

2785_C010.fm Page 349 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

349

taking into account the capital cost, operating cost, and fuel cost and considering a project life of 25 years. The capital will be ﬁnanced with 70% debt at 8% interest. The regulatory return allowed on the 30% equity is 12%. The tax rate can be assumed to be 35%. Fuel consumption is 0.2 MCF per day per HP and fuel gas cost is $3 per MCF. Assume 350 days of operation per year. Calculate the annualized cost of service and transportation tariff for both options. It is expected that the annual O&M cost will increase by $2 million for phase 1 and an additional $3 million for phase 2 compressor station options. For the looping option, the incremental O&M cost is $0.5 million for phase 1 and $0.75 million for phase 2. Solution Phase 1 expansion This expansion results in a ﬂow rate of 238.41 MMSCFD, and the compressor station option requires installing the following HP: Windsor compressor station—8468 HP Avon compressor station—3659 HP Total HP = 8468 + 3659 = 12,127 HP The incremental HP for phase 1 was calculated as ∆HP = 12,127 – 7064 = 5063 HP The cost of this incremental HP based on $2000 per installed HP is ∆Capital cost = 5063 × 2000 = $10.13 million The incremental fuel cost for 5063 HP is ∆Fuel cost = 5063 × 0.2 × $3 × 350 = $1.06 million per year. The incremental capital cost of $10.13 million will be funded by 70% debt and 30% equity. The debt capital = 10.13 × 0.7 = $7.09 million. Loan amortization cost =

10.13 × 0.7 × 0.08 1−

( 1.108 )

25

= $0.66 million per year

The remaining capital of ($10.13 – $7.09) million or $3.04 million is equity that, according to regulatory guidelines, can earn 12% interest. The annual revenue allowed on equity capital is 3.04 × 0.12 = $0.36 million Accounting for a 35% tax rate, the adjusted annual revenue on equity capital is $0.36 = $0.55 million 1 − 0.35

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 350 Friday, April 1, 2005 3:32 PM

350

GAS PIPELINE HYDRAULICS

Next, add the O&M cost increase of $2 million per year and the fuel cost of $1.06 million to the annual costs for debt and equity just calculated to arrive at the annual cost of service as follows: Annual Annual Annual Annual

payment to retire debt = $0.66 million revenue on equity capital = $0.36 million operating cost = $2.0 million fuel cost = $1.06 million

Therefore, the incremental annual cost of service for phase 1 expansion compressor station option is $0.66 + 0.36 + 2.0 + 1.06 = $4.08 million. This amount is the incremental annual cost of service over and above the cost of service for the initial ﬂow rate of 188.41 MMSCFD. The incremental tariff for an incremental ﬂow rate of 50 MMSCFD for phase 1 expansion is Incremental tariff =

4.08 × 10 6 × 10 3 = $0.2331 per MCF 50 × 10 6 × 350

Next, we calculate the cost of service and tariff considering the looping option. In Example 1 of Chapter 5, for phase 1 expansion, we required installation of 50.03 mi of loop at a cost of $25.02 million. In addition to this cost of pipe loop, we must include the cost of the increased horsepower requirement at Windsor for the phase 1 ﬂow rate, which was calculated at 1404 HP. At $2000 per installed HP, the extra cost for incremental HP is $2.81 million. Thus, for phase 1 the total cost of looping pipe upstream of Cardiff and increased HP cost at the Windsor compressor station is calculated as $25.02 + $2.81 = $27.83 million The incremental fuel cost for the extra 1404 HP is ∆Fuel cost = 1404 × 0.2 × $3 × 350 = $0.30 million per year The incremental capital of $27.83 million for the looping option would also be funded by 70% debt and 30% equity. Debt capital = 27.83 × 0.7 = $19.48 million Loan amortization cost =

19.48 × 0.08 1−

( 1.108 )

25

= $1.82 million per year

The remaining capital of ($27.83 – $19.48) million or $8.35 million is equity that, according to regulatory guidelines, can earn 12% interest.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 351 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

351

The annual revenue allowed on equity capital is 8.35 × 0.12 = $1.0 million Accounting for a 35% tax rate, the adjusted annual revenue on equity capital is $1.0 = $1.54 million 1 − 0.35 Next, add the O&M cost increase of $0.5 million per year and the fuel cost of $0.30 million to the annual costs for debt and equity just calculated to arrive at the incremental annual cost of service as follows: Annual Annual Annual Annual

payment to retire debt = $1.82 million revenue on equity capital = $1.0 million operating cost = $0.5 million fuel cost = $0.3 million

Therefore, the incremental annual cost of service for phase 1 expansion looping option is $1.82 + 1.0 + 0.5 + 0.3 = $3.62 million. This amount is the incremental annual cost of service over and above the cost of service for the initial ﬂow rate of 188.41 MMSCFD. The incremental tariff for an incremental ﬂow rate of 50 MMSCFD for the phase 1 expansion looping option is Incremental tariff =

3.62 × 10 6 × 10 3 = $0.2069 per MCF 50 × 10 6 × 350

We can summarize the calculations as follows: For the phase 1 expansion, compressor station option: Incremental annual cost of service = $4.08 million Incremental tariff = $0.2331 per MCF For the phase 1 expansion, looping option: Incremental annual cost of service = $3.62 million Incremental tariff = $0.2069 per MCF It can be seen that the incremental annual cost of service and the incremental tariff for phase 1 expansion are less in the looping option than the compressor station option. Therefore, for phase 1 expansion, the looping option is the preferred choice. For phase 2 expansion, the throughput increase of 50 MMSCFD will be on top of phase 1 expansion. Since the preferred choice for phase 1 expansion is the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 352 Friday, April 1, 2005 3:32 PM

352

GAS PIPELINE HYDRAULICS

looping option, we must consider the increase in facilities required for phase 2 with 50.03 mi of pipe loop already installed. In Example 1 of Chapter 5, for phase 2, the loop required was calculated to be 76.26 mi. The incremental HP at Windsor was calculated as 1775 HP. Also, the incremental looping required and cost of increased HP at Windsor over the phase 1 values were calculated to be $16.66 million. The incremental fuel cost for the extra 1775 HP is ∆Fuel cost = 1775 × 0.2 × $3 × 350 = $0.37 million per year The incremental capital of $16.66 million for the phase 2 looping option would also be funded by 70% debt and 30% equity. Debt capital = 16.66 × 0.7 = $11.66 million Loan amortization cost =

11.66 × 0.08 1−

( 1.108 )

25

= $1.09 million per year

The remaining capital of ($16.66 – $11.66) million or $5.0 million is equity that, according to regulatory guidelines, can earn 12% interest. The annual revenue allowed on equity capital is $5.0 × 0.12 = $0.6 million Accounting for a 35% tax rate, the adjusted annual revenue on equity capital is $0.6 = $0.92 million 1 − 0.35 Next, add the O&M cost increase of $0.75 million per year and the fuel cost of $0.37 million to the annual costs for debt and equity just calculated to arrive at the incremental annual cost of service for the phase 2 looping expansion as follows: Annual Annual Annual Annual

payment to retire debt = $1.09 million revenue on equity capital = $0.6 million operating cost = $0.75 million fuel cost = $0.37 million

Therefore, the incremental annual cost of service for phase 2 expansion looping option is $1.09 + 0.6 + 0.75 + 0.37 = $2.81 million. This amount is the incremental annual cost of service over and above the cost of service for the phase 1 ﬂow rate of 238.41 MMSCFD.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 353 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

353

The incremental tariff for an incremental ﬂow rate of 50 MMSCFD for the phase 2 expansion looping option is Incremental tariff =

2.81 × 10 6 × 10 3 = $0.1606 per MCF 50 × 10 6 × 350

In summary, For the phase 2 expansion, looping option: Incremental annual cost of service = $2.81 million Incremental tariff = $0.1606 per MCF These incremental costs are over and above the phase 1 numbers. It must be noted that we did not consider a compressor station option for phase 2 expansion. This is because the preferred option for phase 1 expansion was installing loop. Since approximately 50 mi of pipe loop was already installed for phase 1, we simply looked at adding approximately 26 mi of extra loop for phase 2. For comparison, we could determine additional compressor station requirements for phase 2 instead of extending the loop. This is left as an exercise for the reader.

10.5 SUMMARY In this chapter the economic aspects of a natural gas pipeline transportation were reviewed. A method for determining the optimum pipe size necessary to transport a certain ﬂow rate was discussed. We introduced concepts of the capital cost of pipeline and compressor stations and the annual operating and maintenance costs. The fuel consumption calculations were also explained. Taking into account time value of money and the rate of return allowed on an equity investment in pipeline facilities, we calculated an annual cost of transporting gas. From this annual cost, the transportation tariff was calculated. The economic pipe size for a particular application was illustrated using three different pipe sizes and estimating the initial capital cost and annual operating costs. A typical pipeline expansion scenario with the option of installing compressors vs. pipe loops was also explained using economic principles. Additionally, the major components of the capital cost of a typical pipeline system were reviewed. PROBLEMS 1. A natural gas pipeline transports 120 MMSCFD at a load factor of 95%. The capital cost is estimated at $70 million and the annual operating cost is $6 million. Amortizing the capital at 8% for a project life of 20 years, calculate the cost of the service and transportation tariffs for this pipeline. 2. A new pipeline is being constructed to transport natural gas from a processing plant to a power plant 150 mi away. An initial phase and an expansion phase

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 354 Friday, April 1, 2005 3:32 PM

354

GAS PIPELINE HYDRAULICS

are contemplated. During the initial phase lasting 10 years, the amount of gas shipped is expected to be a constant volume of 100 MMSCFD at a 95% load factor. A pipe size of NPS 18, 0.250 in. wall thickness, is required to handle the volumes with two compressor stations of 5000 HP total. The total pipeline cost can be estimated at $750,000 per mi and the compressor station cost at $2000 per HP installed. The annual operating costs are estimated at $6 million. The construction project will be ﬁnanced by borrowing 75% of the required capital at an interest rate of 6%. The regulatory rate of return allowed on equity is 13%. Consider a project life of 25 years and an overall tax rate of 36%. a. Calculate the annual cost of service for this pipeline and the transportation tariff in $/MCF. b. The second phase, lasting the next 10 years, is projected to increase throughput to 150 MMSCFD. Calculate the transportation tariff for the expansion phase, considering the capital cost to increase by $30 million and the annual cost to increase by $4 million, with the same load factor as before.

3. A gas pipeline is to be constructed to transport 200 MMSCFD of natural gas from Jackson to Columbus, 180 mi away. Consider three pipe sizes—NPS 18, NPS 20, and NPS 24—all constructed of API 5L-X52 pipe with suitable wall thickness for an MOP of 1400 psig. Determine the most economical pipe diameter, taking into account the pipe material cost, cost of compressor stations, and fuel costs. The selection of pipe size can be based on a 30-year project life and a present value of discounted cash ﬂow at 6% per year. Use $750 per ton for pipe material and $2000 per installed HP for compressor station cost. Fuel gas can be estimated at $3 per MCF.

REFERENCES 1. Nayyar, M.L., Piping Handbook, 7th ed., McGraw-Hill, New York, 2000. 2. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 3. Pipeline Design for Hydrocarbon Gases and Liquids, American Society of Civil Engineers, New York, 1975. 4. Katz, D.L. et al., Handbook of Natural Gas Engineering, McGraw-Hill, New York, 1959.

Copyright 2005 by Taylor & Francis Group, LLC

2785_A001.fm Page 355 Friday, March 11, 2005 5:31 PM

APPENDIX

A

Units and Conversions

355

Copyright 2005 by Taylor & Francis Group, LLC

millimeter (mm) 1 meter (m) = 1,000 mm 1 kilometer (km) = 1,000 m square meter (m2) 1 hectare = 10,000 m2 cubic meter (m3) kilogram/cubic meter (kg/m3) Newton per cubic meter (N/m3) 1 1 1 1 m2/s Stoke (S), centiStoke (cSt)

pound (lb) inch (in.) 1 foot (ft) = 12 in. 1 mile (mi) = 5,280 ft square foot (ft2) 1 acre = 43,560 ft2 cubic foot (ft3) slug per cubic foot (slug/ft3) pound per cubic foot (lb/ft3) lb/ft-s lb-s/ft2

ft2/s cubic foot/hour (ft3/h) cubic foot/day (ft3/day) million Std ft3/day (MMSCFD) pound (lb)

Weight

Copyright 2005 by Taylor & Francis Group, LLC

Length

Area

Volume

Density

Specific weight

Viscosity (absolute or dynamic)

Viscosity (kinematic)

Flow rate

Force

1 N/m3 = 0.0064 lb/ft3

1 lb/ft3 = 157.09 N/m3

1 m3/h = 35.3134 ft3/h

1 ft3/h = 0.02832 m3/h

1 lb = 4.4482 N

cubic meter/hour (m3/h) cubic meter/day (m3/day) million std m3/day (Mm3/day) Newton (N) = kg-m/s2

1 N = 0.2248 lb

1 m2/s = 10.7639 ft2/s 1 cSt = 1.076 × 10−5 ft2/s

1 lb-s/ft = 47.88 N-s/m 1lb-s/ft2 = 478.8 Poise

1 kg/m3 = 0.0019 slug/ft3

1 slug/ft3 = 515.38 kg/m3

1 cP = 6.7197 × 10-4 lb/ft-s 1 N-s/m2 = 0.0209 lb-s/ft2 1 Poise = 0.00209 lb-s/ft2

1 m3 = 35.3134 ft3

1 ft3 = 0.02832 m3

2

1 m2 = 10.764 ft2 1 hectare = 2.4711 acre

1 ft2 = 0.0929 m2 1 acre = 0.4047 hectare

2

1 mm = 0.0394 in 1 m = 3.2808 ft 1 km = 0.6214 mi

kg = 0.0685 slug kg = 2.205 lb t = 1.1023 U.S. ton t = 0.9842 long ton

1 in = 25.4 mm 1 ft = 0.3048 m 1 mi = 1.609 km

1 1 1 1

1 N = 0.2248 lb

lb = 0.45359 kg slug = 14.594 kg U.S. ton = 0.9072 t long ton = 1.016 t

SI to USCS Conversion

1 lb = 4.4482 N

1 1 1 1

USCS to SI Conversion

356

1 ft2/s = 0.092903 m2/s

Poise (P) = 0.1 Pa-s centiPoise (cP) = 0.01 P Poise = 1 dyne-s/cm2 Poise = 0.1 N-s/m2

Newton (N) = kg-m/s2

metric tonne (t) = 1,000 kg

kilogram (kg)

SI Units

slug (slug) pound mass (lbm) 1 U.S. ton = 2,000 lb 1 long ton = 2,240 lb

USCS Units

Mass

Item

USCS Units—U.S. Customary System of Units SI Units—Systeme International Units (modified metric)

2785_A001.fm Page 356 Friday, March 11, 2005 5:31 PM

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

W/m/°C W/m2/°C

Btu/hr/ft/°F Btu/hr/ft2/°F Btu/lb/°F

Thermal conductivity

Heat transfer coefficient

Specific heat

kJ/kg/°C

degree Celsius (°C) 1 degree Kelvin (K) = °C + 273

degree Fahrenheit (°F) 1 degree Rankin (°R) = °F + 460

1°C = (°F − 32)/1.8 1 K = °R/1.8 1 W/m/°C = 0.5778 Btu/hr/ft/F 1 W/m2/°C = 0.1761 Btu/hr/ft2/°F 1 kJ/kg/°C = 0.2388 Btu/lb/F

1°F = 9/5°C + 32 1°R = 1.8 K 1 Btu/hr/ft/°F = 1.7307W/m/°C 1 Btu/hr/ft2/°F = 5.6781 W/m2/°C 1 Btu/lb/°F = 4.1869 kJ/kg/°C

1 kW = 1.3405 HP

1 HP = 0.746 kW

Temperature

1 W = 3.4121 Btu/hr

1 Btu/hr = 0.2931 W

Joule/second (J/s) watt (W) = J/s 1 kilowatt (kW) = 1,000 W

ft-lb/min Btu/hour horsepower (HP) 1 HP = 33,000 ft-lb/min

Power

1 kJ = 0.9478 Btu

1 Btu = 1055.0 J

Joule (J) = N-m

foot-pound (ft-lb) British thermal unit (Btu) 1 Btu = 778 ft-lb

Work and energy

1 m/s = 3.281 ft/s

1 ft/s = 0.3048 m/s

meter/second (m/s)

foot/second (ft/s) mile/hour (mi/h) = 1.4667 ft/s

1 Bar = 14.5 psi 1 kg/cm2 = 14.22 psi

1 psi = 0.069 Bar 1 psi = 0.0703 kg/cm2

Velocity

1 kPa = 0.145 psi

1 psi = 6.895 kPa

Pascal (Pa) = N/m2 1 kiloPascal (kPa) = 1,000 Pa 1 megaPascal (MPa) = 1,000 kPa 1 Bar = 100 kPa kilogram/sq. centimeter (kg/cm2)

pound/square inch, lb/in2 (psi) 1lb/ft2 = 144 psi

Pressure

2785_A001.fm Page 357 Friday, March 11, 2005 5:31 PM

APPENDIX A 357

2785_A002.fm Page 359 Friday, March 11, 2005 5:32 PM

APPENDIX

B

Physical Properties of Various Gases

359

Copyright 2005 by Taylor & Francis Group, LLC

Methane Ethane Propane Isobutane n-butane Iso-pentane n-pentane Neo-pentane n-hexane 2-methyl pentane 3-methyl pentane Neo hexane 2,3-dimethylbutane n-Heptane 2-Methylhexane 3-Methylhexane 3-Ethylpentane 2,2-Dimethylpentane 2,4-Dimethylpentane 3,3-Dimethylpentane Triptane n-octane Di Isobutyl Isooctane n-Nonane n-Decane Cyclopentane Methylcyclopentane Cyclohexane Methylcyclohexane

Gas CH4 C 2H 6 C 3H 8 C4H10 C4H10 C5H12 C5H12 C5H12 C6H14 C6H14 C6H14 C6H14 C6H14 C7H16 C7H16 C7H16 C7H16 C7H16 C7H16 C7H16 C7H16 C8H18 C8H18 C8H18 C9H20 C10H22 C5H10 C6H12 C6H12 C7H14

Formula 16.0430 30.0700 44.0970 58.1230 58.1230 72.1500 72.1500 72.1500 86.1770 86.1770 86.1770 86.1770 86.1770 100.2040 100.2040 100.2040 100.2040 100.2040 100.2040 100.2040 100.2040 114.2310 114.2310 114.2310 128.2580 142.2850 70.1340 84.1610 84.1610 98.1880

Molecular Weight 5,000 800 188.65 72.581 51.706 20.443 15.575 36.72 4.9596 6.769 6.103 9.859 7.406 1.621 2.273 2.13 2.012 3.494 3.294 2.775 3.376 0.5371 1.1020 1.7090 0.17155 0.06088 9.917 4.491 3.267 1.609

Vapor Pressure, psia at 100F 666.0 707.0 617.0 527.9 548.8 490.4 488.1 464.0 436.9 436.6 452.5 446.7 454.0 396.8 396.0 407.6 419.2 401.8 397.4 427.9 427.9 360.7 361.1 372.7 330.7 304.6 653.8 548.8 590.7 503.4

−116.66 90.07 205.93 274.4 305.52 368.96 385.7 321.01 453.8 435.76 448.2 419.92 440.08 512.8 494.44 503.62 513.16 476.98 475.72 505.6 496.24 564.15 530.26 519.28 610.72 652.1 461.1 499.28 536.6 570.2 0.0988 0.0783 0.0727 0.0714 0.0703 0.0684 0.0695 0.0673 0.0688 0.0682 0.0682 0.0667 0.0665 0.0682 0.0673 0.0646 0.0665 0.0665 0.0667 0.0662 0.0636 0.0673 0.0676 0.0657 0.0693 0.0702 0.0594 0.0607 0.0586 0.0600

Critical Constants Pressure, Temp., Volume, F ft3/lb psia 0.5539 1.0382 1.5226 2.0068 2.0068 2.4912 2.4912 2.4912 2.9755 2.9755 2.9755 2.9755 2.9755 3.4598 3.4598 3.4598 3.4598 3.4598 3.4598 3.4598 3.4598 3.9441 3.9441 3.9441 4.4284 4.9127 2.4215 2.9059 2.9059 3.3902

23.654 12.620 8.6059 6.5291 6.5291 5.2596 5.2596 5.2596 4.4035 4.4035 4.4035 4.4035 4.4035 3.7872 3.7872 3.7872 3.7872 3.7872 3.7872 3.7872 3.7872 3.322 3.322 3.322 2.9588 2.6671 5.411 4.509 4.509 3.8649

Ideal Gas, 14.696 psia, 60F Spgr (air = 1.00) ft3/lb-gas 0.52676 0.40789 0.38847 0.38669 0.39500 0.38448 0.38831 0.39038 0.38631 0.38526 0.37902 0.38231 0.37762 0.38449 0.38170 0.37882 0.38646 0.38651 0.39627 0.38306 0.37724 0.38334 0.37571 0.38222 0.38248 0.38181 0.27122 0.30027 0.29012 0.31902

Specific Heat, Btu/lb/F 14.696 psia, 60F Ideal Gas

2785_A002.fm Page 360 Friday, March 11, 2005 5:32 PM

360 GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

Ethylene Propylene Butylene Cis-2-butene Trans-2-butene Isobutene 1-Pentene 1,2-Butadene 1,3-Butadene Isoprene Acetylene Benzene Toluene Ethyl-benzene o-Xylene m-Xylene p-Xylene Styrene Isopropylbenzene Methyl alcohol Ethyl alcohol Carbon monoxide Carbon dioxide Hydrogen sulfide Sulfur dioxide Ammonia Air Hydrogen Oxygen Nitrogen Chlorine Water Helium Hydrogen Chloride

C 2H 4 C 3H 6 C 4H 8 C 4H 8 C 4H 8 C 4H 8 C5H10 C 4H 6 C 4H 6 C 5H 8 C 2H 2 C 6H 6 C 7H 8 C8H10 C8H10 C8H10 C8H10 C 8H 8 C9H12 CH4 O C 2H 6 O CO CO2 H 2S SO2 NH3 N2+O2 H2 O2 N2 Cl2 H 2O He HCl

28.0540 42.0810 56.1080 56.1080 56.1080 56.1080 70.1340 54.0920 54.0920 68.1190 26.0380 78.1140 92.1410 106.1670 106.1670 106.1670 106.1670 104.1520 120.1940 32.0420 46.0690 28.0100 44.0100 34.0820 64.0650 17.0305 28.9625 2.0159 31.9988 28.0134 70.9054 18.0153 4.0026 36.4606

Copyright 2005 by Taylor & Francis Group, LLC

906.71

157.3 0.95

394.59 85.46 211.9

3.225 1.033 0.3716 0.2643 0.3265 0.3424 0.2582 0.188 4.631 2.313

1400 232.8 62.55 45.97 49.88 64.95 19.12 36.53 59.46 16.68

731.0 676.6 586.4 615.4 574.9 580.2 509.5 656.0 620.3 582.0 890.4 710.4 595.5 523 541.6 512.9 509.2 587.8 465.4 1174 891.7 506.8 1071 1306 1143 1647 546.9 187.5 731.4 493 1157 3200.1 32.99 1205

48.54 198.31 296.18 324.31 311.8 292.49 376.86 354 306 403 95.29 552.15 605.5 651.22 674.85 650.95 649.47 703 676.2 463.01 465.31 −220.51 87.73 212.4 315.7 270.2 −221.29 −400.3 −181.4 −232.48 290.69 705.1 −450.31 124.75

0.0746 0.0717 0.0683 0.0667 0.0679 0.0681 0.0674 0.0700 0.0653 0.0660 0.0693 0.0531 0.0549 0.0564 0.0557 0.0567 0.0572 0.0534 0.0569 0.0590 0.0581 0.0527 0.0342 0.0461 0.0305 0.0681 0.0517 0.5101 0.0367 0.0510 0.0280 0.04975 0.2300 0.0356

0.9686 1.4529 1.9373 1.9373 1.9373 1.9373 2.4215 1.8677 1.8677 2.3520 0.8990 2.6971 3.1814 3.6657 3.6657 3.6657 3.6657 3.5961 4.1500 1.1063 1.5906 0.9671 1.5196 1.1768 2.2120 0.5880 1.0000 0.06960 1.1048 0.9672 2.4482 0.62202 0.1382 1.2589

13.527 9.0179 6.7636 6.7636 6.7636 6.7636 5.411 7.0156 7.0156 5.571 14.574 4.8581 4.1184 3.5744 3.5744 3.5744 3.5744 3.6435 3.1573 11.843 8.2372 13.548 8.6229 11.134 5.9235 22.283 13.103 188.25 11.859 13.546 5.3519 21.065 94.814 10.408

0.35789 0.35683 0.35535 0.33275 0.35574 0.36636 0.35944 0.34347 0.34223 0.35072 0.39754 0.24295 0.26005 0.27768 0.28964 0.27427 0.27470 0.26682 0.30704 0.32429 0.33074 0.24847 0.19909 0.23838 0.14802 0.49678 0.2398 3.4066 0.21897 0.24833 0.11375 0.44469 1.24040 0.19086

2785_A002.fm Page 361 Friday, March 11, 2005 5:32 PM

APPENDIX B 361

2785_A003.fm Page 363 Friday, March 11, 2005 5:33 PM

APPENDIX

C

Pipe Properties—U.S. Customary System of Units

363

Copyright 2005 by Taylor & Francis Group, LLC

Copyright 2005 by Taylor & Francis Group, LLC

21/2

0.84 0.84 0.84 0.84 0.84 0.84 1.05 1.05 1.05 1.05 1.05 1.05 1.315 1.315 1.315 1.315 1.315 1.315 1.900 1.900 1.900 1.900 1.900 1.900 2.375 2.375 2.375 2.375 2.375 2.375 2.875 2.875

Outside Dia, in.

40 80 160

40 80 160

40 80 160

40 80 160

40 80 160

a

XXS

Std XS

XXS

Std XS

XXS

Std XS

XXS

Std XS

XXS

Std XS

5S 10S

5S 10S 40S 80S

5S 10S 40S 80S

5S 10S 40S 80S

5S 10S 40S 80S

5S 10S 40S 80S

Schedule b c 0.065 0.083 0.109 0.147 0.187 0.294 0.065 0.083 0.113 0.154 0.218 0.308 0.065 0.109 0.330 0.179 0.250 0.358 0.065 0.109 0.145 0.200 0.281 0.400 0.065 0.109 0.154 0.218 0.343 0.436 0.083 0.12

Wall Thickness, in. 0.710 0.674 0.622 0.546 0.466 0.252 0.920 0.884 0.824 0.742 0.614 0.434 1.185 1.097 0.655 0.957 0.815 0.599 1.770 1.682 1.610 1.500 1.338 1.100 2.245 2.157 2.067 1.939 1.689 1.503 2.709 2.635

Inside Dia, in. 0.3957 0.3566 0.3037 0.2340 0.1705 0.0499 0.6644 0.6134 0.5330 0.4322 0.2959 0.1479 1.1023 0.9447 0.3368 0.7189 0.5214 0.2817 2.4593 2.2209 2.0348 1.7663 1.4053 0.9499 3.9564 3.6523 3.3539 2.9514 2.2394 1.7733 5.7609 5.4504

Inside Area, in.2 0.22 0.22 0.22 0.22 0.22 0.22 0.27 0.27 0.27 0.27 0.27 0.27 0.34 0.34 0.34 0.34 0.34 0.34 0.50 0.50 0.50 0.50 0.50 0.50 0.62 0.62 0.62 0.62 0.62 0.62 0.75 0.75

Surface Area, ft2/ft 0.0027 0.0025 0.0021 0.0016 0.0012 0.0003 0.0046 0.0043 0.0037 0.0030 0.0021 0.0010 0.0077 0.0066 0.0023 0.0050 0.0036 0.0020 0.0171 0.0154 0.0141 0.0123 0.0098 0.0066 0.0275 0.0254 0.0233 0.0205 0.0156 0.0123 0.0400 0.0379

Volume, ft3/ft 0.54 0.67 0.85 1.09 1.30 1.71 0.68 0.86 1.13 1.47 1.94 2.44 0.87 1.40 3.47 2.17 2.84 3.66 1.27 2.08 2.72 3.63 4.86 6.41 1.60 2.64 3.65 5.02 7.44 9.03 2.47 3.53

Pipe Weight, lb/ft 0.02 0.02 0.02 0.01 0.01 0.00 0.04 0.04 0.03 0.03 0.02 0.01 0.06 0.05 0.02 0.04 0.03 0.02 0.14 0.13 0.12 0.10 0.08 0.06 0.23 0.21 0.19 0.17 0.13 0.10 0.33 0.32

Water Weight, lb/ft

364

2

11/2

1

3/4

1/2

Nominal Pipe Size, NPS

2785_A003.fm Page 364 Friday, March 11, 2005 5:33 PM

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

8

6

4

3

2.875 2.875 2.875 2.875 3.5 3.5 3.5 3.5 3.5 3.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 6.625 6.625 6.625 6.625 6.625 6.625 6.625 8.625 8.625 8.625 8.625 8.625 8.625 8.625 8.625 8.625 8.625 20 30 40 60 80 100 120 140

40 80 120 160

40 80 120 160

40 80 160

40 80 160

40S 80S

XS

5S 10S

5S 10S 40S 80S

5S 10S 40S 80S

5S 10S 40S 80S

Std

XXS

Std XS

XXS

Std XS

XXS

Std XS

XXS

Std XS

0.203 0.276 0.375 0.552 0.083 0.120 0.216 0.300 0.437 0.600 0.083 0.120 0.237 0.337 0.437 0.531 0.674 0.109 0.134 0.280 0.432 0.562 0.718 0.864 0.109 0.148 0.250 0.277 0.322 0.406 0.500 0.593 0.718 0.812

2.469 2.323 2.125 1.771 3.334 3.260 3.068 2.900 2.626 2.300 4.334 4.260 4.026 3.826 3.626 3.438 3.152 6.407 6.357 6.065 5.761 5.501 5.189 4.897 8.407 8.329 8.125 8.071 7.981 7.813 7.625 7.439 7.189 7.001

4.7853 4.2361 3.5448 2.4621 8.7257 8.3427 7.3889 6.6019 5.4133 4.1527 14.7451 14.2459 12.7238 11.4910 10.3211 9.2786 7.7991 32.2240 31.7230 28.8756 26.0535 23.7549 21.1367 18.8248 55.4820 54.4572 51.8223 51.1357 50.0016 47.9187 45.6404 43.4409 40.5702 38.4760

0.75 0.75 0.75 0.75 0.92 0.92 0.92 0.92 0.92 0.92 1.18 1.18 1.18 1.18 1.18 1.18 1.18 1.73 1.73 1.73 1.73 1.73 1.73 1.73 2.26 2.26 2.26 2.26 2.26 2.26 2.26 2.26 2.26 2.26

0.0332 0.0294 0.0246 0.0171 0.0606 0.0579 0.0513 0.0458 0.0376 0.0288 0.1024 0.0989 0.0884 0.0798 0.0717 0.0644 0.0542 0.2238 0.2203 0.2005 0.1809 0.1650 0.1468 0.1307 0.3853 0.3782 0.3599 0.3551 0.3472 0.3328 0.3169 0.3017 0.2817 0.2672

5.79 7.66 10.01 13.69 3.03 4.33 7.58 10.25 14.30 18.58 3.92 5.61 10.79 14.98 18.96 22.51 27.54 7.59 9.29 18.97 28.57 36.39 45.30 53.16 9.91 13.40 22.36 24.70 28.55 35.64 43.39 50.87 60.63 67.76

APPENDIX C

(continued )

0.28 0.25 0.21 0.14 0.51 0.48 0.43 0.38 0.31 0.24 0.85 0.83 0.74 0.67 0.60 0.54 0.45 1.87 1.84 1.67 1.51 1.38 1.22 1.09 3.21 3.15 3.00 2.96 2.90 2.78 2.64 2.52 2.35 2.23

2785_A003.fm Page 365 Friday, March 11, 2005 5:33 PM

365

Copyright 2005 by Taylor & Francis Group, LLC

8.625 8.625 10.75 10.75 10.75 10.75 10.75 10.75 10.75 10.75 10.75 10.75 10.75 10.75 12.75 12.75 12.75 12.75 12.75 12.75 12.75 12.75 12.75 12.75 12.75 12.75 12.75 14.00 14.00 14.00 14.00 14.00

Outside Dia, in.

10 20 30

60 80 100 120 140 160

40

20 30

30 40 60 80 100 120 140 160

20

160

a

Std

XS

Std

Std XS

XXS

5S 10S

80S

40S

5S 10S

40S 80S

5S 10S

Schedule b c 0.875 0.906 0.134 0.165 0.250 0.279 0.307 0.365 0.500 0.593 0.718 0.843 1.000 1.125 0.156 0.180 0.250 0.330 0.375 0.406 0.500 0.562 0.687 0.843 1.000 1.125 1.312 0.156 0.188 0.250 0.312 0.375

Wall Thickness, in. 6.875 6.813 10.482 10.420 10.250 10.192 10.136 10.020 9.750 9.564 9.314 9.064 8.750 8.500 12.438 12.390 12.250 12.090 12.000 11.938 11.750 11.626 11.376 11.064 10.750 10.500 10.126 13.688 13.624 13.500 13.376 13.250

Inside Dia, in. 37.1035 36.4373 86.2498 85.2325 82.4741 81.5433 80.6497 78.8143 74.6241 71.8040 68.0992 64.4925 60.1016 56.7163 121.4425 120.5070 117.7991 114.7420 113.0400 111.8749 108.3791 106.1036 101.5895 96.0935 90.7166 86.5463 80.4907 147.0787 145.7065 143.0663 140.4501 137.8166

Inside Area, in.2 2.26 2.26 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.67 3.67 3.67 3.67 3.67

Surface Area, ft2/ft 0.2577 0.2530 0.5990 0.5919 0.5727 0.5663 0.5601 0.5473 0.5182 0.4986 0.4729 0.4479 0.4174 0.3939 0.8434 0.8369 0.8181 0.7968 0.7850 0.7769 0.7526 0.7368 0.7055 0.6673 0.6300 0.6010 0.5590 1.0214 1.0119 0.9935 0.9754 0.9571

Volume, ft3/ft 72.42 74.69 15.19 18.65 28.04 31.20 34.24 40.48 54.74 64.33 76.93 89.20 104.13 115.64 20.98 24.16 33.38 43.77 49.56 53.52 65.42 73.15 88.51 107.20 125.49 139.67 160.27 23.07 27.73 36.71 45.61 54.57

Pipe Weight, lb/ft 2.15 2.11 5.00 4.94 4.78 4.72 4.67 4.57 4.32 4.16 3.94 3.74 3.48 3.29 7.03 6.98 6.82 6.65 6.55 6.48 6.28 6.15 5.88 5.57 5.26 5.01 4.66 8.52 8.44 8.29 8.14 7.98

Water Weight, lb/ft

366

14

12

10

Nominal Pipe Size, NPS

2785_A003.fm Page 366 Friday, March 11, 2005 5:33 PM

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

18

16

14.00 14.00 14.00 14.00 14.00 14.00 14.00 14.00 14.00 14.00 14.00 14.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 18.00 18.00 18.00 XS

40

10

100 120 140 160

80

60

Std

XS

10 20 30

100 120 140 160

80

60

40

5S 10S

5S 10S

0.437 0.500 0.562 0.593 0.625 0.687 0.750 0.875 0.937 1.093 1.250 1.406 0.165 0.188 0.250 0.312 0.375 0.437 0.500 0.562 0.625 0.656 0.687 0.750 0.843 0.875 1.031 1.218 1.437 1.593 0.165 0.188 0.250

13.126 13.000 12.876 12.814 12.750 12.626 12.500 12.250 12.126 11.814 11.500 11.188 15.670 15.624 15.500 15.376 15.250 15.126 15.000 14.876 14.750 14.688 14.626 14.500 14.314 14.250 13.938 13.564 13.126 12.814 17.670 17.624 17.500

135.2491 132.6650 130.1462 128.8959 127.6116 125.1415 122.6563 117.7991 115.4263 109.5629 103.8163 98.2595 192.7559 191.6259 188.5963 185.5908 182.5616 179.6048 176.6250 173.7169 170.7866 169.3538 167.9271 165.0463 160.8391 159.4041 152.5003 144.4259 135.2491 128.8959 245.0997 243.8252 240.4063

3.67 3.67 3.67 3.67 3.67 3.67 3.67 3.67 3.67 3.67 3.67 3.67 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.71 4.71 4.71

0.9392 0.9213 0.9038 0.8951 0.8862 0.8690 0.8518 0.8181 0.8016 0.7609 0.7209 0.6824 1.3386 1.3307 1.3097 1.2888 1.2678 1.2473 1.2266 1.2064 1.1860 1.1761 1.1662 1.1462 1.1169 1.1070 1.0590 1.0030 0.9392 0.8951 1.7021 1.6932 1.6695

63.30 72.09 80.66 84.91 89.28 97.68 106.13 122.65 130.72 150.67 170.21 189.11 27.90 31.75 42.05 52.27 62.58 72.64 82.77 92.66 102.63 107.50 112.35 122.15 136.46 141.34 164.82 192.29 223.50 245.11 31.43 35.76 47.39

APPENDIX C

(continued )

7.83 7.69 7.54 7.47 7.39 7.25 7.11 6.82 6.69 6.35 6.01 5.69 11.17 11.10 10.93 10.75 10.58 10.40 10.23 10.06 9.89 9.81 9.73 9.56 9.32 9.23 8.83 8.37 7.83 7.47 14.20 14.12 13.93

2785_A003.fm Page 367 Friday, March 11, 2005 5:33 PM

367

Copyright 2005 by Taylor & Francis Group, LLC

20

Nominal Pipe Size, NPS

18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00

Outside Dia, in.

XS

30

80 100 120

60

5S 10S

0.312 0.375 0.437 0.500 0.562 0.625 0.687 0.750 0.875 0.937 1.156 1.375 1.562 1.781 0.188 0.218 0.250 0.312 0.375 0.437 0.500 0.562 0.593 0.625 0.687 0.750 0.812 0.875 1.031 1.281 1.500

Wall Thickness, in. 17.376 17.250 17.126 17.000 16.876 16.750 16.626 16.500 16.250 16.126 15.688 15.250 14.876 14.438 19.624 19.564 19.500 19.376 19.250 19.126 19.000 18.876 18.814 18.750 18.626 18.500 18.376 18.250 17.938 17.438 17.000

Inside Dia, in. 237.0114 233.5866 230.2404 226.8650 223.5675 220.2416 216.9927 213.7163 207.2891 204.1376 193.1990 182.5616 173.7169 163.6378 302.3046 300.4588 298.4963 294.7121 290.8916 287.1560 283.3850 279.6982 277.8638 275.9766 272.3384 268.6663 265.0767 261.4541 252.5909 238.7058 226.8650

Inside Area, in.2 4.71 4.71 4.71 4.71 4.71 4.71 4.71 4.71 4.71 4.71 4.71 4.71 4.71 4.71 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24

Surface Area, ft2/ft 1.6459 1.6221 1.5989 1.5755 1.5526 1.5295 1.5069 1.4841 1.4395 1.4176 1.3417 1.2678 1.2064 1.1364 2.0993 2.0865 2.0729 2.0466 2.0201 1.9941 1.9680 1.9424 1.9296 1.9165 1.8912 1.8657 1.8408 1.8157 1.7541 1.6577 1.5755

Volume, ft3/ft 58.94 70.59 81.97 93.45 104.67 115.98 127.03 138.17 160.03 170.75 207.96 244.14 274.22 308.50 39.78 46.06 52.73 65.60 78.60 91.30 104.13 116.67 122.91 129.33 141.70 154.19 166.40 178.72 208.87 256.10 296.37

Pipe Weight, lb/ft 13.73 13.53 13.34 13.14 12.95 12.76 12.57 12.38 12.01 11.83 11.19 10.58 10.06 9.48 17.51 17.41 17.29 17.07 16.85 16.63 16.42 16.20 16.10 15.99 15.78 15.56 15.36 15.15 14.63 13.83 13.14

Water Weight, lb/ft

368

40

Std

XS

Std

Schedule b c

20

10

80 100 120 140 160

60

40

30

20

a

2785_A003.fm Page 368 Friday, March 11, 2005 5:33 PM

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

26

24

22

20.00 20.00 22.00 22.00 22.00 22.00 22.00 22.00 22.00 22.00 22.00 22.00 22.00 22.00 22.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 26.00 26.00 10

60 80 100 120 140 160

40

30

20

10

60 80 100 120 140 160

10 20 30

140 160

XS

Std

Std XS

5S 10S

5S 10S

1.750 1.968 0.188 0.218 0.250 0.375 0.500 0.625 0.750 0.875 1.125 1.375 1.625 1.875 2.125 0.188 0.218 0.250 0.312 0.375 0.437 0.500 0.562 0.593 0.625 0.812 1.031 1.281 1.500 1.750 1.968 0.250 0.312

16.500 16.064 21.624 21.564 21.500 21.250 21.000 20.750 20.500 20.250 19.750 19.250 18.750 18.250 17.750 23.624 23.564 23.500 23.376 23.250 23.126 23.000 22.876 22.814 22.750 22.376 21.938 21.438 21.000 20.500 20.064 25.500 25.376

213.7163 202.5709 367.0639 365.0298 362.8663 354.4766 346.1850 337.9916 329.8963 321.8991 306.1991 290.8916 275.9766 261.4541 247.3241 438.1033 435.8807 433.5163 428.9533 424.3416 419.8273 415.2650 410.7994 408.5757 406.2866 393.0380 377.8015 360.7765 346.1850 329.8963 316.0128 510.4463 505.4940

5.24 5.24 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.81 6.81

1.4841 1.4067 2.5491 2.5349 2.5199 2.4616 2.4041 2.3472 2.2910 2.2354 2.1264 2.0201 1.9165 1.8157 1.7175 3.0424 3.0270 3.0105 2.9789 2.9468 2.9155 2.8838 2.8528 2.8373 2.8214 2.7294 2.6236 2.5054 2.4041 2.2910 2.1945 3.5448 3.5104

341.09 379.00 43.80 50.71 58.07 86.61 114.81 142.68 170.21 197.41 250.81 302.88 353.61 403.00 451.06 47.81 55.37 63.41 78.93 94.62 109.97 125.49 140.68 148.24 156.03 201.09 252.91 310.82 360.45 415.85 463.07 68.75 85.60

APPENDIX C

(continued )

12.38 11.73 21.26 21.15 21.02 20.53 20.05 19.58 19.11 18.65 17.74 16.85 15.99 15.15 14.33 25.38 25.25 25.11 24.85 24.58 24.32 24.06 23.80 23.67 23.54 22.77 21.89 20.90 20.05 19.11 18.31 29.57 29.28

2785_A003.fm Page 369 Friday, March 11, 2005 5:33 PM

369

Copyright 2005 by Taylor & Francis Group, LLC

26.00 26.00 26.00 26.00 26.00 26.00 26.00 28.00 28.00 28.00 28.00 28.00 28.00 28.00 28.00 28.00 30.00 30.00 30.00 30.00 30.00 30.00 30.00 30.00 30.00 32.00 32.00 32.00 32.00 32.00 32.00

Outside Dia, in.

20 30 40

10

20 30 40

10

20 30

10

20

a

Std XS

Std XS

Std XS

Std XS

5S 10S

Schedule b c 0.375 0.500 0.625 0.750 0.875 1.000 1.125 0.250 0.312 0.375 0.500 0.625 0.750 0.875 1.000 1.125 0.250 0.312 0.375 0.500 0.625 0.750 0.875 1.000 1.125 0.250 0.312 0.375 0.500 0.625 0.688

Wall Thickness, in. 25.250 25.000 24.750 24.500 24.250 24.000 23.750 27.500 27.376 27.250 27.000 26.750 26.500 26.250 26.000 25.750 29.500 29.376 29.250 29.000 28.750 28.500 28.250 28.000 27.750 31.500 31.376 31.250 31.000 30.750 30.624

Inside Dia, in. 500.4866 490.6250 480.8616 471.1963 461.6291 452.1600 442.7891 593.6563 588.3146 582.9116 572.2650 561.7166 551.2663 540.9141 530.6600 520.5041 683.1463 677.4153 671.6166 660.1850 648.8516 637.6163 626.4791 615.4400 604.4991 778.9163 772.7959 766.6016 754.3850 742.2666 736.1961

Inside Area, in.2 6.81 6.81 6.81 6.81 6.81 6.81 6.81 7.33 7.33 7.33 7.33 7.33 7.33 7.33 7.33 7.33 7.85 7.85 7.85 7.85 7.85 7.85 7.85 7.85 7.85 8.38 8.38 8.38 8.38 8.38 8.38

Surface Area, ft2/ft 3.4756 3.4071 3.3393 3.2722 3.2058 3.1400 3.0749 4.1226 4.0855 4.0480 3.9741 3.9008 3.8282 3.7564 3.6851 3.6146 4.7441 4.7043 4.6640 4.5846 4.5059 4.4279 4.3506 4.2739 4.1979 5.4092 5.3667 5.3236 5.2388 5.1546 5.1125

Volume, ft3/ft 102.63 136.17 169.38 202.25 234.79 267.00 298.87 74.09 92.26 110.64 146.85 182.73 218.27 253.48 288.36 322.90 79.43 98.93 118.65 157.53 196.08 234.29 272.17 309.72 346.93 84.77 105.59 126.66 168.21 209.43 230.08

Pipe Weight, lb/ft 28.99 28.42 27.86 27.30 26.74 26.19 25.65 34.39 34.08 33.77 33.15 32.54 31.93 31.33 30.74 30.15 39.57 39.24 38.91 38.24 37.59 36.94 36.29 35.65 35.02 45.12 44.77 44.41 43.70 43.00 42.65

Water Weight, lb/ft

370

32

30

28

Nominal Pipe Size, NPS

2785_A003.fm Page 370 Friday, March 11, 2005 5:33 PM

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

42

36

34

32.00 32.00 32.00 32.00 34.00 34.00 34.00 34.00 34.00 34.00 34.00 34.00 34.00 34.00 36.00 36.00 36.00 36.00 36.00 36.00 36.00 36.00 36.00 42.00 42.00 42.00 42.00 42.00 42.00 42.00 42.00 20 30 40

20 30 40

10

20 30 40

10

Std XS

Std XS

Std XS

0.750 0.875 1.000 1.125 0.250 0.312 0.375 0.500 0.625 0.688 0.750 0.875 1.000 1.125 0.250 0.312 0.375 0.500 0.625 0.750 0.875 1.000 1.125 0.250 0.375 0.500 0.625 0.750 1.000 1.250 1.500

30.500 30.250 30.000 29.750 33.500 33.376 33.250 33.000 32.750 32.624 32.500 32.250 32.000 31.750 35.500 35.376 35.250 35.000 34.750 34.500 34.250 34.000 33.750 41.500 41.250 41.000 40.750 40.500 40.000 39.500 39.000

730.2463 718.3241 706.5000 694.7741 880.9663 874.4565 867.8666 854.8650 841.9616 835.4954 829.1563 816.4491 803.8400 791.3291 989.2963 982.3972 975.4116 961.6250 947.9366 934.3463 920.8541 907.4600 894.1641 1351.9663 1335.7266 1319.5850 1303.5416 1287.5963 1256.0000 1224.7963 1193.9850

8.38 8.38 8.38 8.38 8.90 8.90 8.90 8.90 8.90 8.90 8.90 8.90 8.90 8.90 9.42 9.42 9.42 9.42 9.42 9.42 9.42 9.42 9.42 11.00 11.00 11.00 11.00 11.00 11.00 11.00 11.00

5.0712 4.9884 4.9063 4.8248 6.1178 6.0726 6.0269 5.9366 5.8470 5.8021 5.7580 5.6698 5.5822 5.4954 6.8701 6.8222 6.7737 6.6780 6.5829 6.4885 6.3948 6.3018 6.2095 9.3887 9.2759 9.1638 9.0524 8.9417 8.7222 8.5056 8.2916

250.31 290.86 331.08 370.96 90.11 112.25 134.67 178.89 222.78 244.77 266.33 309.55 352.44 394.99 95.45 118.92 142.68 189.57 236.13 282.35 328.24 373.80 419.02 111.47 166.71 221.61 276.18 330.41 437.88 544.01 648.81

42.30 41.61 40.93 40.25 51.03 50.66 50.27 49.52 48.77 48.40 48.03 47.30 46.57 45.84 57.31 56.91 56.50 55.71 54.91 54.13 53.34 52.57 51.80 78.32 77.38 76.44 75.51 74.59 72.76 70.95 69.17

2785_A003.fm Page 371 Friday, March 11, 2005 5:33 PM

APPENDIX C 371

2785_A004.fm Page 373 Wednesday, April 13, 2005 2:36 PM

APPENDIX

D

GASMOD Output Report The following is a report from GASMOD—Gas Pipeline Hydraulics Simulation software (www.systek.us)—for a pipeline transporting natural gas in an NPS 16 pipeline 450 miles long from Compton to Harvard. ******* GASMOD - GAS PIPELINE HYDRAULIC SIMULATION ******** ************ 32-bit Version 5.00.200 ************

DATE:

21-September-2004

TIME: 09:45:06

PROJECT DESCRIPTION: Pipeline from Compton to Harvard 16" pipeline Case Number:

1264

Pipeline data file:

C:\GASMOD32\MYPIPE001.TOT

Pressure drop formula: Pipeline efficiency: Compressibility Factor Method: Inlet Gas Gravity(Air=1.0): Inlet Gas Viscosity:

AGA Turbulent 1.00 AGA NX19 0.67883 0.0000070(lb/ft-sec)

CALCULATION OPTIONS: Branch pipe calculations: Loop pipe calculations: Compressor Fuel Calculated: Joule Thompson effect included: Customized Output: Holding Delivery Pressure at terminus

NO NO YES NO NO

373

Copyright 2005 by Taylor & Francis Group, LLC

2785_A004.fm Page 374 Wednesday, April 13, 2005 2:36 PM

374

GAS PIPELINE HYDRAULICS

***************** Calculations Based on Specified Thermal Conductivities of Pipe, Soil and Insulation ************** Origin suction temperature: Base temperature: Base pressure: Origin suction pressure: Delivery pressure: Minimum pressure: Gas specific heat ratio: Maximum gas velocity:

70.00(degF) 60.00(degF) 14.700(psig) 800.00(psig) 500.33(psig) 200.00(psig) 1.26 50.00(ft/sec)

Inlet Flow rate: Outlet Flow rate:

150.0000(MMSCFD) 128.2744(MMSCFD)

**************** PIPELINE PROFILE DATA *********** Distance (mi)

Elevation (ft)

Diameter (in)

Thickness (in)

Roughness (in)

0.00 15.00 20.00 30.00 42.00 45.00 48.90 85.00 128.00 130.00 140.00 150.00 154.00 155.00 160.00 180.00 238.40 240.00 250.00 260.00 290.00 292.00 300.00 320.00 330.00 346.00 350.00 356.00 360.00 400.00 420.00 450.00

620.00 969.70 1086.26 1319.39 1599.14 1669.08 1760.00 2929.00 1260.00 1338.46 1730.77 2123.08 2280.00 2263.53 2181.18 1851.80 890.00 883.18 840.54 797.91 670.00 727.00 955.00 1525.00 1563.33 1624.67 1640.00 1543.40 1479.00 835.00 881.00 950.00

16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000

0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375

0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700

Copyright 2005 by Taylor & Francis Group, LLC

2785_A004.fm Page 375 Wednesday, April 13, 2005 2:36 PM

APPENDIX D

375

******** THERMAL CONDUCTIVITY AND INSULATION DATA ********

Distance (mi)

Cover (in)

0.000 15.000 20.000 30.000 42.000 45.000 48.900 85.000 128.000 130.000 140.000 150.000 154.000 155.000 160.000 180.000 238.400 240.000 250.000 260.000 290.000 292.000 300.000 320.000 330.000 346.000 350.000 356.000 360.000 400.000 420.000 450.000

36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000

Thermal Conductivity (Btu/hr/ft/degF) Pipe Soil Insulation 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000

****************

0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750

Insul.Thk (in)

Soil Temp (degF)

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00

0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200

COMPRESSOR STATION DATA **************

FLOW RATES, PRESSURES AND TEMPERATURES:

Name Compton Sta-2 Sta-4 Sta-5

Flow Suct. Disch. Suct. Disch. Suct. Disch. MaxPipe Rate Press. Press. Compr. Loss. Loss. Temp. Temp Temp (MMSCFD) (psig) (psig) Ratio (psig) (psig) (degF) (degF) (degF) 149.39 98.90 98.45 98.27

Copyright 2005 by Taylor & Francis Group, LLC

795.00 752.05 768.45 953.88

1210.00 1210.00 1210.00 1149.13

1.5125 1.5973 1.5638 1.2016

5.00 5.00 5.00 5.00

10.00 10.00 10.00 10.00

70.00 80.00 80.00 80.00

129.17 148.62 145.37 106.12

140.00 140.00 140.00 140.00

2785_A004.fm Page 376 Wednesday, April 13, 2005 2:36 PM

376

GAS PIPELINE HYDRAULICS

INACTIVE COMPRESSOR STATIONS:

Name

Distance (mi)

Dimpton Jackson

180 420

********* COMPRESSOR EFFICIENCY, HP AND FUEL USED *********

Name Compton Sta-2 Sta-4 Sta-5

Distance (mi)

Compr Effy. (%)

Mech. Effy. (%)

Overall Effy. (%)

0.00 85.00 250.00 330.00

80.00 80.00 80.00 80.00

98.00 98.00 98.00 98.00

78.40 78.40 78.40 78.40

Horse Power

Fuel Factor (MCF/ day/HP)

Fuel Used (MMSCFD)

3,059.55 2,419.97 2,284.08 864.48

0.2000 0.2000 0.2000 0.2000

0.6119 0.4840 0.4568 0.1729

Total Compressor Station Horsepower:

8,628.08

Total Fuel consumption:

1.7256(MMSCFD)

**************** LOCATIONS AND FLOW RATES **************** Flow Distance in/out Viscosity Pressure GasTemp. Location (mi) (MMSCFD) Gravity (lb/ft-sec) (psig) (degF) GasName Compton

0.00

150.0000 0.6788

0.00000700

1200.00

129.17

Harvard

30.00 350.00 450.00

-50.0000 0.6788 30.0000 0.6000 -128.2744 0.6606

0.00000700 0.00000800 0.00000723

986.29 1083.08 500.33

83.52 82.05 80.00

SAN JUAN GAS

****** REYNOLD'S NUMBER AND HEAT TRANSFER COEFFICIENT ***** HeatTransCoeff CompressibilityDistance FrictFactor Transmission (Btu/hr/ Factor (mi) Reynold'sNum. (Darcy) Factor ft2/degF) (AGA NX19) 0.000 15.000 20.000 30.000 42.000 45.000 48.900 85.000 128.000 130.000 140.000

12,831,042. 12,831,042. 12,831,042. 8,536,437. 8,536,437. 8,536,437. 8,536,437. 8,493,784. 8,493,784. 8,493,784. 8,493,784.

Copyright 2005 by Taylor & Francis Group, LLC

0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104

19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63

0.4999 0.4999 0.4999 0.4992 0.4992 0.4992 0.4992 0.4991 0.4991 0.4991 0.4991

0.8027 0.7887 0.7913 0.7979 0.8024 0.8052 0.8241 0.7672 0.7518 0.7574 0.7670

2785_A004.fm Page 377 Wednesday, April 13, 2005 2:36 PM

APPENDIX D

150.000 154.000 155.000 160.000 180.000 238.400 240.000 250.000 260.000 290.000 292.000 300.000 320.000 330.000 346.000 350.000 356.000 360.000 400.000 420.000 450.000

377

8,493,784. 8,493,784. 8,493,784. 8,493,784. 8,493,784. 8,493,784. 8,493,784. 8,451,888. 8,451,888. 8,451,888. 8,451,888. 8,451,888. 8,451,888. 8,435,063. 8,435,063. 10,716,727. 10,716,727. 10,716,727. 10,716,727. 10,716,727. 10,716,727.

0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104

19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63

0.4991 0.4991 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4685 0.4685 0.4685 0.4685 0.4685 0.4685

0.7738 0.7761 0.7777 0.7849 0.8094 0.8298 0.8345 0.8069 0.7629 0.7635 0.7676 0.7797 0.7923 0.7808 0.7706 0.7892 0.7925 0.8124 0.8451 0.8805 0.8805

******** PIPELINE TEMPERATURE AND PRESSURE PROFILE ******** Distance Diameter Flow Velocity Press. GasTemp. SoilTemp. (mi) (in) (MMSCFD) (ft/sec) (psig) (degF) (degF)

MAOP (psig) Location

0.00 15.00 20.00 30.00 42.00 45.00 48.90 85.00

16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000

149.3881 149.3881 149.3881 99.3881 99.3881 99.3881 99.3881 98.9041

16.53 18.08 18.68 13.34 13.97 14.15 14.38 17.22

1200.00 1095.78 1060.08 986.29 940.95 929.42 914.30 757.05

129.17 93.61 88.70 83.52 80.68 80.45 80.26 80.00

80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00

1400.00 Compton 1400.00 1400.00 1400.00 1400.00 1400.00 Branch1 1400.00 1400.00 Sta-2

85.00 128.00 130.00 140.00 150.00 154.00 155.00 160.00 180.00 238.40 240.00 250.00

16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000

98.9041 98.9041 98.9041 98.9041 98.9041 98.9041 98.9041 98.9041 98.9041 98.9041 98.9041 98.4473

10.94 11.42 11.50 11.91 12.36 12.55 12.58 12.72 13.34 16.01 16.12 16.78

1200.00 1149.26 1141.23 1100.95 1060.36 1044.01 1041.70 1030.05 981.36 815.32 809.68 773.45

140.00 80.23 80.17 80.04 80.01 80.01 80.01 80.00 80.00 80.00 80.00 80.00

80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00

1400.00 Sta-2 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 Dimpton 1400.00 1400.00 1400.00 Sta-4

250.00 260.00 290.00 292.00 300.00 320.00 330.00

16.000 16.000 16.000 16.000 16.000 16.000 16.000

98.4473 98.4473 98.4473 98.4473 98.4473 98.4473 98.2744

10.89 11.12 11.85 11.93 12.25 13.17 13.56

1200.00 1175.10 1101.69 1094.39 1065.01 990.04 958.88

140.00 98.11 80.39 80.30 80.11 80.01 80.00

80.00 80.00 80.00 80.00 80.00 80.00 80.00

1400.00 Sta-4 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 Sta-5

330.00 346.00

16.000 16.000

98.2744 98.2744

11.44 11.91

1139.13 1094.37

106.12 83.45

80.00 80.00

1400.00 Sta-5 1400.00

Copyright 2005 by Taylor & Francis Group, LLC

2785_A004.fm Page 378 Wednesday, April 13, 2005 2:36 PM

378

350.00 356.00 360.00 400.00 420.00 450.00

GAS PIPELINE HYDRAULICS

16.000 16.000 16.000 16.000 16.000 16.000

128.2744 128.2744 128.2744 128.2744 128.2744 128.2744

15.70 16.05 16.30 19.78 23.01 33.47

1083.08 1058.92 1042.45 856.54 734.40 500.33

82.05 81.11 80.74 80.01 80.00 80.00

80.00 80.00 80.00 80.00 80.00 80.00

1400.00 1400.00 1400.00 1400.00 1400.00 Jackson 1400.00 Harvard

************ LINE PACK VOLUMES AND PRESSURES ************** Distance (mi)

Pressure (psig)

Line Pack (million std.cu.ft)

0.00 15.00 20.00 30.00 42.00 45.00 48.90 85.00 128.00 130.00 140.00 150.00 154.00 155.00 160.00 180.00 238.40 240.00 250.00 260.00 290.00 292.00 300.00 320.00 330.00 346.00 350.00 356.00 360.00 400.00 420.00 450.00

1200.00 1095.78 1060.08 986.29 940.95 929.42 914.30 757.05 1149.26 1141.23 1100.95 1060.36 1044.01 1041.70 1030.05 981.36 815.32 809.68 773.45 1175.10 1101.69 1094.39 1065.01 990.04 958.88 1094.37 1083.08 1058.92 1042.45 856.54 734.40 500.33

0.0000 9.6000 3.0086 5.7524 6.4677 1.5581 1.9909 16.4027 23.2986 1.3544 6.5846 6.2702 2.4211 0.5982 2.9652 11.4175 28.9974 0.7002 4.2420 5.1998 20.1395 1.2790 5.0078 11.7426 5.4816 9.2580 2.5092 3.6990 2.4088 21.2972 8.5521 9.6571

Total line pack in main pipeline = 239.8614(million std.cu.ft) Started simulation at: 09:44:17 Finished simulation at: 09:45:06 Time elapsed: 49 seconds DATE: 21-September-2004

Copyright 2005 by Taylor & Francis Group, LLC

2785_A005.fm Page 379 Friday, April 1, 2005 3:06 PM

APPENDIX

E

Summary of Formulas CHAPTER 1 1. Density

ρ=

m V

(1.1)

where r = density of gas m = mass of gas V = volume of gas 2. Gas gravity G=

ρg ρair

(1.2)

where G = gas gravity, dimensionless rg = density of gas rair = density of air G=

Mg 29

where Mg = molecular weight of gas Mair = molecular weight of air = 28.9625

379

Copyright 2005 by Taylor & Francis Group, LLC

(1.4)

2785_A005.fm Page 380 Friday, April 1, 2005 3:06 PM

380

GAS PIPELINE HYDRAULICS

3. Kinematic viscosity

ν=

µ ρ

(1.5)

where, in USCS units, n = kinematic viscosity, ft2/s m = dynamic viscosity, lb/ft-s r = density, lb/ft3 and, in SI units, n = kinematic viscosity, cSt m = dynamic viscosity, cP r = density, kg/m3 4. Viscosity of mixture

µ=

where m = mi = yi = Mi =

(

Σ µi yi Mi

(

Σ yi Mi

)

)

(1.6)

dynamic viscosity of gas mixture dynamic viscosity of gas component i mole fraction or percent of gas component i molecular weight of gas component i

5. Ideal gas law or perfect gas equation PV = nRT where P = V = n = R = T =

(1.8)

absolute pressure, pounds per square inch absolute (psia) gas volume, ft3 number of lb moles as deﬁned in Equation 1.7 universal gas constant, psia ft3/lb mole °R absolute temperature of gas, °R (°F + 460)

6. Absolute pressure Pabs = Pgauge + Patm

(1.10)

P1 V2 = or P1V1 = P2V2 P2 V1

(1.13)

7. Boyle’s law

Copyright 2005 by Taylor & Francis Group, LLC

2785_A005.fm Page 381 Friday, April 1, 2005 3:06 PM

APPENDIX E

381

8. Charles’s law V1 T1 = V2 T2

at constant pressure

(1.14)

P1 T1 = P2 T2

at constant volume

(1.15)

9. Modiﬁed ideal gas equation PV = ZnRT (USCS units) where P = V = Z = T = n = R =

(1.16)

absolute pressure of gas, psia volume of gas, ft3 gas compressibility factor, dimensionless absolute temperature of gas, °R number of lb moles as deﬁned in Equation 1.7 universal gas constant, 10.73 psia ft3/lb mole °R

10. Reduced temperature and reduced pressure Tr =

T Tc

(1.17)

Pr =

P Pc

(1.18)

11. Pseudo-reduced temperature and pseudo-reduced pressure

where P T Tpr Ppr Tpc Ppc

= = = = = =

Tpr =

T Tpc

(1.19)

Ppr =

P Ppc

(1.20)

absolute pressure of gas mixture, psia absolute temperature of gas mixture, °R pseudo-reduced temperature, dimensionless pseudo-reduced pressure, dimensionless pseudo-critical temperature, °R pseudo-critical pressure, psia

12. Apparent molecular weight of gas mixture M a = Σyi Mi

Copyright 2005 by Taylor & Francis Group, LLC

(1.21)

2785_A005.fm Page 382 Friday, April 1, 2005 3:06 PM

382

GAS PIPELINE HYDRAULICS

where Ma = apparent molecular weight of gas mixture yi = mole fraction of gas component i Mi = molecular weight of gas component i 13. Kay’s rule to calculate the average pseudo-critical properties of the gas mixture Tpc = ΣyiTc

(1.22)

Ppc = Σyi Pc

(1.23)

14. Pseudo-critical properties from gas gravity Tpc = 170.491 + 307.344 G

(1.24)

Ppc = 709.604 − 58.718 G

(1.25)

where G = gas gravity (air = 1.00) Tpc = pseudo-critical temperature, °R Ppc = pseudo-critical pressure, psia 15. Supercompressibility factor Z=

1 ( Fpv )2

(1.30)

16. Dranchuk, Purvis, and Robinson method A A ρ5 A7 ρr3 A A A Z = 1 + A + 2 + 33 ρr + A4 + 5 ρr2 + 5 6 r + 2 Tpr Tpr Tpr Tpr Tpr3 1 + A8 ρr2 e( − A8ρr )

(

)

(1.31) where

ρr = and A1 A3 A5 A7 Ppr Tpr

= = = = = =

0.27 Ppr ZTpr

0.31506237; A2 = −1.04670990; −0.57832729; A4 = 0.53530771; −0.61232032; A6 = −0.10488813; 0.68157001; A8 = 0.68446549; pseudo-reduced pressure pseudo-reduced temperature

Copyright 2005 by Taylor & Francis Group, LLC

(1.32)

2785_A005.fm Page 383 Friday, April 1, 2005 3:06 PM

APPENDIX E

383

17. CNGA method Z=

1 Pavg 344 ,400 (10 )1.785G 1 + T f3.825

(1.34)

for the average gas pressure Pavg > 100 psig. For Pavg < 100 psig, Z = 1.00 where Pavg = average gas pressure, psig Tf = average gas temperature, °R G = gas gravity (air = 1.00) 18. Average pressure in a pipe segment Pavg =

2 P13 − P23 3 P12 − P2 2

(1.36)

19. Heating value H m = Σ( yi Hi )

(1.37)

where Hm = gross heating value of mixture, Btu/ft3 yi = mole fraction or percent of gas component i Hi = heating value of gas component, Btu/ft3

CHAPTER 2 1. General Flow equation using friction factor T P 2 − P22 Q = 77.54 b 1 Pb GT f LZf where Q = f = Pb = Tb = P1 = P2 = G = Tf =

0.5

D 2.5

(USCS units)

gas ﬂow rate, measured at standard conditions, ft3/day (SCFD) friction factor, dimensionless base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂowing temperature, °R (460 + °F)

Copyright 2005 by Taylor & Francis Group, LLC

(2.2)

2785_A005.fm Page 384 Friday, April 1, 2005 3:06 PM

384

GAS PIPELINE HYDRAULICS

L = pipe segment length, mi Z = gas compressibility factor at the ﬂowing temperature, dimensionless D = pipe inside diameter, in.

(

)

2 2 T P1 − P2 Q = 1.1494 × 10 −3 b Pb GT f LZf

0.5

D 2.5

(SI units)

(2.3)

where Q = gas ﬂow rate, measured at standard conditions, m3/day f = friction factor, dimensionless Pb = base pressure, kPa Tb = base temperature, K (273 + °C) P1 = upstream pressure, kPa P2 = downstream pressure, kPa G = gas gravity (air = 1.00) Tf = average gas ﬂowing temperature, K (273 + °C) L = pipe segment length, km Z = gas compressibility factor at the ﬂowing temperature, dimensionless D = pipe inside diameter, mm 2. General Flow equation using transmission factor T P 2 − e s P22 Q = 38.77 F b 1 Pb GT f Le Z

Q = 5.747 × 10 −4

0.5

(

D 2.5

2 s 2 T P1 − e P2 F b Pb GT f Le Z

)

(USCS units)

(2.7)

D 2.5

(2.8)

0.5

(SI units)

where the elevation correction is as follows: Le =

L (e s − 1) s

H −H 1 s = 0.0375 G 2 Tf Z where s = H1 = H2 = e =

(2.9)

(USCS units)

elevation adjustment parameter, dimensionless upstream elevation, ft downstream elevation, ft base of natural logarithms

Copyright 2005 by Taylor & Francis Group, LLC

(2.10)

2785_A005.fm Page 385 Friday, April 1, 2005 3:06 PM

APPENDIX E

385

and H −H 1 s = 0.0684 G 2 Tf Z

(SI units)

(2.11)

where H1 = upstream elevation, m H2 = downstream elevation, m 3. The equivalent length Le = j1L1 + j2L2es1 + j3L3es2 + …

(2.13)

Q P ZT u = 0.002122 b2 b D Tb P

(2.28)

4. The gas velocity

where u = gas velocity, ft/s Qb = gas ﬂow rate, measured at standard conditions, ft3/day (SCFD) D = pipe inside diameter, in. Pb = base pressure, psia Tb = base temperature, °R (460 + °F) P = upstream pressure, psia T = upstream gas temperature, °R (460 + °F) Z = gas compressibility factor at upstream conditions, dimensionless In SI units, the gas velocity at any point in a gas pipeline is given by Q P ZT u = 14.7349 b2 b D Tb P where u Qb D Pb Tb P T Z

(SI units)

(2.29)

= gas velocity, m/s = gas ﬂow rate, measured at standard conditions, m3/day = pipe inside diameter, mm = base pressure, kPa = base temperature, K (273 + °C) = pressure, kPa = average gas ﬂowing temperature, K (273 + °C) = gas compressibility factor at the ﬂowing temperature, dimensionless

5. Maximum velocity umax = 100

Copyright 2005 by Taylor & Francis Group, LLC

ZRT 29GP

(USCS units)

(2.31)

2785_A005.fm Page 386 Friday, April 1, 2005 3:06 PM

386

GAS PIPELINE HYDRAULICS

where Z = R = T = G = P =

compressibility factor of gas, dimensionless gas constant = 10.73 ft3 psia/lb-moleR gas temperature, °R gas gravity (air = 1.00) gas pressure, psia

6. Reynolds number P GQ Re = 0.0004778 b Tb µ D

(USCS units)

(2.34)

(SI units)

(2.35)

where Pb = base pressure, psia Tb = base temperature, °R (460 + °F) G = speciﬁc gravity of gas (air = 1.0) Q = gas ﬂow rate, standard ft3/day (SCFD) D = pipe inside diameter, in. m = viscosity of gas, lb/ft-s In SI units, the Reynolds number is P GQ Re = 0.5134 b Tb µ D

where Pb = base pressure, kPa Tb = base temperature, °K (273 + °C) G = speciﬁc gravity of gas (air = 1.0) Q = gas ﬂow rate, m3/day (standard conditions) D = pipe inside diameter, mm m = viscosity of gas, Poise 7. Colebrook-White equation e 2.51 = − 2Log10 + f 3.7 D Re f

1

where f D e Re

= = = =

friction factor, dimensionless pipe inside diameter, in. absolute pipe roughness, in. Reynolds number of ﬂow, dimensionless

Copyright 2005 by Taylor & Francis Group, LLC

for Re > 4000

(2.39)

2785_A005.fm Page 387 Friday, April 1, 2005 3:06 PM

APPENDIX E

387

2.51 = −2Log10 f Re f

1

e = −2Log10 3.7 D f

1

for turbulent ﬂow in smooth pipes

(2.40)

for turbulent ﬂow in fully rough pipes

(2.41)

The transmission factor F is related to the friction factor f as follows: F=

2

f=

4 F2

(2.42)

f

Therefore, (2.43)

where f = friction factor F = transmission factor 8. Colebrook equation in terms of transmission factor F e 1.255F F = −4 Log10 + Re 3.7 D

(2.45)

9. Modiﬁed Colebrook-White equation for turbulent ﬂow using friction factor e 2.825 = −2Log10 + f 3.7 D Re f

1

(2.46)

Modiﬁed Colebrook-White equation in terms of the transmission factor e 1.4125F F = −4 Log10 + Re 3.7 D

(USCS and SI units)

(2.47)

10. AGA equation 3.7 D F = 4 Log10 e

(2.48)

11. Bend index BI =

Copyright 2005 by Taylor & Francis Group, LLC

total degrees of all bends in pipe sectionn total length of pipe section

(2.51)

2785_A005.fm Page 388 Friday, April 1, 2005 3:06 PM

388

GAS PIPELINE HYDRAULICS

12. Weymouth equation T P 2 − e s P22 Q = 433.5E b 1 Pb GT f Le Z where Q = E = Pb = Tb = P1 = P2 = G = Tf = Le = Z = D =

0.5

D 2.667

(in USCS units)

(2.52)

volume ﬂow rate, standard ft3/day (SCFD) pipeline efﬁciency, a decimal value less than or equal to 1.0 base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂow temperature, °R (460 + °F) equivalent length of pipe segment, mi gas compressibility factor, dimensionless pipe inside diameter, in.

Weymouth transmission factor F = 11.18(D)1/6

(in USCS units)

(2.53)

Weymouth equation T P 2 − e s P22 Q = 3.7435 × 10 E b 1 Pb GT f Le Z

0.5

−3

D 2.667

(in SI units)

(2.54)

where Q = gas ﬂow rate, standard m3/day Tb = base temperature, K (273 + °C) Pb = base pressure, kPa Tf = average gas ﬂow temperature, K (273 + °C) P1 = upstream pressure, kPa P2 = downstream pressure, kPa Le = equivalent length of pipe segment, km Other symbols are as deﬁned previously. Weymouth transmission factor F = 6.521(D)1/6

(in SI units)

(2.53a)

13. Panhandle A equation T Q = 435.87 E b Pb

1.0788

P 2 − es P 2 1 2 0.8539 T f Le Z G

T Q = 4.5965 × 10 E b Pb −3

Copyright 2005 by Taylor & Francis Group, LLC

1.0788

0.5394

D 2.6182

P 2 − es P 2 1 2 0.8539 T f Le Z G

(USCS units)

(2.55)

0.5394

D 2.6182

(SI units)

(2.56)

2785_A005.fm Page 389 Friday, April 1, 2005 3:06 PM

APPENDIX E

389

The equivalent transmission factor for Panhandle A equation QG F = 7.2111E D

0.07305

(USCS)

(2.57)

(SI)

(2.58)

and in SI units, it is QG F = 11.85E D

0.07305

13. Panhandle B equation T Q = 737 E b Pb

1.02

P 2 − es P 2 1 2 0.961 G T L f eZ

T Q = 1.002 × 10 E b Pb −2

1.02

0.51

D 2.53

P 2 − es P 2 1 2 0.961 G T f Le Z

(USCS units)

(2.59)

0.51

D 2.53 (SI units)

(2.60)

The equivalent transmission factor for Panhandle B equation is QG F = 16.7 E D

0.01961

QG F = 19.08 E D

(USCS units)

(2.61)

0.01961

(SI units)

(2.62)

14. IGT equation T P 2 − e s P22 Q = 136.9 E b 01.8 0.2 Pb G T f Le µ

0.555

D 2.667

T P 2 − e s P22 Q = 1.2822 × 10 E b 01.8 0.2 Pb G T f Le µ

(USCS units)

(2.63)

D 2.667

(2.64)

0.555

−3

(SI units)

15. Spitzglass equation The low pressure (less than or equal to 1 psig) version T P1 − P2 Q = 3.839 × 10 E b 3 . 6 Pb GT f Le Z 1 + D + 0.03D 3

(

)

0.5

D 2.5

(USCS units) (2.65)

Copyright 2005 by Taylor & Francis Group, LLC

2785_A005.fm Page 390 Friday, April 1, 2005 3:06 PM

390

GAS PIPELINE HYDRAULICS

The high pressure (more than 1 psig) version T P12 − e s P22 Q = 729.6087 E b Pb GT f Le Z 1 + 3D.6 + 0.03D

(

0.5

D 2.5

)

(USCS units) (2.67)

The low pressure (less than 6.9 kPa) version T P1 − P2 Q = 5.69 × 10 E b 91 . 44 Pb GT f Le Z 1 + D + 0.0012 D −2

(

0.5

D 2.5

)

(SI units) (2.66)

The high pressure (more than 6.9 kPa) version T P12 − e s P22 Q = 1.0815 × 10 E b Pb GT f Le Z 1 + 91D.44 + 0.0012 D −2

(

0.5

)

D 2.5

(SI units) (2.68)

16. The Mueller equation T P12 − e s P22 Q = 85.7368 E b 0.7391 0.2609 T f Le µ Pb G

0.575

D 2.725

T P12 − e s P22 Q = 3.0398 × 10 E b 0.7391 T f Le µ 0.2609 Pb G

(USCS units)

(2.69)

0.575

−2

D 2.725

(SI units) (2.70)

17. Fritzsche formula T P2 − P2 Q = 410.1688 E b 01.8587 2 T f Le Pb G T P 2 − es P 2 Q = 2.827 E b 10.8587 2 T f Le Pb G

0.538

D 2.69

(USCS units)

(2.71)

0.538

D 2.69

(SI units)

(2.72)

CHAPTER 3 1. Total equivalent length—series piping 5

D D Le = L1 + L2 1 + L3 1 D 2 D3

Copyright 2005 by Taylor & Francis Group, LLC

5

(3.6)

2785_A005.fm Page 391 Friday, April 1, 2005 3:06 PM

APPENDIX E

391

2. Equivalent diameter—parallel pipes 1/ 5

1 + Const1 2 De = D1 Const1

(3.17)

where 5

D L Const1 = 1 2 D2 L1

(3.18)

Flow rates Q1 and Q2 are calculated from Q1 =

QConst1 1 + Const1

(3.19)

Q2 =

Q 1 + Const1

(3.20)

and

3. Temperature proﬁle of gas in a pipe segment T2 = Ts + (T1 − Ts )e −θ

θ= where U ∆L ∆A T1 T2 Ts D m

(3.29)

πUD∆ L mCp

(3.28)

= overall heat transfer coefﬁcient, Btu/h/ft2/°F = length of pipe segment = surface area of pipe for heat transfer = p D∆L = gas temperature upstream of pipe segment, °F = gas temperature downstream of pipe segment, °F = average soil temperature surrounding pipe segment, °F = pipe inside diameter, ft = mass ﬂow rate of gas, lb/s

4. Line pack T P Vb = 28.798 b avg ( D 2 L ) Pb ZavgTavg

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(3.34)

2785_A005.fm Page 392 Friday, April 1, 2005 3:06 PM

392

GAS PIPELINE HYDRAULICS

where Vb = line pack in pipe segment, standard ft3 D = pipe inside diameter, in. L = pipe segment length, mi T P Vb = 7.855 × 10 −4 b avg ( D 2 L) Pb ZavgTavg

(SI units)

(3.35)

where Vb = line pack in pipe segment, standard m3 D = pipe inside diameter, mm L = pipe segment length, km CHAPTER 4 1. Compression ratio r=

Pd Ps

(4.1)

2. Isothermal work done Wi = where Wi G T1 P1 P2 Loge

= = = = = =

P 53.28 T Loge 2 G 1 P1

(4.4)

isothermal work done, ft-lb/lb of gas gas gravity, dimensionless suction temperature of gas, °R suction pressure of gas, psia discharge pressure of gas, psia natural logarithm to base e(e = 2.718)

The ratio

( ) is also called the compression ratio. P2 P1

Wi = where Wi = T1 = P1 = P2 =

(USCS units)

P 286.76 T1Loge 2 G P1

isothermal work done, J/kg of gas suction temperature of gas, K suction pressure of gas, kPa absolute discharge pressure of gas, kPa absolute

Other symbols are as deﬁned earlier.

Copyright 2005 by Taylor & Francis Group, LLC

(SI units)

(4.5)

2785_A005.fm Page 393 Friday, April 1, 2005 3:06 PM

APPENDIX E

393

3. Adiabatic work done γ −1 53.28 γ P2 γ − 1 Wa = T G 1 γ − 1 P1

where Wa G T1 g P1 P2

= = = = = =

(USCS units)

adiabatic work done, ft-lb/lb of gas gas gravity, dimensionless suction temperature of gas, °R ratio of speciﬁc heats of gas, dimensionless suction pressure of gas, psia discharge pressure of gas, psia γ −1 286.76 γ P2 γ Wa = T1 − 1 G γ − 1 P1

where Wa T1 P1 P2

= = = =

(4.8)

(SI units)

(4.9)

adiabatic work done, J/kg of gas suction temperature of gas, K suction pressure of gas, kPa absolute discharge pressure of gas, kPa absolute

Other symbols are as deﬁned earlier. 4. Horsepower γ −1 Z1 + Z 2 1 P2 γ γ − 1 HP = 0.0857 QT γ − 1 1 2 ηa P1

where HP g Q T1 P1 P2 Z1 Z2 ha

= compression horsepower = ratio of speciﬁc heats of gas, dimensionless = gas ﬂow rate, MMSCFD = suction temperature of gas, °R = suction pressure of gas, psia = discharge pressure of gas, psia = compressibility of gas at suction conditions, dimensionless = compressibility of gas at discharge conditions, dimensionless = compressor adiabatic (isentropic) efﬁciency, decimal value

Copyright 2005 by Taylor & Francis Group, LLC

(4.15)

2785_A005.fm Page 394 Friday, April 1, 2005 3:06 PM

394

GAS PIPELINE HYDRAULICS

In SI units, the Power equation is as follows: γ −1 Z1 + Z 2 1 P2 γ γ − Power = 4.0639 QT 1 γ − 1 1 2 ηa P1

where Power g Q T1 P1 P2 Z1 Z2 ha

= = = = = = = = =

(4.16)

compression power, kW ratio of speciﬁc heats of gas, dimensionless gas ﬂow rate, Mm3/day (standard) suction temperature of gas, K suction pressure of gas, kPa discharge pressure of gas, kPa compressibility of gas at suction conditions, dimensionless compressibility of gas at discharge conditions, dimensionless compressor adiabatic (isentropic) efﬁciency, decimal value BHP =

HP ηm

(4.17)

5. Compression ratio 1

r = (rt ) n

(4.25)

where r = compression ratio, dimensionless rt = overall compression ratio, dimensionless n = number of compressors in series CHAPTER 6 1. Barlow’s equation Sh = where Sh = P = D = t =

PD 2t

(6.1)

hoop or circumferential stress in pipe material, psi internal pressure, psi pipe outside diameter, in. pipe wall thickness, in.

Axial or longitudinal stress Sa =

Copyright 2005 by Taylor & Francis Group, LLC

PD 4t

(6.2)

2785_A005.fm Page 395 Friday, April 1, 2005 3:06 PM

APPENDIX E

395

2. Internal design pressure P=

2tSEFT D

(6.8)

where P = internal pipe design pressure, psig D = pipe outside diameter, in. t = pipe wall thickness, in. S = speciﬁed minimum yield strength (SMYS) of pipe material, psig E = seam joint factor, 1.0 for seamless and submerged arc welded (SAW) pipes F = design factor, usually 0.72 for cross-country gas pipelines, but can be as low as 0.4, depending on class location and type of construction T = temperature deration factor = 1.00 for temperatures below 250°F 3. Blowdown calculations 1

1

0.0588 P1 3G 2 D 2 LFc T= d2 where T = P1 = G = D = L = d = Fc =

(USCS units)

(6.9)

(SI units)

(6.10)

blowdown time, min initial pressure, psia gas gravity (air = 1.00) pipe inside diameter, in. length of pipe section, mi inside diameter of blowdown pipe, in. choke factor (as follows)

Choke factor list Ideal nozzle = 1.0 Through gate = 1.6 Regular gate = 1.8 Regular lube plug = 2.0 Venturi lube plug = 3.2

In SI units, 1

T=

where P1 = initial pressure, kPa D = pipe inside diameter, mm

Copyright 2005 by Taylor & Francis Group, LLC

1

0.0192 P1 3G 2 D 2 LFc d2

2785_A005.fm Page 396 Friday, April 1, 2005 3:06 PM

396

GAS PIPELINE HYDRAULICS

L = length of pipe section, km d = pipe inside diameter of blowdown, mm Other symbols are as deﬁned before. 4. Pipe weight w = 10.68 × t × (D − t)

(USCS units)

(6.11)

where w = pipe weight, lb/ft D = pipe outside diameter, in. t = pipe wall thickness, in. w = 0.0246 × t × (D − t)

(SI units)

(6.12)

2[( p1 − p2 )/ρ + g( z1 − z2 )] 1 − Cc2 ( Ao / A)2

(9.2)

where w = pipe weight, kg/m D = pipe outside diameter, mm t = pipe wall thickness, mm

CHAPTER 9 1. The discharge through the oriﬁce meter Q = CcCv Ao where Q = Cc = Cv = Ao = A = p1 = p2 = r = z1 = z2 = g =

ﬂow rate, ft3/s contraction coefﬁcient, dimensionless discharge coefﬁcient, dimensionless cross-sectional area of the oriﬁce, in.2 cross-sectional area of pipe containing the oriﬁce, in.2 upstream pressure, psig downstream pressure, psig density of gas, lb/ft3 upstream elevation, ft downstream elevation, ft acceleration due to gravity

2. Sharp-crested oriﬁce 5

A 2 Cc = 0.595 + 0.29 o A

Copyright 2005 by Taylor & Francis Group, LLC

(9.4)

2785_A005.fm Page 397 Friday, April 1, 2005 3:06 PM

APPENDIX E

397

3. The fundamental oriﬁce meter ﬂow equation described in the ANSI 2530/AGA Report No. 3 is as follows: qm =

π C Y d 2 (2 gρ f ∆ P)0.5 4 0.5 4 (1 − β )

(9.6)

or qm = KY

π 2 d (2 gρ f ∆ P)0.5 4 d D

(9.8)

C CD 2 = (1 − β 4 )0.5 ( D 4 − d 4 )0.5

(9.9)

β= K= where qm rf C b d D Y g ∆P K

= = = = = = = = = =

(9.7)

mass ﬂow rate of gas, lb/s density of gas, lb/ft3 discharge coefﬁcient beta ratio, dimensionless oriﬁce diameter, in. meter tube diameter, in. expansion factor, dimensionless acceleration due to gravity, ft/s2 pressure drop across the oriﬁce, psi ﬂow coefﬁcient, dimensionless

4. Buckingham and Bean equation endorsed by the National Bureau of Standards (NBS) and listed in AGA Report No.3 For ﬂange taps: 1 0.5 0.007 0.076 + 0.364 + 0.5 β 4 + 0.4 1.6 − 0.07 + − β D D D D 5

Ke = 0.5993 +

65 0.034 − 0.009 + (0.5 − β )1.5 + 2 + 3 (β − 0.7)2.5 D D where Ke = D = d = b =

ﬂow coefﬁcient for Reynolds number Rd = d(106/15), dimensionless meter tube diameter, in. oriﬁce diameter, in. beta ratio, dimensionless

Copyright 2005 by Taylor & Francis Group, LLC

2.5

(9.12)

2785_A005.fm Page 398 Friday, April 1, 2005 3:06 PM

398

GAS PIPELINE HYDRAULICS

For pipe taps: Ke = 0.5925 +

0.0182 0.06 2 0.225 5 + 0.440 − β + 0.935 + β D D D

+ 1.35β 14 +

1.43 (0.25 − β )2.5 D 0.5

(9.13)

where all symbols are as deﬁned before. 5. Expansion factor For ﬂange taps: Y1 = 1 − (0.41 + 0.35β 4 )

x1 k

(9.20)

For pipe taps: Y1 = 1 − [0.333 + 1.145(β 2 + 0.7β 5 + 12β 13 )]

x1 k

(9.21)

and the pressure ratio x1 is x1 =

Pf 1 − Pf 2 Pf 1

=

hw 27.707 Pf 1

(9.22)

where Y1 = expansion factor based on upstream pressure x1 = ratio of differential pressure to absolute upstream static pressure hw = differential pressure between upstream and downstream taps in in. of water at 60°F Pf1 = static pressure at upstream tap, psia Pf2 = static pressure at downstream tap, psia x1/k = acoustic ratio, dimensionless k = ratio of speciﬁc heats of gas, dimensionless 6. Supercompressibility factor C = Fb Fr Fpb Ftb Ftf Fgr FpvY where the dimensionless factors are Fb = basic oriﬁce factor Fr = Reynolds number factor Fpb = pressure base factor Ftb = temperature base factor Ftf = ﬂowing temperature factor Fgr = gas relative density factor Fpv = supercompressibility factor Y = expansion factor

Copyright 2005 by Taylor & Francis Group, LLC

(9.34)

2785_A005.fm Page 399 Friday, April 1, 2005 3:06 PM

APPENDIX E

399

CHAPTER 10 1. Present value PV = where PV R i n

= = = =

1 R 1− i (1 + i)n

(10.1)

present value, $ series of cash ﬂows, $ interest rate, decimal value number of periods, years

2. Pipe material cost PMC = where PMC L D T C

= = = = =

10.68( D − T )TLC × 5280 2000

(10.2)

pipe material cost, $ length of pipe, mi pipe outside diameter, in. pipe wall thickness, in. pipe material cost, $/ton

In SI units, PMC = 0.0246( D − T )TLC where PMC L D T C

= = = = =

pipe material cost, $ length of pipe, km pipe outside diameter, mm pipe wall thickness, mm pipe material cost, $/metric ton

Copyright 2005 by Taylor & Francis Group, LLC

(10.3)

Boca Raton London New York Singapore

A CRC title, part of the Taylor & Francis imprint, a member of the Taylor & Francis Group, the academic division of T&F Informa plc.

Copyright 2005 by Taylor & Francis Group, LLC

2785_Discl Page 1 Friday, March 11, 2005 10:28 AM

Published in 2005 by CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2005 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number-10: 0-8493-2785-7 (Hardcover) International Standard Book Number-13: 978-0-8493-2785-8 (Hardcover) Library of Congress Card Number 2004062949 This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data Menon, E. Shashi. Gas pipeline hydraulics / E. Shashi Menon. p. cm. Includes bibliographical references and index. ISBN 0-8493-2785-7 1. Natural gas pipeline--Design and construction. 2. Pipe--Hydrodynamics. I. Title. TN880.5.M455 2005 665.7'44--dc22

2004062949

Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com Taylor & Francis Group is the Academic Division of T&F Informa plc.

Copyright 2005 by Taylor & Francis Group, LLC

and the CRC Press Web site at http://www.crcpress.com

2785_C000.fm Page i Wednesday, April 13, 2005 10:38 AM

Preface Gas Pipeline Hydraulics is a practical handbook for engineers, technicians, and others involved in the design and operation of pipelines transporting natural gas and other compressible fluids. It is based on the author’s 30-year experience in the oil and gas industry. This book will help readers better understand and apply the principles of fluid mechanics to their daily work in the gas pipeline transmission and distribution industry. The book is divided into 10 chapters with several example problems solved fully, as well as additional problems provided as exercises. Chapter 1 introduces the basic properties of natural gas and other compressible fluids that are important in understanding how gas behaves under various conditions of pressure and temperature as it flows through a pipeline. The properties of hydrocarbon gas mixtures, such as gravity, viscosity, and compressibility, are reviewed, and both analytical and graphical methods are explained with illustrative examples. In Chapter 2, the methods of calculating the pressure drop in a gas pipeline are discussed. The General Flow equation is introduced as the basic equation, and the various correlations for friction factor and transmission factors, such as Colebrook and AGA, are explained. Other flow equations, such as Panhandle and Weymouth, are also covered using examples. Chapter 3 extends the concepts of pressure drop calculations further to determine the total pressure required for transporting gas in pipelines under various configurations, such as series and parallel pipelines. The effects of intermediate delivery volumes and injection rates along a distribution pipeline are reviewed. The importance of the contract delivery pressures and the necessity of regulating pressures using a control valve or pressure regulator are also discussed. The effect of gas temperature on the pressure drop in a transmission pipeline is reviewed with example output reports from a commercial gas hydraulics simulation model. Equivalent lengths in series piping and equivalent diameters in parallel piping are covered, as well as pipe looping to increase gas pipeline flow rate. The quantity of gas contained in a section of a pipeline and the calculation of line pack are also reviewed. Chapter 4 discusses compressor stations required to transport gas in a pipeline and how to calculate their numbers and optimum locations. Centrifugal and positive displacement compressors are explained, and their performances are compared. Typical performance characteristics of a centrifugal compressor are analyzed. Isothermal, adiabatic, and polytropic compression processes and horsepower required are discussed with sample calculations. The discharge temperature of the compressed gas and its impact on pipeline throughput, along with the necessity of gas cooling, are explained. In Chapter 5, installing pipe loops to increase the throughput in a gas pipeline is explored. Looping is compared to the option of building intermediate compressor stations. The advantages and disadvantages of looping a pipeline versus installing compressor stations are discussed. Chapter 6 covers the mechanical strength of a pipeline. The effects of pipe diameter, wall thickness, material of construction, and specific safety requirements

Copyright 2005 by Taylor & Francis Group, LLC

2785_C000.fm Page ii Wednesday, April 13, 2005 10:38 AM

dictated by design codes and state and federal regulations are reviewed. Hydrostatic testing requirements and classification of pipelines based upon their proximity to human dwellings and industrial establishments and population density are also covered. Chapter 7 introduces readers to thermal hydraulic analysis. For simplicity, long distance gas pipelines can be treated as isothermal flows. With compressor stations, the higher discharge temperature causes heat transfer between the pipeline gas and its surroundings. The effects of soil thermal conductivity, burial depth of pipeline, and the soil temperature are analyzed in determining the temperature and pressure profile of a gas pipeline. The Joule-Thompson cooling effect is reviewed, and the results from a commercial hydraulic simulation model are discussed. Chapter 8 introduces transient pressure analysis. This is an area that is quite complex, and a complete discussion of the transient hydraulic analysis of gas pipelines requires a separate book. Readers are referred to some excellent references on this subject. Chapter 9 covers valves and flow measurement. The different types of valves used in a gas pipeline and their characteristics are explained. The importance of accurate flow measurement in gas pipeline transactions is stressed. The codes and standards used to ensure proper design, construction, and operation of orifice flow meters are reviewed. Chapter 10 explores economic aspects of gas pipeline systems. Determining the optimum pipe size for a particular gas flow rate, taking into account the initial capital cost and annual operating and maintenance cost, is explained. For a typical gas pipeline system, the various capital cost components are reviewed, along with the recurring annual costs such as operation and maintenance, fuel, and administrative costs. Also, the calculation methodology for determining transportation cost or tariff is covered. At the end of each chapter, additional problems are provided as exercises. A list of references for the material covered in each chapter is included as well. Appendices at the end of the book include tables of conversion factors from USCS units to SI units and vice versa, along with tables of properties of natural gas. Also included are tables showing commonly used pipe sizes, listing allowable internal pressures and hydrotest pressures. A section containing a summary of all hydraulic formulas used in the book is provided as a handy reference. I enjoyed working with the fine staff at CRC Press. In particular, I want to thank Cindy Carelli, acquisitions editor; Theresa Del Forn, project coordinator; and Marsha Hecht, project editor, for their meticulous and prompt attention to all matters concerning the production of this book. They are indeed a very professional group and one of the best I have worked with over the years. I am indebted to my wife Pramila for the enormous amount of typing she did preparing the manuscript and for helping me proofread the final document and check for mathematical accuracy. I also appreciate all the good comments and suggestions that I received from practicing professional engineers such as Ken Zipp, Dan Bhavsar, and Charles Tateosian. In addition, I am thankful to Charles Peterson, Ken Zipp, and Ron Wood for agreeing to review

Copyright 2005 by Taylor & Francis Group, LLC

2785_C000.fm Page iii Wednesday, April 13, 2005 10:38 AM

the manuscript, which resulted in some valuable comments that enhanced the quality of this book. Finally, I would like to dedicate this book to my parents, who encouraged me in all my endeavors throughout my childhood and adult life. E. Shashi Menon

Copyright 2005 by Taylor & Francis Group, LLC

2785_C000.fm Page v Wednesday, April 13, 2005 10:38 AM

Author E. Shashi Menon is an engineering and computer consultant at SYSTEK Technologies, Inc., Lake Havasu City, Arizona. He is a registered professional engineer in California and the author of numerous engineering software programs and professional publications. He is an associate fellow of the Royal Aeronautical Society, London, and a member of the American Society of Mechanical Engineers. Menon earned a bachelor’s degree in mechanical engineering from Bhopal University, India, and an M.S. in mechanical engineering from California State University, Long Beach. He has worked for more than 30 years in manufacturing and the oil and gas industry. He has held the positions of design engineer, project engineer, project manager, chief engineer, and engineering manager for gas pipeline and liquid pipeline companies. Since 1998, Menon has taught pipeline hydraulics courses to engineers in North and South America and has published numerous hydraulics and pump software programs currently used by engineers in the oil and gas industry. He has authored books for Marcel Dekker and McGraw-Hill.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C000.fm Page vii Wednesday, April 13, 2005 10:38 AM

Contents Chapter 1 Gas Properties 1.1 Mass and Weight..............................................................................................1 1.2 Volume..............................................................................................................2 1.3 Density, Specific Weight, and Specific Volume ..............................................3 1.4 Specific Gravity................................................................................................3 1.5 Viscosity ...........................................................................................................4 1.6 Ideal Gases .......................................................................................................9 1.7 Real Gases......................................................................................................14 1.8 Natural Gas Mixtures.....................................................................................16 1.9 Pseudo-Critical Properties from Gas Gravity................................................19 1.10 Impact of Sour Gas and Non-Hydrocarbon Components.............................20 1.11 Compressibility Factor ...................................................................................21 1.11.1 Standing-Katz Method .....................................................................22 1.11.2 Dranchuk, Purvis, and Robinson Method .......................................23 1.11.3 American Gas Association (AGA) Method.....................................23 1.11.4 California Natural Gas Association (CNGA) Method ....................24 1.12 Heating Value .................................................................................................27 1.13 Summary ........................................................................................................27 Problems ..................................................................................................................28 References................................................................................................................29 Chapter 2 Pressure Drop Due to Friction 2.1 Bernoulli’s Equation ......................................................................................31 2.2 Flow Equations...............................................................................................32 2.3 General Flow Equation ..................................................................................33 2.4 Effect of Pipe Elevations ...............................................................................35 2.5 Average Pipe Segment Pressure ....................................................................37 2.6 Velocity of Gas in a Pipeline.........................................................................37 2.7 Erosional Velocity ..........................................................................................40 2.8 Reynolds Number of Flow.............................................................................43 2.9 Friction Factor................................................................................................45 2.10 Colebrook-White Equation ............................................................................47 2.11 Transmission Factor .......................................................................................50 2.12 Modified Colebrook-White Equation ............................................................54 2.13 American Gas Association (AGA) Equation.................................................57 2.14 Weymouth Equation.......................................................................................61 2.15 Panhandle A Equation....................................................................................64 2.16 Panhandle B Equation....................................................................................68 2.17 Institute of Gas Technology (IGT) Equation ................................................70 2.18 Spitzglass Equation ........................................................................................74 2.19 Mueller Equation............................................................................................76 2.20 Fritzsche Equation..........................................................................................77 2.21 Effect of Pipe Roughness ..............................................................................78

Copyright 2005 by Taylor & Francis Group, LLC

2785_C000.fm Page viii Wednesday, April 13, 2005 10:38 AM

2.22 Comparison of Flow Equations .....................................................................80 2.23 Summary ........................................................................................................81 Problems ..................................................................................................................82 References................................................................................................................83 Chapter 3 Pressure Required to Transport 3.1 Total Pressure Drop Required........................................................................85 3.2 Frictional Effect .............................................................................................86 3.3 Effect of Pipeline Elevation ...........................................................................86 3.4 Effect of Changing Pipe Delivery Pressure...................................................90 3.5 Pipeline with Intermediate Injections and Deliveries ...................................93 3.6 Series Piping ................................................................................................104 3.7 Parallel Piping ..............................................................................................111 3.8 Locating Pipe Loop......................................................................................121 3.9 Hydraulic Pressure Gradient........................................................................123 3.10 Pressure Regulators and Relief Valves ........................................................126 3.11 Temperature Variation and Gas Pipeline Modeling ....................................129 3.12 Line Pack......................................................................................................132 3.13 Summary ......................................................................................................135 Problems ................................................................................................................136 References..............................................................................................................137 Chapter 4 Compressor Stations 4.1 Compressor Station Locations .....................................................................139 4.2 Hydraulic Balance........................................................................................146 4.3 Isothermal Compression ..............................................................................146 4.4 Adiabatic Compression ................................................................................148 4.5 Polytropic Compression...............................................................................151 4.6 Discharge Temperature of Compressed Gas ...............................................152 4.7 Horsepower Required...................................................................................153 4.8 Optimum Compressor Locations .................................................................157 4.9 Compressors in Series and Parallel .............................................................163 4.10 Types of Compressors—Centrifugal and Positive Displacement ..............166 4.11 Compressor Performance Curves ................................................................168 4.12 Compressor Station Piping Losses ..............................................................171 4.13 Compressor Station Schematic ....................................................................172 4.14 Summary ......................................................................................................173 Problems ................................................................................................................174 References..............................................................................................................175 Chapter 5 Pipe Loops versus Compression 5.1 Purpose of a Pipe Loop ...............................................................................177 5.2 Purpose of Compression ..............................................................................178 5.3 Increasing Pipeline Capacity .......................................................................179 5.4 Reducing Power Requirements....................................................................189

Copyright 2005 by Taylor & Francis Group, LLC

2785_C000.fm Page ix Wednesday, April 13, 2005 10:38 AM

5.5 Looping in Distribution Piping....................................................................192 5.6 Summary ......................................................................................................198 Problems ................................................................................................................198 References..............................................................................................................199 Chapter 6 Pipe Analysis 6.1 Pipe Wall Thickness.....................................................................................201 6.2 Barlow’s Equation........................................................................................202 6.3 Thick-Walled Pipes ......................................................................................203 6.4 Derivation of Barlow’s Equation .................................................................205 6.5 Pipe Material and Grade ..............................................................................207 6.6 Internal Design Pressure Equation ..............................................................207 6.7 Class Location..............................................................................................209 6.8 Mainline Valves............................................................................................210 6.9 Hydrostatic Test Pressure ............................................................................211 6.10 Blowdown Calculations ...............................................................................242 6.11 Determining Pipe Tonnage ..........................................................................243 6.12 Summary ......................................................................................................246 Problems ................................................................................................................246 References..............................................................................................................247 Chapter 7 Thermal Hydraulics 7.1 Isothermal versus Thermal Hydraulics........................................................249 7.2 Temperature Variation and Gas Pipeline Modeling ....................................251 7.3 Review of Simulation Model Reports .........................................................253 7.4 Summary ......................................................................................................273 Problems ................................................................................................................274 References..............................................................................................................274 Chapter 8 Transient Analysis and Case Studies 8.1 Unsteady Flow .............................................................................................275 8.1.1 Transient Due to Mainline Valve Closure .......................................276 8.1.2 Transient Due to Compressor Shutdown.........................................277 8.2 Case Studies .................................................................................................279 8.2.1 Offshore Pipeline Case ....................................................................279 8.3 Summary ......................................................................................................296 Problems ................................................................................................................296 References..............................................................................................................297 Chapter 9 Valves and Flow Measurements 9.1 Purpose of Valves.........................................................................................299 9.2 Types of Valves ............................................................................................300 9.3 Material of Construction..............................................................................302 9.4 Codes for Design and Construction ............................................................302 9.5 Gate Valve ....................................................................................................303

Copyright 2005 by Taylor & Francis Group, LLC

9.6 9.7 9.8 9.9 9.10 9.11 9.12 9.13 9.14 9.15

Ball Valve .....................................................................................................305 Plug Valve ....................................................................................................305 Butterfly Valve .............................................................................................305 Globe Valve ..................................................................................................306 Check Valve..................................................................................................307 Pressure Control Valve.................................................................................308 Pressure Regulator .......................................................................................309 Pressure Relief Valve ...................................................................................309 Flow Measurement.......................................................................................310 Flow Meters .................................................................................................310 9.15.1 Orifice Meter ..................................................................................310 9.15.1.1 Meter Tube ....................................................................313 9.15.1.2 Expansion Factor ..........................................................314 9.16 Venturi Meter ...............................................................................................321 9.17 Flow Nozzle .................................................................................................323 9.18 Summary ......................................................................................................325 Problems ................................................................................................................325 References..............................................................................................................325 Chapter 10 Pipeline Economics 10.1 Components of Cost.....................................................................................328 10.2 Capital Costs ................................................................................................330 10.2.1 Pipeline...........................................................................................330 10.2.2 Compressor Stations.......................................................................332 10.2.3 Mainline Valve Stations .................................................................333 10.2.4 Meter Stations and Regulators.......................................................333 10.2.5 SCADA and Telecommunication System......................................333 10.2.6 Environmental and Permitting .......................................................334 10.2.7 Right of Way Acquisitions.............................................................334 10.2.8 Engineering and Construction Management .................................335 10.2.9 Other Project Costs ........................................................................335 10.3 Operating Costs............................................................................................336 10.4 Determining Economic Pipe Size................................................................339 10.5 Summary ......................................................................................................353 Problems ................................................................................................................353 References..............................................................................................................354 Appendix A Appendix B Appendix C Appendix D Appendix E

Units and Conversions ..................................................................355 Physical Properties of Various Gases ...........................................359 Pipe Properties—U.S. Customary System of Units.....................363 GASMOD Output Report .............................................................373 Summary of Formulas...................................................................379

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 1 Friday, April 1, 2005 3:46 PM

CHAPTER

1

Gas Properties In this chapter we will discuss the properties of gases that influence gas flow through a pipeline. We will explore the relationship of pressure, volume, and temperature of a gas and how the gas properties such as density, viscosity, and compressibility change with the temperature and pressure. Starting with the ideal or perfect gases that obey the ideal gas equation, we will examine how real gases differ from ideal gases. The concept of compressibility factor, or gas deviation factor, will be introduced and methods of calculating the compressibility factor using some popular graphical correlation and calculation methods explained. The properties of a mixture of gases will be discussed, and how these are calculated will be covered. Understanding the gas properties is an important first step toward analysis of gas pipeline hydraulics. A fluid can be a liquid or a gas. Liquids are generally considered almost incompressible. A gas is classified as a homogenous fluid with low density and viscosity. It expands to fill the vessel that contains the gas. The molecules that constitute the gas are spaced farther apart in comparison with a liquid and, therefore, a slight change in pressure affects the density of gas more than that of a liquid. Gases, therefore, have higher compressibility than liquids. This implies that gas properties such as density, viscosity, and compressibility factor change with pressure.

1.1 MASS AND WEIGHT Mass is the quantity of matter in a substance. It is sometimes used interchangeably with weight. Strictly speaking, mass is a scalar quantity, whereas weight is a force and, therefore, a vector quantity. Mass is independent of the geographic location, whereas weight depends upon the acceleration due to gravity and, therefore, varies slightly with geographic location. Mass is measured in slugs in the U.S. Customary System (USCS) of units and kilograms (kg) in Systeme International (SI) units.

1

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 2 Friday, April 1, 2005 3:46 PM

2

GAS PIPELINE HYDRAULICS

However, for most purposes, we say that a 10-lb mass has a weight of 10 lb. The pound (lb) is a more convenient unit for mass, and to distinguish between mass and weight, the terms pound mass (lbm) and pound force (lbf) are sometimes used. A slug is equal to approximately 32.2 lb. If some gas is contained in a certain volume and the temperature and pressure change, the mass will remain constant unless some gas is taken out or added to the container. This is known as the principle of conservation of mass. Weight is measured in pounds (lb) in USCS units and in Newton (N) in SI units. Sometimes we talk about mass flow rate through a pipeline rather than volume flow rate. Mass flow rate is measured in lb/hr in USCS units or kg/hr in SI units.

1.2 VOLUME Volume of a gas is the space a given mass of gas occupies at a particular temperature and pressure. Since gas is compressible, it will expand to fill the available space. Therefore, the gas volume will vary with temperature and pressure. Hence, a certain volume of a given mass of gas at some temperature and pressure will decrease in volume as the pressure is increased and vice versa. Suppose a quantity of gas is contained in a volume of 100 ft3 at a temperature of 80°F and a pressure of 200 psi. If the temperature is increased to 100°F, keeping the volume constant, the pressure will also increase. Similarly, if the temperature is reduced, gas pressure will also reduce, provided volume remains constant. Charles’s law states that for constant volume, the pressure of a fixed mass of gas will vary directly with the temperature. Thus, if temperature increases by 20%, the pressure will also rise by 20%. Similarly, if pressure is maintained constant, the volume will increase in direct proportion with temperature. Charles’s law, Boyle’s law, and other gas laws will be discussed in detail later in this chapter. Volume of gas is measured in ft3 in USCS units and m3 in SI units. Other units for volume include thousand ft3 (Mft3) and million ft3 (MMft3) in USCS units and thousand m3 (km3) and million m3 (Mm3) in SI units. When referred to standard conditions (also called base conditions) of temperature and pressure (60°F and 14.7 psia in USCS units), the volume is stated as standard volume and, therefore, measured in standard ft3 (SCF) or million standard ft3 (MMSCF). It must be noted that in the USCS units, the practice has been to use M to represent a thousand, and therefore MM refers to a million. This goes back to the Roman days of numerals, when M represented a thousand. In SI units, a more logical step is followed. For thousand, the letter k (for kilo) is used and the letter M (for Mega) is used for a million. Therefore, 500 MSCFD in USCS units refers to 500 thousand standard cubic feet per day (500,000 ft3/day), whereas 15 Mm3/day means 15 million cubic meters per day in SI units. This distinction in the use of the letter M to denote a thousand in USCS units and M for a million in SI units must be clearly understood. Volume flow rate of gas is measured per unit time and can be expressed as ft3/min, ft3/h, ft3/day, SCFD, MMSCFD, etc. in USCS units. In SI units, gas flow rate is expressed in m3/h or Mm3/day.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 3 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

3

1.3 DENSITY, SPECIFIC WEIGHT, AND SPECIFIC VOLUME Density represents the amount of gas that can be packed in a given volume. Therefore, it is measured in terms of mass per unit volume. If 5 lb of a gas is contained in 100 ft3 of volume at some temperature and pressure, we say that the gas density is 5/100 = 0.05 lb/ft3. Strictly speaking, in USCS units density must be expressed as slug/ft3 since mass is customarily referred to in slug. Thus,

ρ=

m V

(1.1)

where r = density of gas m = mass of gas V = volume of gas Density is expressed in slug/ft3 or lb/ft3 in USCS units and kg/m3 in SI units. A companion term called specific weight is also used when referring to the density of gas. Specific weight, represented by the symbol g, is the weight of gas per unit volume measured in lb/ft3 in USCS units, and is, therefore, contrasted with density, which is measured in slug/ft3. In SI units, the specific weight is expressed in Newton per m3 (N/m3). Quite often, density is also referred to in lb/ft3 in USCS units. The reciprocal of the specific weight is known as the specific volume. By definition, therefore, specific volume represents the volume occupied by a unit weight of gas. It is measured in ft3/lb in USCS units and m3/N in SI units. If the specific weight of a particular gas is 0.06 lb/ft3 at some temperature and pressure, its specific volume is 0.106 or 16.67 ft3/lb.

1.4 SPECIFIC GRAVITY Specific gravity of a gas, sometimes called gravity, is a measure of how heavy the gas is compared to air at a particular temperature. It might also be called relative density, expressed as the ratio of the gas density to the density of air. Because specific gravity is a ratio, it is a dimensionless quantity. G= where G = gas gravity, dimensionless rg = density of gas rair = density of air

Copyright 2005 by Taylor & Francis Group, LLC

ρg ρair

(1.2)

2785_C001.fm Page 4 Friday, April 1, 2005 3:46 PM

4

GAS PIPELINE HYDRAULICS

Both densities in Equation 1.2 must be in the same units and measured at the same temperature. For example, natural gas has a specific gravity of 0.60 (air = 1.00) at 60°F. This means that the gas is 60% as heavy as air. If we know the molecular weight of a particular gas, we can calculate its gravity by dividing the molecular weight by the molecular weight of air, as follows. G=

Mg Mg = M air 28.9625

(1.3)

Mg 29

(1.4)

or G=

Rounding off the molecular weight of air to 29 where G = specific gravity of gas Mg = molecular weight of gas Mair = molecular weight of air = 28.9625 Since natural gas consists of a mixture of several gases (methane, ethane, etc.), the molecular weight Mg in Equation 1.4 is referred to as the apparent molecular weight of the gas mixture. When the molecular weight and the percentage or mole fractions of the individual components of a natural gas mixture are known, we can calculate the molecular weight of the gas mixture by using a weighted average method. Thus, a natural gas mixture consisting of 90% methane, 8% ethane, and 2% propane will have a specific gravity of G=

(0.9 × M1) + (0.08 × M 2) + (0.02 × M 3) 29

where M1, M2, and M3 are the molecular weights of methane, ethane, and propane, respectively, and 29 represents the molecular weight of air. Table 1.1 lists the molecular weights and other properties of several hydrocarbon gases.

1.5 VISCOSITY The viscosity of a fluid represents its resistance to flow. The higher the viscosity, the more difficult it is to flow. Lower-viscosity fluids flow easily in pipes and cause less pressure drop. Liquids have much larger values of viscosity compared to gases. For example, water has a viscosity of 1.0 centiPoise (cP), whereas viscosity of natural gas is approximately 0.0008 cP. Even though the gas viscosity is a small number, it

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 5 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

5

has an important function in determining the type of flow in pipelines. The Reynolds number (discussed in Chapter 2) is a dimensionless parameter that is used to classify flow rate in pipelines. It depends on the gas viscosity, flow rate, pipe diameter, temperature, and pressure. The absolute viscosity, also called the dynamic viscosity, is expressed in lb/ft-s in USCS units and Poise (P) in SI units. A related term is the kinematic viscosity. This is simply the absolute viscosity divided by the density. The two viscosities are related as follows:

ν=

µ ρ

(1.5)

where, in USCS units, n = kinematic viscosity, ft2/s m = dynamic viscosity, lb/ft-s r = density, lb/ft3 and in SI units n = kinematic viscosity, St m = dynamic viscosity, P r = density, kg/m3 Kinematic viscosity is expressed in ft2/s in USCS units and Stokes (St) in SI units. Other units of viscosity in SI units include centipoise (cP) for dynamic viscosity and centistokes (cSt) for kinematic viscosity. Appendix A includes conversion factors for converting viscosity from one set of units to another. The viscosity of a gas depends on its temperature and pressure. Unlike liquids, the viscosity of a gas increases with increase in temperature. Since viscosity represents resistance to flow, as the gas temperature increases, the quantity of gas flow through a pipeline will decrease; hence, more throughput is possible in a gas pipeline at lower temperatures. This is in sharp contrast to liquid flow, where the throughput increases with temperature due to decrease in viscosity and vice versa. It must be noted that, unlike liquids, pressure also affects the viscosity of a gas. Like temperature, the gas viscosity increases with pressure. Figure 1.1 shows the variation of viscosity with temperature for a gas. Table 1.2 lists the viscosities of common gases. Since natural gas is a mixture of pure gases such as methane and ethane, the following formula is used to calculate the viscosity from the viscosities of component gases:

µ=

where m mi yi Mi

= = = =

(

Σ µi yi Mi

(

Σ yi Mi

)

)

dynamic viscosity of gas mixture dynamic viscosity of gas component i mole fraction or percent of gas component i molecular weight of gas component i

Copyright 2005 by Taylor & Francis Group, LLC

(1.6)

CH4 C 2H 6 C 3H 8 C4H10 C4H10 C5H12 C5H12 C5H12 C6H14 C6H14 C6H14 C6H14 C6H14 C7H16 C7H16 C7H16 C7H16 C7H16 C7H16 C7H16 C7H16 C8H18 C8H18 C8H18 C9H20 C10H22 C5H10 C6H12 C6H12

16.0430 30.0700 44.0970 58.1230 58.1230 72.1500 72.1500 72.1500 86.1770 86.1770 86.1770 86.1770 86.1770 100.2040 100.2040 100.2040 100.2040 100.2040 100.2040 100.2040 100.2040 114.2310 114.2310 114.2310 128.2580 142.2850 70.1340 84.1610 84.1610

5000 800 188.65 72.581 51.706 20.443 15.575 36.72 4.9596 6.769 6.103 9.859 7.406 1.621 2.273 2.13 2.012 3.494 3.294 2.775 3.376 0.5371 1.1020 1.7090 0.17155 0.06088 9.917 4.491 3.267

Methane Ethane Propane Isobutane n-butane Iso-pentane n-pentane Neo-pentane n-hexane 2-methyl pentane 3-methyl pentane Neo hexane 2,3-dimethylbutane n-Heptane 2-Methylhexane 3-Methylhexane 3-Ethylpentane 2,2-Dimethylpentane 2,4-Dimethylpentane 3,3-Dimethylpentane Triptane n-octane Di Isobutyl Isooctane n-Nonane n-Decane Cyclopentane Methylcyclopentane Cyclohexane

Copyright 2005 by Taylor & Francis Group, LLC

Critical Constants Pressure Temperature Volume °F ft3/lb psia 666.0 707.0 617.0 527.9 548.8 490.4 488.1 464.0 436.9 436.6 452.5 446.7 454.0 396.8 396.0 407.6 419.2 401.8 397.4 427.9 427.9 360.7 361.1 372.7 330.7 304.6 653.8 548.8 590.7

−116.66 90.07 205.93 274.4 305.52 368.96 385.7 321.01 453.8 435.76 448.2 419.92 440.08 512.8 494.44 503.62 513.16 476.98 475.72 505.6 496.24 564.15 530.26 519.28 610.72 652.1 461.1 499.28 536.6

0.0988 0.0783 0.0727 0.0714 0.0703 0.0684 0.0695 0.0673 0.0688 0.0682 0.0682 0.0667 0.0665 0.0682 0.0673 0.0646 0.0665 0.0665 0.0667 0.0662 0.0636 0.0673 0.0676 0.0657 0.0693 0.0702 0.0594 0.0607 0.0586

Ideal Gas 14.696 psia, 60°F Spgr (air=1.00) ft3/lb-gas 0.5539 1.0382 1.5226 2.0068 2.0068 2.4912 2.4912 2.4912 2.9755 2.9755 2.9755 2.9755 2.9755 3.4598 3.4598 3.4598 3.4598 3.4598 3.4598 3.4598 3.4598 3.9441 3.9441 3.9441 4.4284 4.9127 2.4215 2.9059 2.9059

23.654 12.620 8.6059 6.5291 6.5291 5.2596 5.2596 5.2596 4.4035 4.4035 4.4035 4.4035 4.4035 3.7872 3.7872 3.7872 3.7872 3.7872 3.7872 3.7872 3.7872 3.322 3.322 3.322 2.9588 2.6671 5.411 4.509 4.509

Specific Heat, Btu/lb/°F 14.696 psia, 60°F Ideal Gas 0.52676 0.40789 0.38847 0.38669 0.39500 0.38448 0.38831 0.39038 0.38631 0.38526 0.37902 0.38231 0.37762 0.38449 0.38170 0.37882 0.38646 0.38651 0.39627 0.38306 0.37724 0.38334 0.37571 0.38222 0.38248 0.38181 0.27122 0.30027 0.29012

GAS PIPELINE HYDRAULICS

Formula

Molecular Weight

Vapor Pressure psia at 100°F

2785_C001.fm Page 6 Friday, April 1, 2005 3:46 PM

Gas

6

Table 1.1 Properties of Hydrocarbon Gases

1.609 1400 232.8 62.55 45.97 49.88 64.95 19.12 36.53 59.46 16.68 3.225 1.033 0.3716 0.2643 0.3265 0.3424 0.2582 0.188 4.631 2.313

394.59 85.46 211.9

157.3 0.95 906.71

503.4 731.0 676.6 586.4 615.4 574.9 580.2 509.5 656.0 620.3 582.0 890.4 710.4 595.5 523 541.6 512.9 509.2 587.8 465.4 1174 891.7 506.8 1071 1306 1143 1647 546.9 187.5 731.4 493 1157 3200.1 32.99 1205

570.2 48.54 198.31 296.18 324.31 311.8 292.49 376.86 354 306 403 95.29 552.15 605.5 651.22 674.85 650.95 649.47 703 676.2 463.01 465.31 −220.51 87.73 212.4 315.7 270.2 −221.29 −400.3 −181.4 −232.48 290.69 705.1 −450.31 124.75

0.0600 0.0746 0.0717 0.0683 0.0667 0.0679 0.0681 0.0674 0.0700 0.0653 0.0660 0.0693 0.0531 0.0549 0.0564 0.0557 0.0567 0.0572 0.0534 0.0569 0.0590 0.0581 0.0527 0.0342 0.0461 0.0305 0.0681 0.0517 0.5101 0.0367 0.0510 0.0280 0.04975 0.2300 0.0356

3.3902 0.9686 1.4529 1.9373 1.9373 1.9373 1.9373 2.4215 1.8677 1.8677 2.3520 0.8990 2.6971 3.1814 3.6657 3.6657 3.6657 3.6657 3.5961 4.1500 1.1063 1.5906 0.9671 1.5196 1.1768 2.2120 0.5880 1.0000 0.06960 1.1048 0.9672 2.4482 0.62202 0.1382 1.2589

3.8649 13.527 9.0179 6.7636 6.7636 6.7636 6.7636 5.411 7.0156 7.0156 5.571 14.574 4.8581 4.1184 3.5744 3.5744 3.5744 3.5744 3.6435 3.1573 11.843 8.2372 13.548 8.6229 11.134 5.9235 22.283 13.103 188.25 11.859 13.546 5.3519 21.065 94.814 10.408

0.31902 0.35789 0.35683 0.35535 0.33275 0.35574 0.36636 0.35944 0.34347 0.34223 0.35072 0.39754 0.24295 0.26005 0.27768 0.28964 0.27427 0.27470 0.26682 0.30704 0.32429 0.33074 0.24847 0.19909 0.23838 0.14802 0.49678 0.2398 3.4066 0.21897 0.24833 0.11375 0.44469 1.24040 0.19086

2785_C001.fm Page 7 Friday, April 1, 2005 3:46 PM

98.1880 28.0540 42.0810 56.1080 56.1080 56.1080 56.1080 70.1340 54.0920 54.0920 68.1190 26.0380 78.1140 92.1410 106.1670 106.1670 106.1670 106.1670 104.1520 120.1940 32.0420 46.0690 28.0100 44.0100 34.0820 64.0650 17.0305 28.9625 2.0159 31.9988 28.0134 70.9054 18.0153 4.0026 36.4606

7

Copyright 2005 by Taylor & Francis Group, LLC

C7H14 C 2H 4 C 3H 6 C 4H 8 C 4H 8 C 4H 8 C 4H 8 C5H10 C 4H 6 C 4H 6 C 5H 8 C 2H 2 C 6H 6 C 7H 8 C8H10 C8H10 C8H10 C8H10 C 8H 8 C9H12 CH4O C 2H 6O CO CO2 H 2S SO2 NH3 N2+O2 H2 O2 N2 Cl2 H 2O He HCl

GAS PROPERTIES

Methylcyclohexane Ethylene Propylene Butylene Cis-2-butene Trans-2-butene Isobutene 1-Pentene 1,2-Butadene 1,3-Butadene Isoprene Acetylene Benzene Toluene Ethyl-benzene o-Xylene m-Xylene p-Xylene Styrene Isopropylbenzene Methyl alcohol Ethyl alcohol Carbon monoxide Carbon dioxide Hydrogen sulfide Sulfur dioxide Ammonia Air Hydrogen Oxygen Nitrogen Chlorine Water Helium Hydrogen chloride

2785_C001.fm Page 8 Friday, April 1, 2005 3:46 PM

8

GAS PIPELINE HYDRAULICS

0.024

m

Heliu Air

0.022

Viscosity, cP

0.020

e

ioxid

on d

Carb

0.018 0.016

ne Methane Ethyle e Ethan Propannee i-Butatane n-Bu ntane n-Pe

0.014 0.012 0.010 0.008 0.006 0.004 50

100

150

200

250

300

350

Temperature,°F Figure 1.1

Variation of gas viscosity with temperature.

Therefore, a homogeneous mixture that consists of 20% of a gas A (molecular weight = 18) that has a viscosity 6 × 10−6 Poise and 80% of a gas B (molecular weight = 17) that has a viscosity 8 × 10−6 Poise will have a resultant viscosity of

µ=

(0.2 × 6 × 18 ) + (0.8 × 8 × 17 ) (0.2 × 18 ) + (0.8 × 17 )

× 10 −6 = 7.59 × 10−6 Poise

Table 1.2 Viscosities of Common Gases

Copyright 2005 by Taylor & Francis Group, LLC

Gas

Viscosity (cP)

Methane Ethane Propane i-Butane n-Butane i-Pentane n-Pentane Hexane Heptane Octane Nonane Decane Ethylene Carbon Monoxide Carbon Dioxide Hydrogen Sulphide Air Nitrogen Helium

0.0107 0.0089 0.0075 0.0071 0.0073 0.0066 0.0066 0.0063 0.0059 0.0050 0.0048 0.0045 0.0098 0.0184 0.0147 0.0122 0.0178 0.0173 0.0193

2785_C001.fm Page 9 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

9

It must be noted that all viscosities must be measured at the same temperature and pressure. The reader is referred to W. McCain’s book for further discussion on calculation of viscosities of natural gas mixtures. See the Reference section for more details. Example 1 A natural gas mixture consists of four components C1, C2, C3, and nC4. Their mole fractions and viscosities at a particular temperature and pressure are indicated below, along with their molecular weights. Component

Mole Fraction y

Viscosity, cP

Molecular Weight

C1 C2 C3 nC4

0.8200 0.1000 0.0500 0.0300

0.0130 0.0112 0.0098 0.0091

16.04 30.07 44.10 58.12

Total

1.000

Calculate the viscosity of the gas mixture. Solution Using the given data, we prepare a table as follows. M represents the molecular weight of each component and m the viscosity. Component

y

M

M 1/2

yM 1/2

yM 1/2

C1 C2 C3 nC4

0.8200 0.1000 0.0500 0.0300

16.04 30.07 44.10 58.12

4.00 5.48 6.64 7.62

3.2841 0.5484 0.3320 0.2287

0.0130 0.0112 0.0098 0.0091

0.0427 0.0061 0.0033 0.0021

Total

1.000

4.3932

0.0542

From Equation 1.6, the viscosity of the gas mixture is

µ=

0.0542 = 0.0123 cP 4.3932

1.6 IDEAL GASES An ideal gas is defined as a fluid in which the volume of the gas molecules is negligible when compared to the volume occupied by the gas. Also, the attraction or repulsion between the individual gas molecules and the container is negligible. Further, in an ideal gas, the molecules are considered to be perfectly elastic, and there is no internal energy loss resulting from collision between the molecules. Such ideal gases are said

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 10 Friday, April 1, 2005 3:46 PM

10

GAS PIPELINE HYDRAULICS

to obey several gas laws, such as Boyle’s law, Charles’s law, and the ideal gas law or the perfect gas equation. We will first discuss the behavior of ideal gases and then follow it up with the behavior of real gases. If M represents the molecular weight of a gas and the mass of a certain quantity of gas is m, the number of moles is given by n=

m M

(1.7)

where n is the number that represents the number of moles in the given mass. For example, the molecular weight of methane is 16.043. Therefore, 50 lb of methane will contain approximately 3 moles. The ideal gas law, sometimes referred to as the perfect gas equation, simply states that the pressure, volume, and temperature of the gas are related to the number of moles by the following equation: PV = nRT where P = V = n = R = T =

(USCS units)

(1.8)

absolute pressure, pounds per square inch absolute (psia) gas volume, ft3 number of lb moles as defined in Equation 1.7 universal gas constant, psia ft3/lb mole °R absolute temperature of gas, °R (°F + 460)

The universal gas constant R has a value of 10.73 psia ft3/lb mole °R in USCS units. We can combine Equation 1.7 with Equation 1.8 and express the ideal gas equation as follows: PV =

mRT M

(1.9)

All symbols are as defined previously. It should be noted that the constant R is the same for all ideal gases and, hence, it is called the universal gas constant. It has been found that the ideal gas equation is correct only at low pressures close to the atmospheric pressure. Since gas pipelines generally operate at pressures higher than atmospheric pressures, we must modify Equation 1.9 to take into account the effect of compressibility. The latter is accounted for by using a term called the compressibility factor, or gas deviation factor. We will discuss real gases and the compressibility factor later in this chapter. In the ideal gas Equation 1.9, the pressures and temperatures must be in absolute units. Absolute pressure is defined as the gauge pressure (as measured by a gauge) plus the local atmospheric pressure. Therefore, Pabs = Pgauge + Patm

Copyright 2005 by Taylor & Francis Group, LLC

(1.10)

2785_C001.fm Page 11 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

11

Thus, if the gas pressure is 200 psig and the atmospheric pressure is 14.7 psia, we get the absolute pressure of the gas as Pabs = 200 + 14.7 = 214.7 psia Absolute pressure is expressed as psia, whereas the gauge pressure is expressed as psig. The adder to the gauge pressure, which is the local atmospheric pressure, is also called the base pressure. In SI units, 500 kPa gauge pressure is equal to 601 kPa absolute pressure if the base pressure is 101 kPa. Pressure in USCS units is expressed in pounds per square inch, or psi. In SI units, pressure is expressed in kilopascal (kPa), megapascal (MPa), or Bar. Refer to Appendix A for unit conversion charts. The absolute temperature is measured above a certain datum. In USCS units, the absolute scale of temperature is designated as degree Rankin (°R) and is equal to the sum of the temperature in °F and the constant 460. In SI units the absolute temperature scale is referred to as degree Kelvin (K). Absolute temperature in K is equal to °C + 273. Therefore, °R = °F + 460

(1.11)

K = °C + 273

(1.12)

It is customary to drop the degree symbol for absolute temperature in Kelvin. Ideal gases also obey Boyle’s law and Charles’s law. Boyle’s law relates the pressure and volume of a given quantity of gas when the temperature is kept constant. Constant temperature is also called isothermal condition. Boyle’s law is stated as follows: P1 V2 = P2 V1

or P1V1 = P2V2

(1.13)

where P1 and V1 are the pressure and volume at condition 1 and P2 and V2 are the corresponding values at some other condition 2 where the temperature is not changed. Therefore, according to Boyle’s law, a given quantity of gas under isothermal conditions will double in volume if its pressure is halved and vice versa. In other words, the pressure is inversely proportional to the volume, provided the temperature is maintained constant. Since density and volume are inversely related, Boyle’s law also means that the pressure is directly proportional to the density at constant temperature. Thus, a given quantity of gas at a fixed temperature will double its density when the pressure is doubled. Similarly, a 10% reduction in pressure will cause the density to decrease by the same amount. Charles’s law states that for constant pressure, the gas volume is directly proportional to its temperature. Similarly, if volume is kept constant, the pressure varies directly as the temperature. Therefore, we can state the following:

Copyright 2005 by Taylor & Francis Group, LLC

V1 T1 = V2 T2

at constant pressure

(1.14)

P1 T1 = P2 T2

at constant volume

(1.15)

2785_C001.fm Page 12 Friday, April 1, 2005 3:46 PM

12

GAS PIPELINE HYDRAULICS

Therefore, according to Charles’s law, for an ideal gas at constant pressure, the volume will change in the same proportion as its temperature. Thus, a 20% increase in temperature will cause a 20% increase in volume as long as the pressure does not change. Similarly, if volume is kept constant, a 20% increase in temperature will result in the same percentage increase in gas pressure. Constant pressure is also known as isobaric condition. Example 2 A certain mass of gas occupies a volume of 1000 ft3 at 60 psig. If temperature is constant (isothermal) and the pressure increases to 120 psig, what is the final volume of the gas? The atmospheric pressure is 14.7 psi. Solution Boyle’s law can be applied because the temperature is constant. Using Equation 1.13, we can write V2 =

PV 1 1 P2

or V2 =

(60 + 14.7) × 1000 = 554.57 ft3 120 + 14.7

Note that the pressures must be converted to absolute values before being used in Equation 1.13. Example 3 At 75 psig and 70°F, a gas has a volume of 1000 ft3. If the volume is kept constant and the gas temperature increases to 120°F, what is the final pressure of the gas? For constant pressure at 75 psig, if the temperature increases to 120°F, what is the final volume? Use 14.7 psi for base pressure. Solution Since the volume is constant in the first part of the problem, Charles’s law applies. From Equation 1.15 75 + 14.7 70 + 460 = P2 120 + 460 Solving for P2, we get P2 = 98.16 psia or 83.46 psig

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 13 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

13

For the second part, the pressure is constant and Charles’s law can be applied. From Equation 1.14, we get 1000 70 + 460 = 120 + 460 V2 Solving for V2 we get V2 = 1094.34 ft3 Example 4 An ideal gas occupies a tank volume of 250 ft3 at a pressure of 80 psig and temperature of 110°F. (1) What is the gas volume at standard conditions of 14.73 psia and 60°F? Assume atmospheric pressure is 14.6 psia. (2) If the gas is cooled to 90°F, what is the gas pressure? Solution (1) Initial conditions P1 = 80 + 14.6 = 94.6 psia V1 = 250 ft3 T1 = 110 + 460 = 570°R Final conditions P2 = 14.73 psia V2 is to be calculated T2 = 60 + 460 = 520°R Using the ideal gas Equation 1.8, we can state that 94.6 × 250 14.73V2 = 570 520 V2 = 1464.73 ft3 (2) When the gas is cooled to 90°F, the final conditions are T2 = 90 + 460 = 550°R V2 = 250 ft3 P2 is to be calculated

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 14 Friday, April 1, 2005 3:46 PM

14

GAS PIPELINE HYDRAULICS

The initial conditions are P1 = 80 + 14.6 = 94.6 psia V1 = 250 ft3 T1 = 110 + 460 = 570°R It can be seen that the volume of gas is constant (tank volume) and the temperature reduces from 110°F to 90°F. Therefore, using Charles’s law and Equation 1.15, we can calculate the final pressure as follows. 94.6 570 = P2 550 Solving for P2, we get P2 = 91.28 psia = 91.28 – 14.6 = 76.68 psig

1.7 REAL GASES When dealing with real gases, we can apply the ideal gas equation discussed in the preceding sections and get reasonably accurate results only when the pressures are close to the atmospheric pressure. When pressures are higher, the ideal gas equation will not be accurate for most real gases. The error in calculations at high pressures using the ideal gas equation may be as high as 500% in some instances. This compares with errors of 2 to 3% at low pressures. At higher temperatures and pressures, the “equation of state” that relates pressure, volume, and temperature is used to calculate the properties of gases. Many of these correlations require a computer program to get accurate results in a reasonable amount of time. However, we can modify the ideal gas equation and obtain reasonably accurate results fairly quickly using manual calculations. Two terms called critical temperature and critical pressure need to be defined. The critical temperature of a pure gas is defined as the temperature above which a gas cannot be compressed to form a liquid, regardless of the pressure. The critical pressure is defined as the minimum pressure that is required at the critical temperature to compress a gas into a liquid. Real gases can be considered to follow a modified form of the ideal gas law discussed in Section 1.6. The modifying factor is included in the gas property known as the compressibility factor Z. This is also called the gas deviation factor. It can be defined as the ratio of the gas volume at a given temperature and pressure to the volume the gas would occupy if it were an ideal gas at the same temperature and pressure. Z is a dimensionless number less than 1.0 and it varies with temperature, pressure, and composition of the gas. Using the compressibility factor Z, the ideal gas equation is modified for real gas as follows: PV = ZnRT

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(1.16)

2785_C001.fm Page 15 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

where P = V = Z = T = n = R =

15

absolute pressure of gas, psia volume of gas, ft3 gas compressibility factor, dimensionless absolute temperature of gas, °R number of lb moles as defined in Equation 1.7 universal gas constant, 10.73 psia ft3/lb mole °R

The theorem known as corresponding states says that the extent of deviation of a real gas from the ideal gas equation is the same for all real gases when the gases are at the same corresponding state. The corresponding state can be represented by the two parameters called reduced temperature and reduced pressure. The reduced temperature is the ratio of the temperature of the gas to its critical temperature. Similarly, the reduced pressure is the ratio of the gas pressure to its critical pressure as indicated in the following equations:

where P = T = Tr = Pr = Tc = Pc =

Tr =

T Tc

(1.17)

Pr =

P Pc

(1.18)

absolute pressure of gas, psia absolute temperature of gas, °R reduced temperature, dimensionless reduced pressure, dimensionless critical temperature, °R critical pressure, psia

For example, suppose the critical temperature and critical pressure of methane are 343°R and 666 psia, respectively; the reduced temperature and pressure of the gas at 80°F and 1000 psia pressure are as follows: Tr =

80 + 460 = 1.57 343

and Pr =

1000 = 1.50 666

Therefore, according to the theorem of corresponding states, two gases, A and B, may be at different temperatures and pressures; however, if their reduced temperature and reduced pressure are the same, then their gas deviation factors (Z ) will be the same. Therefore, generalized plots showing the variation of Z with reduced temperature and reduced pressure can be used for most gases for calculating the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 16 Friday, April 1, 2005 3:46 PM

16

GAS PIPELINE HYDRAULICS

Pseudo-reduced pressure, Pr

0.8

1.7

1.2 1.6 3 1.

1.3

1.4 1.5 1.5 1.6 1.7 1.81.9 1.4 2.0 2.2

1.25 1.2

0.5 1.15 0.4

1.1

1.3 1.1 1.2 0.95

1.1

1.4 1.35

0.6

8

1.0 1.05

1.5 1.45

0.7 Compressibility factor, Z

7

1.1

2.4 2.6 3.0

Compressibility factor, Z

0.9

6

5

1.0

5

1. 05

3 4 Pseudo reduced temperature 3.0 2.8 2.6 2.4 2.2 2.0 1.9 1.8 1.7 1.6

1.

2

4

1

1.

1.1

0

1.3

1.2

0.3 1.05 0.25

3.0 1.1 2.8 2.6 2.4 2.2 2.0 1.9 1.0 1.8 1.7 1.6 1.4 1.3 0.9 7 8

Figure 1.2

1.1 MW < 40 1.2 1.1 1.05 9

Compressibility of natural gases Jan. 1, 1941

10 11 12 Pseudo reduced pressure, Pr

13

14

1.0

0.9 15

Compressibility factor chart for natural gases. (From Gas Processors Suppliers Association, Eng. Data Book, Vol. II. With permission.)

compressibility factor. Such a plot is shown in Figure 1.2. The calculation of the compressibility factor Z will be discussed in detail in Section 1.11 of this chapter. 1.8 NATURAL GAS MIXTURES As mentioned earlier, the critical temperature of a pure gas is defined as the temperature above which it cannot be liquefied, whatever the pressure of the gas. Similarly, the critical pressure is defined as the pressure above which liquid and gas cannot

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 17 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

17

coexist, regardless of the temperature. When the gas consists of a mixture of different components, the critical temperature and critical pressure are called the pseudocritical temperature and pseudo-critical pressure, respectively. If we know the composition of the gas mixture, we can calculate these pseudo-critical values of the mixture, using the critical pressure and temperature values of the pure components that constitute the gas mixture. The reduced temperature is defined as the ratio of the temperature of the gas to its critical temperature. Similarly, the reduced pressure is the ratio of gas pressure to its critical pressure. Both temperature and pressure are stated in absolute units. Similar to the pseudo-critical temperature and pseudo-critical pressure discussed above, for a gas mixture, we can define the pseudo-reduced temperature and the pseudo-reduced pressure. Thus,

where P T Tpr Ppr Tpc Ppc

= = = = = =

Tpr =

T Tpc

(1.19)

Ppr =

P Ppc

(1.20)

absolute pressure of gas mixture, psia absolute temperature of gas mixture, °R pseudo-reduced temperature, dimensionless pseudo-reduced pressure, dimensionless pseudo-critical temperature, °R pseudo-critical pressure, psia

In hydrocarbon mixtures, frequently we refer to gas components as C1, C2, C3, etc. These are equivalent to CH4 (methane), C2H6 (ethane), C3H8 (propane), and so on. A natural gas mixture that consists of components such as C1, C2, C3, and so forth is said to have an apparent molecular weight as defined by the equation M a = Σyi Mi

(1.21)

where Ma = apparent molecular weight of gas mixture yi = mole fraction of gas component i Mi = molecular weight of gas component i In a similar manner, from the given mole fractions of the gas components, we use Kay’s rule to calculate the average pseudo-critical properties of the gas mixture.

Copyright 2005 by Taylor & Francis Group, LLC

Tpc = ΣyiTc

(1.22)

Ppc = Σyi Pc

(1.23)

2785_C001.fm Page 18 Friday, April 1, 2005 3:46 PM

18

GAS PIPELINE HYDRAULICS

where Tc and Pc are the critical temperature and pressure, respectively, of the pure component (C1, C2, etc.) and yi refers to the mole fraction of the component. Tpc and Ppc are the average pseudo-critical temperature and pseudo-critical pressure, respectively, of the gas mixture. Example 5 Calculate the apparent molecular weight of a natural gas mixture that has 85% methane, 9% ethane, 4% propane, and 2% normal butane as shown below: Component

Percent

Molecular Weight

C1 C2 C3 n-C4

85 9 4 2

16.01 30.10 44.10 58.10

Total

100

Solution Using Equation 1.21, we get M a = (0.85 × 16.01) + (0.09 × 30.1) + (0.04 × 44.1) + (0.02 × 58.1) = 19.24 Therefore, the apparent molecular weight of the gas mixture is 19.24. Example 6 Calculate the pseudo-critical temperature and the pseudo-critical pressure of a natural gas mixture consisting of 83% methane, 12% ethane, and 5% propane. The critical properties of C1, C2, and C3 components are as follows:

Components

Critical Temperature, °R

Critical Pressure, psia

C1 C2 C3

343 550 666

666 707 617

Solution Using the given data, from Equation 1.22 and Equation 1.23, we calculate the pseudocritical properties as follows: Tpc = (0.83 × 343) + (0.12 × 550) + (0.05 × 666) = 383.99°R

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 19 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

19

and Ppc = (0.83 × 666) + (0.12 × 707) + (0.05 × 617) = 668.47 psia Therefore, the pseudo-critical temperature of the gas mixture = 383.99°R and the pseudo-critical pressure of the gas mixture = 668.47 psia. Example 7 If the temperature of the gas in the previous example is 70°F and the average gas pressure is 1200 psig, what are the pseudo-reduced temperature and pseudo-reduced pressure of this gas? Use 14.7 psia for base pressure. Solutions From Equation 1.19 and Equation 1.20, we get

The pseudo-reduced temperature Tpr = The pseudo-reduced pressure Ppr =

70 + 460 = 1.38 383.99

1200 + 14.7 = 1.82 668.47

1.9 PSEUDO-CRITICAL PROPERTIES FROM GAS GRAVITY If the percentages of the various components in the natural gas mixture are not available, we can calculate approximate values of the pseudo-critical properties of the gas mixture if we know the gas gravity. The pseudo-critical properties are calculated, approximately, from the following equations: Tpc = 170.491 + 307.344 G

(1.24)

Ppc = 709.604 – 58.718 G

(1.25)

where G = gas gravity (air = 1.00) Tpc = pseudo-critical temperature, °R Ppc = pseudo-critical pressure, psia Example 8 Calculate the gravity of a natural gas mixture consisting of 83% methane, 12% ethane, and 5% propane. From the gas gravity, calculate the pseudo-critical temperature and pseudo-critical pressure for this natural gas mixture.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 20 Friday, April 1, 2005 3:46 PM

20

GAS PIPELINE HYDRAULICS

Solution Using Kay’s rule for multicomponent mixtures and Equation 1.4 for gas gravity, we get G=

(0.83 × 16.04) + (0.12 × 30.07) + (0.05 × 44.10) = 0.6595 29

Therefore, the gas gravity is 0.6595. From Equation 1.24 and Equation 1.25, we calculate the pseudo-critical properties as follows: Tpc = 170.491 + 307.344 × (0.6595) = 373.18°R Ppc = 709.604 – 58.718 × (0.6595) = 670.88 psia Comparing the above values with the values calculated using the more accurate method in the previous example, we find that the value of Tpc is off by 2.8% and Ppc is off by 0.4%. These differences are small enough for most calculations related to natural gas pipeline hydraulics.

1.10 IMPACT OF SOUR GAS AND NON-HYDROCARBON COMPONENTS The Standing-Katz chart used for determining the compressibility factor (discussed in Section 1.11) of a gas mixture is accurate only if the amount of non-hydrocarbon components is small. Since sour gases contain carbon dioxide and hydrogen sulfide, adjustments must be made to take into account these components in calculations of the pseudo-critical temperature and pseudo-critical pressure. This method is described below. Depending on the amounts of carbon dioxide and hydrogen sulfide present in the sour gas, we calculate an adjustment factor e from e = 120(A0.9 − A1.6) + 15(B 0.5 – B 4.0)

(1.26)

where e = adjustment factor, °R A = sum of the mole fractions of CO2 and H2S B = mole fraction of H2S The pseudo-critical temperature is modified to get the adjusted pseudo-critical temperature T pc ′ from the following equation: T pc ′ = Tpc – e where T′pc = adjusted pseudo-critical temperature, °R

Copyright 2005 by Taylor & Francis Group, LLC

(1.27)

2785_C001.fm Page 21 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

21

Similarly, the pseudo-critical pressure is adjusted as follows: P′pc =

Ppc × Tpc′ Tpc + B(1 − B)ε

(1.28)

where P′pc = adjusted pseudo-critical pressure, psia. Example 9 The pseudo-critical temperature and the pseudo-critical pressure of a natural gas mixture were calculated as 370°R and 670 psia, respectively. If the CO2 content is 10% and H2S is 20%, calculate the adjustment factor e and the adjusted values of the pseudo-critical temperature and pressure. Solution A = 0.10 + 0.20 = 0.30 B = 0.20 The adjustment factor e from Equation 1.26 is e = 120 (0.30

0.9

– 0.301.6) + 15 (0.200.5 – 0.204) = 29.8082°R

Therefore, the adjustment factor e is 29.81°R. The adjusted values of the pseudo-critical properties are found using Equation 1.27 and Equation 1.28 as follows: T′pc = 370 – 29.81 = 340.19°R and P′pc =

670 × 340.19 = 608.18 psia 370 + 0.20(1 − 0.20) × 29.8082

Therefore, the adjusted pseudo-critical temperature = 340.19°R and the adjusted pseudo-critical pressure = 608.18 psia.

1.11 COMPRESSIBILITY FACTOR The compressibility factor, or gas deviation factor, was briefly mentioned in Section 1.6. It is a measure of how close a real gas is to an ideal gas. The compressibility factor is defined as the ratio of the gas volume at a given temperature and pressure to the volume the gas would occupy if it were an ideal gas at the same temperature and pressure. The compressibility factor is a dimensionless number close to 1.00 and is

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 22 Friday, April 1, 2005 3:46 PM

22

GAS PIPELINE HYDRAULICS

a function of the gas gravity, gas temperature, gas pressure, and the critical properties of the gas. As an example, a particular natural gas mixture can have a compressibility factor equal to 0.87 at 1000 psia and 80°F. Charts have been constructed that depict the variation of Z with the reduced temperature and reduced pressure. Another term, the “supercompressibility factor,” Fpv , which is related to the compressibility factor Z, is defined as follows: Fpv =

1 Z

(1.29)

or Z=

1 ( Fpv )2

(1.30)

As an example, if the compressibility factor Z = 0.85, using Equation 1.29, we calculate the supercompressibility factor, Fpv , as follows: Fpv =

1 0.85

= 1.0847

There are several approaches to calculating the compressibility factor for a particular gas temperature T and pressure P. One method uses the critical temperature and critical pressure of the gas mixture. First, the reduced temperature, Tr, and reduced pressure, Pr, are calculated from the given gas temperature and gas pressure and the critical temperature and critical pressure using Equation 1.17 and Equation 1.18. Once we know the values of Tr and Pr, the compressibility factor can be found from charts similar to the Standing-Katz chart. This will be illustrated using an example. The following methods are available to calculate the compressibility factor: a. b. c. d.

Standing-Katz method Dranchuk, Purvis, and Robinson method AGA method CNGA method

Although the Standing-Katz method is the most popular, we will discuss this as well as the AGA and CNGA methods. 1.11.1

Standing-Katz Method

The Standing-Katz method of calculating compressibility factor is based on the use of a graph that has been constructed for binary mixtures and saturated hydrocarbon vapor. This method is used generally for sweet natural gas mixtures containing various hydrocarbon components. When the natural gas mixture contains appreciable amounts of non-hydrocarbons such as nitrogen, hydrogen sulfide, and carbon dioxide, certain corrections must be applied for these components. These adjustments are applied to

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 23 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

23

the critical temperatures and pressures and were discussed in Section 1.10. The Standing-Katz chart for compressibility factors is shown in Figure 1.2. The use of this chart will be explained in an example problem. 1.11.2

Dranchuk, Purvis, and Robinson Method

In this method of calculating the compressibility factor, the Benedict-Webb-Rubin equation of state is used to correlate the Standing-Katz chart. The coefficients A1, A2, etc. are used in a polynomial function of the reduced density ρr as follows: A A A A ρ5 A7 ρr3 A Z = 1 + A1 + 2 + 33 ρr + A4 + 5 ρr2 + 5 6 r + (− A ρ2 ) Tpr Tpr Tpr Tpr Tpr3 1 + A8 ρr2 e 8 r

(

)

(1.31) where

ρr = and A1 A2 A3 A4 A5 A6 A7 A8 Ppr Tpr

= = = = = = = = = =

0.27 Ppr ZTpr

(1.32)

0.31506237 −1.04670990 −0.57832729 0.53530771 −0.61232032 −0.10488813 0.68157001 0.68446549 pseudo-reduced pressure pseudo-reduced temperature

Other symbols have been defined earlier. 1.11.3

American Gas Association (AGA) Method

The AGA method for the compressibility factor uses a complicated mathematical algorithm and, therefore, does not lend itself easily to manual calculations. Generally, a computer program is used to calculate the compressibility factor. Mathematically, the AGA method is represented by the following function: Z = Function (gas properties, pressure, temperature)

(1.33)

where gas properties include the critical temperature, critical pressure, and gas gravity.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 24 Friday, April 1, 2005 3:46 PM

24

GAS PIPELINE HYDRAULICS

The AGA-IGT Report No. 10 describes in detail this method of calculating Z. This approach is valid for gas temperatures in the range of 30 to 120°F and for pressures not exceeding 1380 psig. The compressibility factors calculated using this method are quite accurate and generally within 0.03% of those calculated using the Standing-Katz chart in this range of temperatures and pressures. When temperatures and pressures are higher than these values, the compressibility factor calculated using this method is within 0.07% of the value obtained from the Standing-Katz chart. The reader may also refer to the AGA publication Report No. 8, Second Edition, November 1992, for more information on compressibility factor calculation methods. 1.11.4

California Natural Gas Association (CNGA) Method

This is a fairly simple equation for quickly calculating the compressibility factor when the gas gravity, temperature, and pressure are known. The following equation is used for calculating the compressibility factor Z: Z=

1 Pavg 344 ,400 (10 )1.785G 1 + T f3.825

(1.34)

This formula for the compressibility factor is valid when the average gas pressure, Pavg, is more than 100 psig. For pressures less than 100 psig, Z is approximately equal to 1.00 where Pavg = average gas pressure, psig Tf = average gas temperature, °R G = gas gravity (air = 1.00) Note that the pressure used in Equation 1.34 is the gauge pressure. In a gas pipeline, the pressure varies along the length of the pipeline. The compressibility factor Z also varies and must therefore be calculated for an average pressure at any location on the pipeline. If two points along the pipeline are at pressures P1 P +P and P2, we could use an average pressure of 1 2 2 . However, the following formula is used for a more accurate value of the average pressure: Pavg =

2 P ×P P1 + P2 − 1 2 3 P1 + P2

(1.35)

Another form of the average pressure in a pipe segment is

Pavg =

Copyright 2005 by Taylor & Francis Group, LLC

2 P13 − P23 3 P12 − P22

(1.36)

2785_C001.fm Page 25 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

25

Example 10 Using the Standing-Katz compressibility chart, calculate the compressibility factor for the gas in Example 7 at 70°F and 1200 psig. Use the values of Tpc and Ppc calculated in Example 7. Solutions From previous Example 7, The pseudo-reduced temperature Tpr = 1.38 The pseudo-reduced pressure Ppr = 1.82 Using the Standing-Katz chart in Figure 1.2, we read the value of Z as Z = 0.770 Example 11 A natural gas mixture consists of the following components: Component

Mole Fraction y

C1 C2 C3 N2 CO2 H 2S

0.780 0.005 0.002 0.013 0.016 0.184

(a) Calculate the apparent molecular weight of the gas, gas gravity, and the pseudocritical temperature and pseudo-critical pressure. (b) Calculate the compressibility factor of the gas at 90°F and 1200 psia. Solution Using Table 1.1, we create the following table showing the molecular weight (M), critical temperature (Tc), and critical pressure (Pc) for each of the component gases. The molecular weight of the mixture and the pseudo-critical temperature and pseudocritical pressure are then calculated using Equation 1.22 and Equation 1.23. Component

y

M

yM

TC

PC

yTC

yPC

C1 C2 C3 N2 CO2 H 2S

0.780 0.005 0.002 0.013 0.016 0.184

16.04 30.07 44.10 28.01 44.01 34.08

12.5112 0.1504 0.0882 0.3641 0.7042 6.2707

343.34 550.07 665.93 227.52 547.73 672.40

667.00 707.80 615.00 492.80 1070.00 1300.00

267.81 2.75 1.33 2.96 8.76 123.72

520.26 3.54 1.23 6.41 17.12 239.20

407.33

787.76

Total

Copyright 2005 by Taylor & Francis Group, LLC

1.000

20.0888

2785_C001.fm Page 26 Friday, April 1, 2005 3:46 PM

26

GAS PIPELINE HYDRAULICS

Therefore, the apparent molecular weight of the natural gas is Mw = ΣyM = 20.09 The gas gravity, using Equation 1.4, is G=

20.09 = 0.6928 29

Next, we calculate the pseudo-critical values Pseudo-critical temperature = ΣyTc = 407.33°R Pseudo-critical pressure = ΣyPc = 787.76 psia Since this sour gas contains more than 5% non-hydrocarbons, we will adjust the pseudo-critical properties using Equation 1.26 through Equation 1.28. We first calculate the temperature adjustment factor e, using Equation 1.26, as follows: A = (0.016 + 0.184) = 0.20 and B = 0.184 Therefore, the adjustment factor is e = 120[(0.2)0.9 – (0.2)1.6] + 15[(0.184)0.5 – (0.184)4.0] = 25.47°R The adjusted pseudo-critical temperature from Equation 1.27 is T′pc = 407.33 – 25.47 = 381.86°R and the adjusted pseudo-critical pressure from Equation 1.28 is P′pc =

787.76 × 381.86 = 731.63 psia 407.33 + 0.184 × (1 − 0.184) × 25.47

Next, we calculate the compressibility factor at 90°F and 1200 psia pressure using these values as follows: From Equation 1.19, pseudo-reduced temperature =

90 + 460 = 1.44 381.86

From Equation 1.20, pseudo-reduced pressure =

1200 = 1.64 731.63

Finally, from the Standing-Katz chart, we get the compressibility factor for the reduced temperature and reduced pressure as Z = 0.825

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 27 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

27

Example 12 The gravity of a natural gas mixture is 0.60. Calculate the compressibility factor of this gas at 1200 psig pressure and a temperature of 70°F, using the CNGA method. Solution Gas temperature Tf = 70 + 460 = 530°R From Equation 1.34, the Z factor can be written as 1 1200 × 344, 400 × (10)1.785 x 0.60 =1+ = 1.1849 Z 530 3.825 Therefore, solving for Z, we get Z = 0.8440

1.12 HEATING VALUE The heating value of a gas is defined as the thermal energy per unit volume of the gas. It is expressed in Btu/ft3. For natural gas, it is approximately in the range of 900 to 1200 Btu/ft3. There are two heating values used in the industry. These are the lower heating value (LHV) and higher heating value (HHV). For a gas mixture, the term gross heating value is used. It is calculated based upon the heating values of the component gases and their mole fractions using the following equation: H m = Σ( yi Hi )

(1.37)

where Hm = gross heating value of mixture, Btu/ft3 yi = mole fraction or percent of gas component i Hi = heating value of gas component, Btu/ft3 For example, a natural gas mixture consisting of 80% of gas A (heating value = 900 Btu/ft3) and 20% of gas B (heating value = 1000 Btu/ft3) will have a gross heating value of Hm = (0.8 × 900) + (0.2 × 1000) = 920 Btu/ft3. 1.13 SUMMARY We discussed several gas properties that influence gas pipeline transportation. The ideal gas equation was introduced along with Boyle’s law and Charles’s law, and how they can be applied with modifications to real gases and real gas mixtures was explained. The gas deviation factor, or compressibility factor, which modifies ideal gas behavior, was introduced. Critical properties of hydrocarbon gases and mixtures and the reduced temperature and pressure that determine the state of a gas were

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 28 Friday, April 1, 2005 3:46 PM

28

GAS PIPELINE HYDRAULICS

explained. Variation of the compressibility factor with pressure and temperature was explored, and calculation methodologies using analytical and graphical approaches were covered. The influence of non-hydrocarbon components in a natural gas mixture was also discussed, along with correction factors for CO2 and H2S in sour gas.

PROBLEMS 1. A natural gas mixture consists of three components, C1, C2, and C3. Their mole fractions and viscosities at a particular temperature are indicated below: Component

Mole Fraction y

Viscosity, cP

C1 C2 C3

0.9000 0.0800 0.0200

0.0130 0.0112 0.0098

Total

1.000

Calculate the viscosity of the gas mixture. 2. At 100 psig and 75°F, a gas has a volume of 800 ft3. If the volume is kept constant and the gas temperature increases to 100°F, what is the final pressure of the gas? Keeping the pressure constant at 100 psig, if the temperature increases to 100°F, what is the final volume? Use 14.73 psi for base pressure. 3. Calculate the apparent molecular weight of a natural gas mixture that has 89% methane, 8% ethane, 2% propane, and 1% normal butane as shown below. Component

Percent

Molecular Weight

C1 C2 C3 C4

89 8 2 1

16.01 30.10 44.10 58.10

Total

100

4. Calculate the pseudo-critical temperature and the pseudo-critical pressure of a natural gas mixture consisting of 89% methane, 8% ethane, and 3% propane. The critical properties of C1, C2, and C3 components are as follows: Components

Critical Temperature, °R

Critical Pressure, psia

C1 C2 C3

343 550 666

667 708 615

5. If the temperature of the gas in the previous example is 80°F and the average gas pressure is 1000 psig, what are the pseudo-reduced temperature and pseudoreduced pressure of this gas? Use 14.7 psia for base pressure.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 29 Friday, April 1, 2005 3:46 PM

GAS PROPERTIES

29

6. Calculate the gravity of a natural gas mixture consisting of 84% methane, 10% ethane, and 6% propane. From the gas gravity, calculate the pseudo-critical temperature and pseudo-critical pressure for this natural gas mixture. 7. The pseudo-critical temperature and pressure of a natural gas mixture were calculated as 380°R and 675 psia. If the CO2 content is 12% and H2S is 22%, calculate the adjustment factor e and the adjusted values of the pseudo-critical temperature and pseudo-critical pressure. 8. Using the Standing-Katz compressibility chart, calculate the compressibility factor for the gas in Problem 7 at 80°F and 1000 psig. Use the values of Tpc and Ppc calculated in Problem 7. 9. A natural gas mixture consists of the following components: Component

Mole Fraction y

C1 C2 C3 N2 CO2 H 2S

0.850 0.004 0.002 0.014 0.010 0.120

(a) Calculate the apparent molecular weight of the gas, gravity, and the pseudocritical temperature and pseudo-critical pressure. (b) Calculate the compressibility factor of the gas at 100°F and 1400 psia. 10. The gravity of a natural gas mixture is 0.62. Calculate the compressibility factor of this gas at 1400 psig and a temperature of 80°F, using the CNGA method.

REFERENCES 1. McCain, W.D. Jr., The Properties of Petroleum Fluids, Petroleum Publishing Company, Tulsa, OK, 1973. 2. Holman, J.P., Thermodynamics, 2nd ed., McGraw-Hill, New York, 1974. 3. Gas Processors Suppliers Association, Engineering Data Book, 10th ed., Tulsa, OK, 1994. 4. AGA Report No. 10, Steady Flow in Gas Pipelines, AGA, 1965. 5. AGA Report No. 8, Compressibility Factors, AGA, Nov. 1992.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C001.fm Page 30 Friday, April 1, 2005 3:46 PM

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 31 Friday, April 1, 2005 3:23 PM

CHAPTER

2

Pressure Drop Due to Friction In this chapter we will discuss the various methods of calculating the pressure drop due to friction in a gas pipeline. The pipeline throughput (ﬂow rate) will depend upon the gas properties, pipe diameter and length, initial gas pressure and temperature, and the pressure drop due to friction. Commonly used formulas will be reviewed and illustrated using examples. The impact of internal conditions of the pipe on the pipe capacity will also be explored. 2.1 BERNOULLI’S EQUATION As gas ﬂows through a pipeline, the total energy of the gas at various points consists of energy due to pressure, energy due to velocity, and energy due to position or elevation above an established datum. Bernoulli’s equation simply connects these components of the energy of the ﬂowing ﬂuid to form an energy conservation equation. Bernoulli’s equation is stated as follows, considering two points, 1 and 2, as shown in Figure 2.1. ZA +

P V2 PA VA 2 + + H p = ZB + B + B + hf γ 2g γ 2g

(2.1)

where Hp is the equivalent head added to the ﬂuid by a compressor at A and hf represents the total frictional pressure loss between points A and B. Starting with the basic energy Equation 2.1, applying gas laws, and after some simpliﬁcation, various formulas were developed over the years to predict the performance of a pipeline transporting gas. These formulas are intended to show the relationship between the gas properties, such as gravity and compressibility factor, with the ﬂow rate, pipe diameter and length, and the pressures along the pipeline. Thus, for a given pipe size and length, we can predict the ﬂow rate possible through a pipeline based upon an inlet pressure and an outlet pressure of a pipe segment. Simpliﬁcations are sometimes introduced, such as uniform gas temperature and no heat transfer between the gas and the surrounding soil in a buried pipeline, in order 31

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 32 Friday, April 1, 2005 3:23 PM

32

GAS PIPELINE HYDRAULICS

Velocity VA Pressure PA

B

Velocity VB Pressure PB

Flow A

Datum for Elevations Figure 2.1

Energy of flow of a fluid.

to adopt these equations for manual calculations. With the advent of microcomputers, we are able to introduce heat transfer effects and, therefore, more accurately model gas pipelines, taking into consideration gas ﬂow temperatures, soil temperatures, and thermal conductivities of pipe material, insulation, and soil. In this chapter we will concentrate on steady-state isothermal ﬂow of gas in pipelines. Appendix D includes an output report from a commercial gas pipeline simulation model that takes into account heat transfer. For most practical purposes, the assumption of isothermal ﬂow is good enough, since in long transmission lines the gas temperature reaches constant values, anyway. 2.2 FLOW EQUATIONS Several equations are available that relate the gas ﬂow rate with gas properties, pipe diameter and length, and upstream and downstream pressures. These equations are listed as follows: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

General Flow equation Colebrook-White equation Modiﬁed Colebrook-White equation AGA equation Weymouth equation Panhandle A equation Panhandle B equation IGT equation Spitzglass equation Mueller equation Fritzsche equation

We will discuss each of these equations, their limitations, and their applicability to compressible ﬂuids, such as natural gas, using illustrated examples. A comparison of these equations will also be discussed using an example pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 33 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

33

2.3 GENERAL FLOW EQUATION The General Flow equation, also called the Fundamental Flow equation, for the steady-state isothermal ﬂow in a gas pipeline is the basic equation for relating the pressure drop with ﬂow rate. The most common form of this equation in the U.S. Customary System (USCS) of units is given in terms of the pipe diameter, gas properties, pressures, temperatures, and ﬂow rate as follows. Refer to Figure 2.2 for an explanation of symbols used. T P 2 − P22 Q = 77.54 b 1 Pb GT f LZf where Q = f = Pb = Tb = P1 = P2 = G = Tf = L = Z = D =

0.5

D 2.5

(USCS units)

(2.2)

gas ﬂow rate, measured at standard conditions, ft3/day (SCFD) friction factor, dimensionless base pressure, psia base temperature, °R(460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂowing temperature, °R (460 + °F) pipe segment length, mi gas compressibility factor at the ﬂowing temperature, dimensionless pipe inside diameter, in.

It must be noted that for the pipe segment from section 1 to section 2, the gas temperature Tf is assumed to be constant (isothermal ﬂow). In SI units, the General Flow equation is stated as follows:

(

)

2 2 T P1 − P2 Q = 1.1494 × 10 −3 b Pb GT f LZf

0.5

D 2.5

(SI units)

Diameter D Temperature Tf

Pressure P1

1 Figure 2.2

Steady flow in gas pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

Pressure P2

Flow Q

Length L

2

(2.3)

2785_C002.fm Page 34 Friday, April 1, 2005 3:23 PM

34

GAS PIPELINE HYDRAULICS

where Q = gas ﬂow rate, measured at standard conditions, m3/day f = friction factor, dimensionless Pb = base pressure, kPa Tb = base temperature, K (273 + °C) P1 = upstream pressure, kPa P2 = downstream pressure, kPa G = gas gravity (air = 1.00) Tf = average gas ﬂowing temperature, K (273 + °C) L = pipe segment length, km Z = gas compressibility factor at the ﬂowing temperature, dimensionless D = pipe inside diameter, mm Due to the nature of Equation 2.3, the pressures can also be in MPa or Bar, as long as the same consistent unit is used. Equation 2.2 relates the capacity (ﬂow rate or throughput) of a pipe segment of length L, based on an upstream pressure of P1 and a downstream pressure of P2 as shown in Figure 2.2. It is assumed that there is no elevation difference between the upstream and downstream points; therefore, the pipe segment is horizontal. Upon examining the General Flow Equation 2.2, we see that for a pipe segment of length L and diameter D, the gas ﬂow rate Q (at standard conditions) depends on several factors. Q depends on gas properties represented by the gravity G and the compressibility factor Z. If the gas gravity is increased (heavier gas), the ﬂow rate will decrease. Similarly, as the compressibility factor Z increases, the ﬂow rate will decrease. Also, as the gas ﬂowing temperature Tf increases, throughput will decrease. Thus, the hotter the gas, the lower the ﬂow rate will be. Therefore, to increase the ﬂow rate, it helps to keep the gas temperature low. The impact of pipe length and inside diameter is also clear. As the pipe segment length increases for given pressure P1 and P2, the ﬂow rate will decrease. On the other hand, the larger the diameter, the larger the ﬂow rate will be. The term P12 – P22 represents the driving force that causes the ﬂow rate from the upstream end to the downstream end. As the downstream pressure P2 is reduced, keeping the upstream pressure P1 constant, the ﬂow rate will increase. It is obvious that when there is no ﬂow rate, P1 is equal to P2. It is due to friction between the gas and pipe walls that the pressure drop (P1 – P2) occurs from the upstream point 1 to downstream point 2. The friction factor f depends on the internal condition of the pipe as well as the type of ﬂow (laminar or turbulent) and will be discussed in detail beginning in Section 2.8. Sometimes the General Flow equation is represented in terms of the transmission factor F instead of the friction factor f. This form of the equation is as follows. T P 2 − P22 Q = 38.77 F b 1 Pb GT f LZ

Copyright 2005 by Taylor & Francis Group, LLC

0.5

D 2.5

(USCS units)

(2.4)

2785_C002.fm Page 35 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

35

where the transmission factor F and friction factor f are related by F=

2

(2.5)

f

and in SI units

(

2 2 Tb P1 − P2 −4 Q = 5.747 × 10 F Pb GT f LZ

)

0.5

D 2.5

(SI units)

(2.6)

We will discuss several aspects of the General Flow equation before moving on to the other formulas for pressure drop calculation.

2.4 EFFECT OF PIPE ELEVATIONS When elevation difference between the ends of a pipe segment is included, the General Flow equation is modiﬁed as follows: T P 2 − e s P22 Q = 38.77 F b 1 Pb GT f Le Z

0.5

D 2.5

(USCS units)

(2.7)

D 2.5

(2.8)

and in SI units

(

s 2 2 T P1 − e P2 Q = 5.747 × 10 F b Pb GT f Le Z −4

)

0.5

(SI units)

where Le =

L (e s − 1) s

(2.9)

The equivalent length, Le, and the term es take into account the elevation difference between the upstream and downstream ends of the pipe segment. The parameter s depends upon the gas gravity, gas compressibility factor, the ﬂowing temperature, and the elevation difference. It is deﬁned as follows in USCS units: H − H1 s = 0.0375G 2 Tf Z

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(2.10)

2785_C002.fm Page 36 Friday, April 1, 2005 3:23 PM

36

where s = H1 = H2 = e =

GAS PIPELINE HYDRAULICS

elevation adjustment parameter, dimensionless upstream elevation, ft downstream elevation, ft base of natural logarithms (e = 2.718…)

Other symbols are as deﬁned earlier. In SI units, the elevation adjustment parameter s is deﬁned as follows: H − H1 s = 0.0684G 2 Tf Z

(SI units)

(2.11)

where H1 = upstream elevation, m H2 = downstream elevation, m Other symbols are as deﬁned earlier. In the calculation of Le in Equation 2.9, we have assumed that there is a single slope between the upstream point 1 and the downstream point 2 in Figure 2.2. If, however, the pipe segment of length L has a series of slopes, then we introduce a parameter j as follows for each individual pipe subsegment that constitutes the pipe length from point 1 to point 2. j=

es − 1 s

(2.12)

The parameter j is calculated for each slope of each pipe subsegment of length L1, L2, etc. that make up the total length L. The equivalent length term Le in Equation 2.7 and Equation 2.8 is calculated by summing the individual slopes as deﬁned below. Le = j1L1 + j2L2es1 + j3L3es2 + …

(2.13)

The terms j1, j2, etc. for each rise or fall in the elevations of individual pipe subsegments are calculated for the parameters s1, s2, etc. for each segment in accordance with Equation 2.12, from the pipeline inlet to the end of each segment. In the subsequent sections of this chapter, we will discuss how the friction factor and transmission factor are calculated using various equations such as ColebrookWhite and AGA. It is important to note that the General Flow equation is the most commonly used equation to calculate the ﬂow rate and pressure in a gas pipeline. In order to apply it correctly, we must use the correct friction factor or transmission factor. The Colebrook equation, AGA equation, and other empirical equations are used to calculate the friction factor to be used in the General Flow equation. Several other equations, such as Panhandle A, Panhandle B, and Weymouth, calculate the ﬂow rate for a given pressure without using a friction factor or transmission factor. However, an equivalent friction factor (or transmission factor) can be calculated using these methods as well.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 37 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

37

2.5 AVERAGE PIPE SEGMENT PRESSURE In the General Flow equation, the compressibility factor Z is used. This must be calculated at the gas ﬂowing temperature and average pressure in the pipe segment. Therefore, it is important to ﬁrst calculate the average pressure in a pipe segment, described in Figure 2.2. Consider a pipe segment with upstream pressure P1 and downstream pressure P2, as in Figure 2.2. An average pressure for this segment must be used to calculate the compressibility factor of gas at the average gas temperature Tf. As a ﬁrst approximation, we may use an arithmetic average of (P1 + P2)/2. However, it has been found that a more accurate value of the average gas pressure in a pipe segment is Pavg =

PP 2 P1 + P2 − 1 2 P1 + P2 3

(2.14)

Another form of the average pressure in a pipe segment is

Pavg =

3 3 2 P1 − P2 3 P12 − P2 2

(2.15)

It must be noted that the pressures used in the General Flow equation are all in absolute units. Therefore, gauge pressure units should be converted to absolute pressure by adding the base pressure. For example, the upstream and downstream pressures are 1000 psia and 900 psia, respectively. From Equation 2.14, the average pressure is Pavg =

1000 × 900 2 1000 + 900 − = 950.88 psia 1900 3

Compare this to the arithmetic average of Pavg =

1 (1000 + 900) = 950 psia 2

2.6 VELOCITY OF GAS IN A PIPELINE The velocity of gas ﬂow in a pipeline represents the speed at which the gas molecules move from one point to another. Unlike a liquid pipeline, due to compressibility, the gas velocity depends upon the pressure and, hence, will vary along the pipeline even if the pipe diameter is constant. The highest velocity will be at the downstream end, where the pressure is the least. Correspondingly, the least velocity will be at the upstream end, where the pressure is higher.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 38 Friday, April 1, 2005 3:23 PM

38

GAS PIPELINE HYDRAULICS

Consider a pipe transporting gas from point A to point B as shown in Figure 2.2. Under steady state ﬂow, at A, the mass ﬂow rate of gas is designated as M and will be the same as the mass ﬂow rate at point B, if between A and B there is no injection or delivery of gas. The mass being the product of volume and density, we can write the following relationship for point A: M = Qr

(2.16)

The volume rate Q can be expressed in terms of the ﬂow velocity u and pipe cross sectional area A as follows: Q=uA

(2.17)

Therefore, combining Equation 2.16 and Equation 2.17 and applying the conservation of mass to points A and B, we get M1 = u1 A1r1 = M2 = u2 A2 r2

(2.18)

where subscripts 1 and 2 refer to points A and B, respectively. If the pipe is of uniform cross section between A and B, then A1 = A2 = A. Therefore, the area term in Equation 2.18 can be dropped, and the velocities at A and B are related by the following equation: u1 r1 = u2 r2

(2.19)

Since the ﬂow of gas in a pipe can result in variation of temperature from point A to point B, the gas density will also vary with temperature and pressure. If the density and velocity at one point are known, the corresponding velocity at the other point can be calculated using Equation 2.19. If inlet conditions are represented by point A and the volume ﬂow rate Q at standard conditions of 60°F and 14.7 psia are known, we can calculate the velocity at any point along the pipeline at which the pressure and temperature of the gas are P and T, respectively. The velocity of gas at section 1 is related to the ﬂow rate Q1 at section 1 and pipe cross-sectional area A as follows from Equation 2.17: Q1 = u1 A The mass ﬂow rate M at section 1 and 2 is the same for steady-state ﬂow. Therefore, M = Q1r1 = Q2 r2 = Qb rb

(2.20)

where Qb is the gas ﬂow rate at standard conditions and rb is the corresponding gas density. Therefore, simplifying Equation 2.20, ρ Q1 = Qb b ρ1

Copyright 2005 by Taylor & Francis Group, LLC

(2.21)

2785_C002.fm Page 39 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

39

Applying the gas law Equation 1.9, we get P1 = Z1 RT1 ρ1 or

ρ1 =

P1 Z1 RT1

(2.22)

where P1 and T1 are the pressure and temperature at pipe section 1. Similarly, at standard conditions,

ρb =

Pb Z b RTb

(2.23)

From Equation 2.21, Equation 2.22, and Equation 2.23, we get P T Z Q1 = Qb b 1 1 Tb P1 Z b

(2.24)

Since Zb = 1.00, approximately, we can simplify this to P T Q1 = Qb b 1 Z1 Tb P1

(2.25)

Therefore, the gas velocity at section 1 is, using Equation 2.17 and Equation 2.25, u1 =

P T Qb Z1 Pb T1 4 × 144 = Qb Z1 b 1 2 A Tb P1 Tb P1 πD

or Q P ZT u1 = 0.002122 b2 b 1 1 D Tb P1 where u1 = Qb = D = Pb = Tb = P1 = T1 = Z1 =

(USCS units)

upstream gas velocity, ft/s gas ﬂow rate, measured at standard conditions, ft3/day (SCFD) pipe inside diameter, in. base pressure, psia base temperature, °R(460 + °F) upstream pressure, psia upstream gas temperature, °R(460 + °F) gas compressibility factor at upstream conditions, dimensionless

Copyright 2005 by Taylor & Francis Group, LLC

(2.26)

2785_C002.fm Page 40 Friday, April 1, 2005 3:23 PM

40

GAS PIPELINE HYDRAULICS

Similarly, the gas velocity at section 2 is given by Q P Z T u2 = 0.002122 b2 b 2 2 D Tb P2

(USCS units)

(2.27)

In general, the gas velocity at any point in a pipeline is given by Q P ZT u = 0.002122 b2 b D Tb P

(2.28)

In SI units, the gas velocity at any point in a gas pipeline is given by Q P ZT u = 14.7349 b2 b D Tb P

(SI units)

(2.29)

where u = gas velocity, m/s Qb = gas ﬂow rate, measured at standard conditions, m3/day D = pipe inside diameter, mm Pb = base pressure, kPa Tb = base temperature, K(273 + °C) P = pressure, kPa T = average gas ﬂowing temperature, K(273 + °C) Z = gas compressibility factor at the ﬂowing temperature, dimensionless Since the right-hand side of Equation 2.29 contains ratios of pressures, any consistent unit can be used, such as kPa, MPa, or Bar.

2.7 EROSIONAL VELOCITY We have seen from the preceding section that the gas velocity is directly related to the ﬂow rate. As ﬂow rate increases, so does the gas velocity. How high can the gas velocity be in a pipeline? As the velocity increases, vibration and noise are evident. In addition, higher velocities will cause erosion of the pipe interior over a long period of time. The upper limit of the gas velocity is usually calculated approximately from the following equation: umax =

100

ρ

where umax = maximum or erosional velocity, ft/s r = gas density at ﬂowing temperature, lb/ft3

Copyright 2005 by Taylor & Francis Group, LLC

(2.30)

2785_C002.fm Page 41 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

41

Since the gas density r may be expressed in terms of pressure and temperature, using the gas law Equation 1.8, the maximum velocity Equation 2.30 can be rewritten as umax = 100 where Z = R = T = G= P =

ZRT 29GP

(USCS units)

(2.31)

compressibility factor of gas, dimensionless gas constant = 10.73 ft3 psia/lb-moleR gas temperature, °R gas gravity (air = 1.00) gas pressure, psia

Usually, an acceptable operational velocity is 50% of the above. Example 1 A gas pipeline, NPS 20 with 0.500 in. wall thickness, transports natural gas (speciﬁc gravity = 0.6) at a ﬂow rate of 250 MMSCFD at an inlet temperature of 60°F. Assuming isothermal ﬂow, calculate the velocity of gas at the inlet and outlet of the pipe if the inlet pressure is 1000 psig and the outlet pressure is 850 psig. The base pressure and base temperature are 14.7 psia and 60°F, respectively. Assume compressibility factor Z = 1.00. What is the erosional velocity for this pipeline based on the above data and a compressibility factor Z = 0.90? Solution If we assume compressibility factor Z = 1.00, then using Equation 2.26, the velocity of gas at the inlet pressure of 1000 psig is 250 × 10 6 14.7 60 + 460 u1 = 0.002122 = 21.29 ft/s 19.0 2 60 + 460 1014.7 and the gas velocity at the outlet is by proportions u2 = 21.29 ×

1014.7 = 24.98 ft/s 864.7

The erosional velocity is found for Z = 0.90, using Equation 2.31, umax = 100

0.9 × 10.73 × 520 = 53.33 ft/s 29 × 0.6 × 1014.7

Example 2 A gas pipeline, DN 500 with 12 mm wall thickness, transports natural gas (speciﬁc gravity = 0.6) at a ﬂow rate of 7.5 Mm3/day at an inlet temperature of 15°C. Assuming

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 42 Friday, April 1, 2005 3:23 PM

42

GAS PIPELINE HYDRAULICS

isothermal ﬂow, calculate the velocity of gas at the inlet and outlet of the pipe if the inlet pressure is 7 MPa and the outlet pressure is 6 MPa. The base pressure and base temperature are 0.1 MPa and 15°C. Assume compressibility factor Z = 0.95. Solution Inside diameter of pipe D = 500 – (2 × 12) = 476 mm. Flow rate at standard conditions Qb = 7.5 × 106 m3/day. Using Equation 2.29, the velocity of gas at the inlet pressure of 7 MPa is 7.5 × 10 6 0.1 0.95 × 288 u1 = 14.7349 = 6.62 m/s 7.0 4762 15 + 273 and the gas velocity at the outlet is by proportions u2 = 6.62 ×

7.0 = 7.72 m/s 6.0

In the preceding Examples 1 and 2, we have assumed the value of compressibility factor Z to the constant. A more accurate solution will be to calculate the value of Z using one of the methods outlined in Chapter 1, such as the CNGA or Standing-Katz method. For example, if we used the CNGA Equation 1.34, the compressibility factor in Example 1 will be 1 1 + 1000×344400×(10 )1.785× 0.6 5203.825 = 0.8578 at an inlet pressure of 1000 psig.

Z1 =

and Z2 =

1

1 + 5203.825 = 0.8765 at an outlet pressure of 850 psig. 850 × 344400 × (10 )1.785× 0.6

The inlet and outlet gas velocities then will be modiﬁed as follows: Inlet velocity u1 = 0.8578 × 21.29 = 18.26 ft/s Outlet velocity u2 = 0.8765 × 24.98 = 21.90 ft/s

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 43 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

43

2.8 REYNOLDS NUMBER OF FLOW An important parameter in ﬂow of ﬂuids in a pipe is the nondimensional term Reynolds number. The Reynolds number is used to characterize the type of ﬂow in a pipe, such as laminar, turbulent, or critical ﬂow. It is also used to calculate the friction factor in pipe ﬂow. We will ﬁrst outline the calculation of the Reynolds number based upon the properties of the gas and pipe diameter and then discuss the range of Reynolds number for the various types of ﬂow and how to calculate the friction factor. The Reynolds number is a function of the gas ﬂow rate, pipe inside diameter, and the gas density and viscosity and is calculated from the following equation: Re = where Re u D r m

= = = = =

uDρ µ

(USCS units)

(2.32)

Reynolds number, dimensionless average velocity of gas in pipe, ft/s inside diameter of pipe, ft gas density, lb/ft3 gas viscosity, lb/ft-s

The above equation for the Reynolds number is in USCS units. The corresponding equation for the Reynolds number in SI units is as follows: Re = where Re u D r m

uDρ µ

(SI units)

(2.33)

= Reynolds number, dimensionless = average velocity of gas in pipe, m/s = inside diameter of pipe, m = gas density, kg/m3 = gas viscosity, kg/m-s

In gas pipeline hydraulics, using customary units, a more suitable equation for the Reynolds number is as follows: P GQ Re = 0.0004778 b Tb µ D where Pb = base pressure, psia Tb = base temperature, °R (460 + °F) G = speciﬁc gravity of gas (air = 1.0) Q = gas ﬂow rate, standard ft3/day (SCFD) D = pipe inside diameter, in. m = viscosity of gas, lb/ft-s

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(2.34)

2785_C002.fm Page 44 Friday, April 1, 2005 3:23 PM

44

GAS PIPELINE HYDRAULICS

In SI units, the Reynolds number is P GQ Re = 0.5134 b Tb µ D

(SI units)

(2.35)

where Pb = base pressure, kPa Tb = base temperature, °K (273 + °C) G = speciﬁc gravity of gas (air = 1.0) Q = gas ﬂow rate, m3/day (standard conditions) D = pipe inside diameter, mm m = viscosity of gas, Poise Laminar ﬂow occurs in a pipeline when the Reynolds number is below a value of approximately 2000. Turbulent ﬂow occurs when the Reynolds number is greater than 4000. For Reynolds numbers between 2000 and 4000, the ﬂow is undeﬁned and is referred to as critical ﬂow. Thus, For laminar ﬂow, Re ≤ 2000 For turbulent ﬂow, Re > 4000 For critical ﬂow, Re > 2000 and Re ≤ 4000

Most natural gas pipelines operate in the turbulent ﬂow region. Therefore, the Reynolds number is greater than 4000. Turbulent ﬂow is further divided into three regions known as smooth pipe ﬂow, fully rough pipe ﬂow, and transition ﬂow. We will discuss these ﬂow regions in more detail in the subsequent sections of this chapter. Example 3 A natural gas pipeline, NPS 20 with 0.500 in. wall thickness, transports 100 MMSCFD. The speciﬁc gravity of gas is 0.6 and viscosity is 0.000008 lb/ft-s. Calculate the value of the Reynolds number of ﬂow. Assume the base temperature and base pressure are 60°F and 14.7 psia, respectively. Solution Pipe inside diameter = 20 – 2 × 0.5 = 19.0 in. The base temperature = 60 + 460 = 520 °R Using Equation 2.34, we get 14.7 0.6 × 100 × 10 6 Re = 0.0004778 = 5,331,726 520 0.000008 × 19 Since Re is greater than 4000, the ﬂow is in the turbulent region.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 45 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

45

Example 4 A natural gas pipeline, DN 500 with 12 mm wall thickness, transports 3 Mm3/day. The speciﬁc gravity of gas is 0.6 and viscosity is 0.00012 Poise. Calculate the value of the Reynolds number. Assume the base temperature and base pressure are 15°C and 101 kPa, respectively. Solution Pipe inside diameter = 500 – 2 × 12 = 476 mm The base temperature = 15 + 273 = 288 K Using Equation 2.35, we get 101 0.6 × 3 × 10 6 Re = 0.5134 = 5,673,735 15 + 273 0.00012 × 476 Since Re is greater than 4000, the ﬂow is in the turbulent region.

2.9 FRICTION FACTOR In order to calculate the pressure drop in a pipeline at a given ﬂow rate, we must ﬁrst understand the concept of friction factor. The term friction factor is a dimensionless parameter that depends upon the Reynolds number of ﬂow. In engineering literature, we ﬁnd two different friction factors mentioned. The Darcy friction factor is more common and will be used throughout this book. Another friction factor known as the Fanning friction factor is preferred by some engineers. The Fanning friction factor is numerically equal to one-fourth the Darcy friction factor as below. ff =

fd 4

(2.36)

where ff = Fanning friction factor fd = Darcy friction factor To avoid confusion, in subsequent discussions, the Darcy friction factor is used and will be represented by the symbol f. For laminar ﬂow, the friction factor is inversely proportional to the Reynolds number, as indicated below. f=

64 Re

(2.37)

For turbulent ﬂow, the friction factor is a function of the Reynolds number, pipe inside diameter, and internal roughness of the pipe. Many empirical relationships for

Copyright 2005 by Taylor & Francis Group, LLC

Laminar Critical Flow Zone Transition Zone

Complete turbulence, rough pipes 0.05 0.04

0.07

4 / Re

0.01 0.008 0.006 0.004

0.025

0.002

0.02

0.001 0.0008 0.0006 0.0004

Sm

0.015

oo

th

0.0002 0.0001

pip

es

103

2 3 4 56 8104 ×103

2 3 4 56 8105 ×104

2 3 4 56 8106 ×105

Reynolds number Re = VD n Moody diagram.

2 3 4 56 8107 ×106

0.000,01 2 3 4 56 8108 e e D = 0. 000D = 0.00 0,0 ,00 05 1

GAS PIPELINE HYDRAULICS

0.000,05

0.01 0.009 0.008

Figure 2.3

Relative roughness e D

f=6

0.03 Friction factor f

0.02 0.015

flow

0.04

0.03

inar

0.05

Lam

0.06

2785_C002.fm Page 46 Friday, April 1, 2005 3:23 PM

46

Copyright 2005 by Taylor & Francis Group, LLC

0.10 0.09 0.08

2785_C002.fm Page 47 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

47

calculating f have been put forth by researchers. The more popular correlations include the Colebrook-White and AGA equations. Before we discuss the equations for calculating the friction factor in turbulent ﬂow, it is appropriate to analyze the turbulent ﬂow regime. Turbulent ﬂow in pipes (Re > 4000) is subdivided into three separate regions as follows: 1. Turbulent ﬂow in smooth pipes 2. Turbulent ﬂow in fully rough pipes 3. Transition ﬂow between smooth pipes and rough pipes

For turbulent ﬂow in smooth pipes, the friction factor f depends only on the Reynolds number. For fully rough pipes, f depends more on the pipe internal roughness and less on the Reynolds number. In the transition zone between smooth pipe ﬂow and ﬂow in fully rough pipes, f depends on the pipe roughness, pipe inside diameter, and the Reynolds number. The various ﬂow regimes are depicted in the Moody diagram, shown in Figure 2.3. The Moody diagram is a graphic plot of the variation of the friction factor with the Reynolds number for various values of relative pipe roughness. The latter term is simply a dimensionless parameter obtained by dividing the absolute (or internal) pipe roughness by the pipe inside diameter as follows: Relative roughness =

e D

(2.38)

where e = absolute or internal roughness of pipe, in. D = pipe inside diameter, in. The terms absolute pipe roughness and internal pipe roughness are equivalent. Generally, the internal pipe roughness is expressed in microinches (one-millionth of an inch). For example, an internal roughness of 0.0006 in. is referred to as 600 microinches or 600 µin. If the pipe inside diameter is 15.5 in., the relative roughness is, in this case, Relative roughness =

0.0006 = 0.0000387 = 3.87 × 10−5 15.5

For example, from the Moody diagram in Figure 2.3, for Re = 10 million and e/D = 0.0001, we ﬁnd that f = 0.012.

2.10 COLEBROOK-WHITE EQUATION The Colebrook-White equation, sometimes referred to simply as the Colebrook equation, is a relationship between the friction factor and the Reynolds number, pipe roughness, and inside diameter of pipe. The following form of the Colebrook

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 48 Friday, April 1, 2005 3:23 PM

48

GAS PIPELINE HYDRAULICS

equation is used to calculate the friction factor in gas pipelines in turbulent ﬂow. e 2.51 = −2 Log10 + f 3.7 D Re f

1

where f D e Re

= = = =

for Re > 4000

(2.39)

friction factor, dimensionless pipe inside diameter, in. absolute pipe roughness, in. Reynolds number of ﬂow, dimensionless

Since Re and f are dimensionless, as long as consistent units are used for both e and D, the Colebrook equation is the same regardless of the units employed. Therefore, in SI units, Equation 2.39 is used with e and D expressed in mm. It can be seen from Equation 2.39 that in order to calculate the friction factor f, we must use a trial-and-error approach. It is an implicit equation in f, since f appears on both sides of the equation. We ﬁrst assume a value of f (such as 0.01) and substitute it in the right-hand side of the equation. This will yield a second approximation for f, which can then be used to calculate a better value of f, and so on. Generally 3 to 4 iterations are sufﬁcient to converge on a reasonably good value of the friction factor. It can be seen from the Colebrook Equation 2.39, for turbulent ﬂow in smooth pipes, the ﬁrst term within the square brackets is negligible compared to the second term, since pipe roughness e is very small. Therefore, for smooth pipe ﬂow, the friction factor equation reduces to 2.51 = −2 Log10 , f Re f

1

for turbulent ﬂow in smooth pipes

(2.40)

Similarly, for turbulent ﬂow in fully rough pipes, with Re being a large number, f depends mostly on the roughness e and, therefore, the friction factor equation reduces to e = −2 Log10 , 3.7 D f

1

for turbulent ﬂow in fully rough pipes

(2.41)

Table 2.1 lists typical values of pipe internal roughness used to calculate the friction factor. As an example, if Re = 100 million or larger and e/D = 0.0002, the friction factor from Equation 2.41 is 0.0002 = −2 Log10 3.7 f

1

or f = 0.0137, which correlates well with the friction factor obtained from the Moody diagram in Figure 2.3.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 49 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

49

Table 2.1 Pipe Internal Roughness Pipe Material Riveted steel Commercial steel/welded steel Cast iron Galvanized iron Asphalted cast iron Wrought iron PVC, drawn tubing, glass Concrete

Roughness, in.

Roughness, mm

0.0354 to 0.354 0.0018 0.0102 0.0059 0.0047 0.0018 0.000059 0.0118 to 0.118

0.9 to 9.0 0.045 0.26 0.15 0.12 0.045 0.0015 0.3 to 3.0

Example 5 A natural gas pipeline, NPS 20 with 0.500 in. wall thickness, transports 200 MMSCFD. The speciﬁc gravity of gas is 0.6 and viscosity is 0.000008 lb/ft-s. Calculate the friction factor using the Colebrook equation. Assume absolute pipe roughness = 600 µ in. The base temperature and base pressure are 60°F and 14.7 psia, respectively. Solution Pipe inside diameter = 20 – 2 × 0.5 = 19.0 in. Absolute pipe roughness = 600 µ in. = 0.0006 in. First, we calculate the Reynolds number using Equation 2.34: 14.7 0.6 × 200 × 10 6 Re = 0.0004778 = 10,663,452 60 + 460 0.000008 × 19 Using Equation 2.39, 0.0006 2.51 = −2 Log10 + f 3.7 × 19 10, 663, 452 f

1

This equation will be solved by successive iteration. Assume f = 0.01 initially; substituting above, we get a better approximation as f = 0.0101. Repeating the iteration, we get the ﬁnal value as f = 0.0101. Therefore, the friction factor is 0.0101. Example 6 A natural gas pipeline, DN 500 with 12 mm wall thickness, transports 6 Mm3/day. The speciﬁc gravity of gas is 0.6 and viscosity is 0.00012 Poise. Calculate the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 50 Friday, April 1, 2005 3:23 PM

50

GAS PIPELINE HYDRAULICS

friction factor using the Colebrook equation. Assume absolute pipe roughness = 0.03 mm and assume the base temperature and base pressure are 15°C and 101 kPa, respectively. Solution Pipe inside diameter = 500 – 2 × 12 = 476 mm First, we calculate the Reynolds number using Equation 2.35: 101 0.6 × 6 × 10 6 Re = 0.5134 = 11,347,470 15 + 273 0.00012 × 476 Using Equation 2.39, the friction factor is 0.030 2.51 = −2 Log10 + . 3 7 476 × 11, 347, 470 f f

1

This equation will be solved by successive iteration. Assume f = 0.01 initially; substituting above, we get a better approximation as f = 0.0112. Repeating the iteration, we get the ﬁnal value as f = 0.0112. Therefore, the friction factor is 0.0112.

2.11 TRANSMISSION FACTOR The transmission factor F is considered the opposite of the friction factor f. Whereas the friction factor indicates how difﬁcult it is to move a certain quantity of gas through a pipeline, the transmission factor is a direct measure of how much gas can be transported through the pipeline. As the friction factor increases, the transmission factor decreases and, therefore, the gas ﬂow rate also decreases. Conversely, the higher the transmission factor, the lower the friction factor and, therefore, the higher the ﬂow rate will be. The transmission factor F is related to the friction factor f as follows: F=

2

f =

4 F2

(2.42)

f

Therefore,

Copyright 2005 by Taylor & Francis Group, LLC

(2.43)

2785_C002.fm Page 51 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

51

where f = friction factor F = transmission factor It must be noted that the friction factor f in the above equation is the Darcy friction factor. Since some engineers prefer to use the Fanning friction factor, the relationship between the transmission factor F and the Fanning friction factor is given below for reference. F=

1

(2.44)

ff

where ff is the Fanning friction factor. For example, if the Darcy friction factor is 0.025, the transmission factor is, using Equation 2.42, F=

2 0.025

= 12.65

The Fanning friction factor in this case will be 0.025 = 0.00625. Therefore, the 4 1 transmission factor using Equation 2.44 is F = 0.00625 = 12.65, which is the same as calculated using the Darcy friction factor. Thus, it must be noted that there is only one transmission factor, whereas there are two different friction factors. Having deﬁned a transmission factor, we can rewrite the Colebrook Equation 2.39 in terms of the transmission factor using Equation 2.42 as follows: e 1.255F F = − 4 Log10 + Re 3.7 D

(2.45)

Since Re and F are dimensionless, as long as consistent units are used for both e and D, the transmission factor equation is the same regardless of the units employed. Therefore, in SI units, Equation 2.45 is used with e and D expressed in mm. Similar to the calculation of the friction factor f from Equation 2.39, to calculate the transmission factor F from Equation 2.45, an iterative approach must be used. This will be illustrated using an example. Example 7 For a gas pipeline, ﬂowing 100 MMSCFD gas of speciﬁc gravity 0.6 and viscosity of 0.000008 lb/ft-s, calculate the friction factor and transmission factor considering an NPS 20 pipeline, 0.500-in. wall thickness, and an internal roughness of 600 microinches. Assume the base temperature and base pressure are 60°F and 14.7 psia, respectively. If the ﬂow rate increases by 50%, what is the impact on the friction factor and transmission factor?

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 52 Friday, April 1, 2005 3:23 PM

52

GAS PIPELINE HYDRAULICS

Solution The base temperature = 60 + 460 = 520°R Pipe inside diameter = 20 – 2 × 0.500 = 19.0 in. Using Equation 2.34, we calculate the Reynolds number as 14.7 0.6 × 100 × 10 6 Re = 0.0004778 = 5,331,726 520 0.000008 × 19 The relative roughness =

600 × 10 −6 = 0.0000316 19

Using Equation 2.39, the friction factor is 0.0000316 2.51 = −2 Log10 + 3.7 5, 331, 726 f f

1

Solving by successive iteration, we get f = 0.0105 Therefore, the transmission factor F is found from Equation 2.42 as follows: F=

2 0.0105

= 19.53

It must be noted that the friction factor calculated above is the Darcy friction factor. The corresponding Fanning friction factor will be one-fourth the calculated value. When ﬂow rate is increased by 50%, the Reynolds number becomes, by proportion, Re = 1.5 × 5,331,726 = 7,997,589 The new friction factor from Equation 2.39 is 0.0000316 2.51 = −2 Log10 + 3.7 7, 997, 589 f f

1

Solving for f by successive iteration, we get f = 0.0103 The corresponding transmission factor is F=

Copyright 2005 by Taylor & Francis Group, LLC

2 0.0103

= 19.74

2785_C002.fm Page 53 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

53

Compared to the previous values of 0.0105 for the friction factor and 19.53 for the transmission factor, we see the following changes:

Decrease in friction factor =

0.0105 − 0.0103 = 0.019 or 1.9% 0.0105

Increase in transmission factor =

19.74 − 19.53 = 0.0108 or 1.08% 19.53

Thus, increasing the ﬂow rate by 50% reduces the friction factor by 1.9% and increases the transmission factor by 1.08%. Example 8 For a gas pipeline, ﬂowing 3 Mm3/day gas of speciﬁc gravity 0.6 and viscosity of 0.000119 Poise, calculate the friction factor and transmission factor considering a DN 400 pipeline, 10 mm wall thickness, and an internal roughness of 0.02 mm. The base temperature and base pressure are 15°C and 101 kPa, respectively. If the ﬂow rate is doubled, what is the impact on the friction factor and transmission factor? Solution The base temperature = 15 + 273 = 288 K Pipe inside diameter = 400 – 2 × 10 = 380 mm Using Equation 2.35, we calculate the Reynolds number as 101 0.6 × 3 × 10 6 Re = 0.5134 = 7,166,823 288 0.000119 × 380 The relative roughness =

0.02 = 0.0000526 380

Using Equation 2.39, the friction factor is 0.0000526 2.51 = −2 Log10 + . 3 7 7,166, 823 f f

1

Solving by iteration, we get f = 0.0111 Therefore, the transmission factor F is found from Equation 2.42 as follows: F=

Copyright 2005 by Taylor & Francis Group, LLC

2 0.0111

= 18.98

2785_C002.fm Page 54 Friday, April 1, 2005 3:23 PM

54

GAS PIPELINE HYDRAULICS

It must be noted that the friction factor calculated above is the Darcy friction factor. The corresponding Fanning friction factor will be one-fourth the calculated value. When the ﬂow rate is doubled, the Reynolds number becomes Re = 2 × 7,166,823 = 14,333,646 The new value of the friction factor from Equation 2.39 is 0.0000526 2.51 = −2 Log10 + 3.7 14, 333, 646 f f

1

Solving for f by successive iteration, we get f = 0.0109 and the transmission factor is F=

2 0.0109

= 19.16

Therefore, doubling the ﬂow rate increases the transmission factor and decreases the friction factor as follows: Decrease in friction factor =

0.0111 − 0.0109 = 0.018 or 1.8% 0.0111

Increase in transmission factor =

19.16 − 18.98 = 0.0095 or 0.95% 18.98

2.12 MODIFIED COLEBROOK-WHITE EQUATION The Colebrook-White equation discussed in the preceding section has been in use for many years in both liquid ﬂow and gas ﬂow. The U.S. Bureau of Mines, in 1956, published a report that introduced a modiﬁed form of the Colebrook-White equation. The modiﬁcation results in a higher friction factor and, hence, a smaller value of the transmission factor. Because of this, a conservative value of ﬂow rate is obtained due to the higher friction and pressure drop. The modiﬁed version of the ColebrookWhite equation for turbulent ﬂow is as follows: e 2.825 = −2 Log10 + f 3.7 D Re f

1

(2.46)

Rewriting Equation 2.46 in terms of the transmission factor, we get the following version of the modiﬁed Colebrook-White equation: e 1.4125F F = −4 Log10 + Re 3.7 D

Copyright 2005 by Taylor & Francis Group, LLC

(USCS and SI units)

(2.47)

2785_C002.fm Page 55 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

55

Since Re, f, and F are dimensionless, as long as consistent units are used for both e and D, the modiﬁed Colebrook equation is the same, regardless of the units employed. Therefore, in SI units, Equation 2.46 and Equation 2.47 are used with e and D expressed in mm. Upon comparing Equation 2.39 with Equation 2.46, it is seen that the difference between the Colebrook equation and the modiﬁed Colebrook equation lies in the second constant term within the square brackets. The constant 2.51 in Equation 2.39 is replaced with the constant 2.825 in Equation 2.46. Similarly, in the transmission factor equations, the modiﬁed equation has 1.4125 instead of 1.255 in the original Colebrook-White equation. Many commercial hydraulic simulation programs list both Colebrook-White equations. Some use only the original Colebrook-White equation. Example 9 For a gas pipeline, ﬂowing 100 MMSCFD gas of speciﬁc gravity 0.6 and viscosity of 0.000008 lb/ft-s, calculate, using the modiﬁed Colebrook-White equation, the friction factor and transmission factor assuming an NPS 20 pipeline, 0.500 in. wall thickness, and an internal roughness of 600 µ in. The base temperature and base pressure are 60°F and 14.7 psia, respectively. How do these numbers compare with those calculated, using the original Colebrook equation? Solution The base temperature = 60 + 460 = 520°R Pipe inside diameter = 20 – 2 × 0.500 = 19.0 in. Using Equation 2.34, we calculate the Reynolds number as 14.7 0.6 × 100 × 10 6 Re = 0.0004778 = 5,331,726 520 0.000008 × 19 The relative roughness is e 600 × 10 6 = = 3.16 × 10 −5 D 19 From Equation 2.46, the friction factor using the modiﬁed Colebrook equation is 0.0000316 2.825 = −2 Log10 + 3.7 5, 331, 726 f f

1

Solving by successive iteration, we get f = 0.0106

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 56 Friday, April 1, 2005 3:23 PM

56

GAS PIPELINE HYDRAULICS

Therefore, the transmission factor F is found from Equation 2.42 as follows: F=

2 0.0106

= 19.43

By comparing these results with the friction factor and the transmission factor calculated in Example 7 using the unmodiﬁed Colebrook equation, it can be seen that the modiﬁed friction factor is approximately 0.95% higher than that calculated using the original Colebrook-White equation, whereas the transmission factor is approximately 0.51% lower than that calculated using the original Colebrook-White equation. Example 10 A gas pipeline, NPS 20 with 0.500 in. wall thickness, ﬂows 200 MMSCFD gas of speciﬁc gravity 0.6 and viscosity of 0.000008 lb/ft-s. Using the modiﬁed ColebrookWhite equation, calculate the pressure drop in a 50 mi segment of pipe, based on an upstream pressure of 1000 psig. Assume an internal pipe roughness of 600 µ in. and the base temperature and base pressure of 60°F and 14.73 psia, respectively. Neglect elevation effects and use 60°F for gas ﬂowing temperature and compressibility factor Z = 0.88. Solution Inside diameter of pipe = 20 – 2 × 0.5 = 19.0 in. The base temperature = 60 + 460 = 520 °R Gas ﬂow temperature = 60 + 460 = 520 °R First, we calculate the Reynolds number using Equation 2.34. 14.73 0.6 × 200 × 10 6 Re = 0.0004778 = 10,685,214 520 0.000008 × 19 The transmission factor F is calculated from Equation 2.47 as follows: 600 × 10 −6 1.4125F + F = −4 Log10 10, 685, 214 3.7 × 19 Solving for F by successive iteration, F = 19.81 Next, using General Flow Equation 2.4, we calculate the downstream pressure P2 as follows: 60 + 460 1014.732 − P2 2 200 × 10 6 = 38.77 × 19.81 14.73 0.6 × 520 × 50 × 0.88

Copyright 2005 by Taylor & Francis Group, LLC

0.5

× 192.5

2785_C002.fm Page 57 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

57

Solving for P2, we get P2 = 853.23 psia = 838.5 psig Therefore, the pressure drop = 1014.73 – 853.23 = 161.5 psi.

2.13 AMERICAN GAS ASSOCIATION (AGA) EQUATION In 1964 and 1965, the American Gas Association (AGA) published a report on how to calculate the transmission factor for gas pipelines to be used in the General Flow equation. This is sometimes referred to as the AGA NB-13 method. Using the method outlined in this report, the transmission factor F is calculated using two different equations. First, F is calculated for the rough pipe law (referred to as the fully turbulent zone). Next, F is calculated based on the smooth pipe law (referred to as the partially turbulent zone). Finally, the smaller of the two values of the transmission factor is used in the General Flow Equation 2.4 for pressure drop calculation. Even though the AGA method uses the transmission factor F instead of the friction factor f, we can still calculate the friction factor using the relationship shown in Equation 2.42. For the fully turbulent zone, AGA recommends using the following formula for F, based on relative roughness e/D and independent of the Reynolds number: 3.7 D F = 4 Log10 e

(2.48)

Equation 2.48 is also known as the Von Karman rough pipe ﬂow equation. For the partially turbulent zone, F is calculated from the following equations using the Reynolds number, a parameter Df known as the pipe drag factor, and the Von Karman smooth pipe transmission factor Ft: Re F = 4 D f Log10 1.4125Ft

(2.49)

Re Ft = 4 Log10 − 0.6 Ft

(2.50)

and

where Ft = Von Karman smooth pipe transmission factor Df = pipe drag factor that depends on the Bend Index (BI) of the pipe The pipe drag factor Df is a parameter that takes into account the number of bends, ﬁttings, etc. Its value ranges from 0.90 to 0.99. The Bend index is the sum of all the angles and bends in the pipe segment, divided by the total length of the pipe section under consideration. BI =

Copyright 2005 by Taylor & Francis Group, LLC

total degrees of all bends in pipe sectionn total length of pipe section

(2.51)

2785_C002.fm Page 58 Friday, April 1, 2005 3:23 PM

58

GAS PIPELINE HYDRAULICS

Table 2.2 Bend Index and Drag Factor

Bare steel Plastic lined Pig burnished Sand blasted

Extremely Low 5° to 10°

Bend Index Average 60° to 80°

Extremely High 200° to 300°

0.975–0.973 0.979–0.976 0.982–0.980 0.985–0.983

0.960–0.956 0.964–0.960 0.968–0.965 0.976–0.970

0.930–0.900 0.936–0.910 0.944–0.920 0.951–0.930

Note: The drag factors above are based on 40-ft joints of pipelines and mainline valves at 10-mile spacing.

The value of Df is generally chosen from Table 2.2. For further discussion on the bend index and drag factor, the reader is referred to Steady Flow in Gas Pipelines listed in the Reference section. Example 11 Using the AGA method, calculate the transmission factor and friction factor for gas ﬂow in an NPS 20 pipeline with 0.500 in. wall thickness. The ﬂow rate is 200 MMSCFD, gas gravity = 0.6, and viscosity = 0.000008 lb/ft-sec. The absolute pipe roughness is 700 µ in. Assume a bend index of 60°, base pressure of 14.73 psia, and base temperature of 60°F. Solution Inside diameter of pipe = 20 – 2 × 0.5 = 19.0 in. The base temperature = 60 + 460 = 520°R We will ﬁrst calculate the Reynolds number using Equation 2.34. Re =

0.0004778 × 200 × 10 6 × 0.6 × 14.73 = 10,685,214 19 × 0.000008 × 520

Next, calculate the two transmission factors. The fully turbulent transmission factor, using Equation 2.48, is 3.7 × 19 F = 4 Log10 = 20.01 0.0007 For the smooth pipe zone, using Equation 2.50, the Von Karman transmission factor is 10, 685, 214 Ft = 4 Log10 − 0.6 Ft

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 59 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

59

Solving this equation by trial and error, we get Ft = 22.13. From Table 2.2, for a bend index of 60°, the drag factor Df is 0.96. Therefore, for the partially turbulent ﬂow zone, using Equation 2.49, the transmission factor is 10, 685, 214 = 21.25 F = 4 × 0.96 Log10 1.4125 × 22.13 From the above two values of F, using the smaller number, we get the AGA transmission factor as F = 20.01 Therefore, the corresponding friction factor f is found from Equation 2.42 as 2 f

= 20.01

or f = 0.0100 Example 12 Using the AGA method, calculate the transmission factor and friction factor for gas ﬂow in a DN 500 pipeline with 12 mm wall thickness. The ﬂow rate is 6 Mm3/day, gas gravity = 0.6, and viscosity = 0.00012 Poise. The absolute pipe roughness is 0.02 mm. Assume a bend index of 60°, base pressure of 101 kPa, and base temperature of 15°C. For a 60 km pipe length, calculate the upstream pressure needed to hold a downstream pressure of 5 MPa (absolute). Assume ﬂow temperature = 20°C and compressibility factor Z = 0.85. Neglect elevation effects. Solution Inside diameter of pipe = 500 – 2 × 12 = 476 mm The base temperature = 15 + 273 = 288 K Gas ﬂowing temperature = 20 + 273 = 293 K We ﬁrst calculate the Reynolds number from Equation 2.35. 101 0.6 × 6 × 10 6 = 11,347,470 Re = 0.5134 288 0.00012 × 476

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 60 Friday, April 1, 2005 3:23 PM

60

GAS PIPELINE HYDRAULICS

Next, calculate the two transmission factors as follows: The fully turbulent transmission factor, using Equation 2.48, is 3.7 × 476 F = 4 Log10 = 19.78 0.02 For the smooth pipe zone, using Equation 2.50, the Von Karman transmission factor is 11, 347, 470 Ft = 4 Log10 − 0.6 Ft Solving by successive iteration, we get Ft = 22.23 From Table 2.2, for a bend index of 60°, the drag factor is 0.96. Therefore, for the partially turbulent ﬂow zone, using Equation 2.49, the transmission factor is 11, 347, 470 = 21.34 F = 4 × 0.96 Log10 1.4125 × 22.23 Using the smaller of the two values of F, the AGA transmission factor is F = 19.78 Therefore, the corresponding friction factor is found from Equation 2.42 as 2 f

= 19.78

or f = 0.0102 Using the General Flow Equation 2.8, we calculate the upstream pressure P1 as follows: 288 P12 − 5000 2 6 × 10 6 = 5.747 × 10 −4 × 19.78 × 101 0.6 × 293 × 60 × 0.85 Solving for P1, we get P1 = 6130 kPa = 6.13 MPa

Copyright 2005 by Taylor & Francis Group, LLC

0.5

× 4762.5

2785_C002.fm Page 61 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

61

2.14 WEYMOUTH EQUATION The Weymouth equation is used for high pressure, high ﬂow rate, and large diameter gas gathering systems. This formula directly calculates the ﬂow rate through a pipeline for given values of gas gravity, compressibility, inlet and outlet pressures, pipe diameter, and length. In USCS units, the Weymouth equation is stated as follows: T P 2 − e s P22 Q = 433.5E b 1 Pb GT f Le Z where Q = E = Pb = Tb = P1 = P2 = G = Tf = Le = Z = D =

0.5

D 2.667

(2.52)

volume ﬂow rate, standard ft3/day (SCFD) pipeline efﬁciency, a decimal value less than or equal to 1.0 base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂow temperature, °R (460 + °F) equivalent length of pipe segment, mi gas compressibility factor, dimensionless pipe inside diameter, in.

where the equivalent length Le and s were deﬁned earlier in Equation 2.9 and Equation 2.10. By comparing the Weymouth equation with the General Flow equation, we can isolate an equivalent transmission factor as follows: The Weymouth transmission factor in USCS units is F = 11.18(D)1/6

(2.53)

In SI units, the Weymouth equation is as follows: T P 2 − e s P22 Q = 3.7435 × 10 E b 1 Pb GT f Le Z −3

where Q = Tb = Pb = Tf = P1 = P2 = Le =

gas ﬂow rate, standard m3/day base temperature, K (273 + °C) base pressure, kPa average gas ﬂow temperature, K (273 + °C) upstream pressure, kPa downstream pressure, kPa equivalent length of pipe segment, km

Other symbols are as deﬁned previously.

Copyright 2005 by Taylor & Francis Group, LLC

0.5

D 2.667

(2.54)

2785_C002.fm Page 62 Friday, April 1, 2005 3:23 PM

62

GAS PIPELINE HYDRAULICS

The Weymouth transmission factor in SI units is F = 6.521(D)1/6

(2.53a)

You will notice that a pipeline efﬁciency factor, E, is used in the Weymouth equation so we can compare the throughput performance of a pipeline using the General Flow equation that does not include an efﬁciency factor. Example 13 Calculate the ﬂow rate using the Weymouth equation in a gas pipeline system, 15 miles long, NPS 12 pipe with 0.250 in. wall thickness, at an efﬁciency of 0.95. The upstream pressure is 1200 psia, and the delivery pressure required at the end of the pipe segment is 750 psia. Use gas gravity = 0.59 and viscosity = 0.000008 lb/ft-sec. The ﬂowing temperature of gas = 75°F, base pressure = 14.7 psia, and base temperature = 60°F. Assume compressibility factor to be 0.94. Neglect elevation difference along the pipe. How does this compare with the ﬂow rate calculated using the General Flow equation with the Colebrook friction factor? Assume a pipe roughness of 700 µ in. Solution Using Equation 2.52, we get the ﬂow rate for the Weymouth equation as follows: 60 + 460 1200 2 − 750 2 Q = 433.5 × 0.95 14.7 0.59 × (75 + 460) × 15 × 0.94

0.5

× 12.252.667

Q = 163,255,858 SCFD or Q = 163.26 MMSCFD Next, we will calculate the Reynolds number using Equation 2.34. Re =

0.0004778 × Q × 0.59 × 14.7 12.25 × 0.000008 × 520

where Q is the ﬂow rate in SCFD. Simplifying, we get Re = 0.0813 Q. Since Q is unknown, we will ﬁrst assume a transmission factor F = 20 and calculate the ﬂow rate from the General Flow Equation 2.4. 520 1200 2 − 750 2 Q = 38.77 × 20 14.7 0.59 × 535 × 15 × 0.94

Copyright 2005 by Taylor & Francis Group, LLC

0.5

× 12.252.5 = 202,284,747 SCFD

2785_C002.fm Page 63 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

63

or Q = 202.28 MMSCFD Next, we will calculate the Reynolds number and the transmission factor based on this ﬂow rate as Re = 0.0813 × 202,284,747 = 16.45 million and, using Equation 2.45, 700 × 10 −6 1.255F + F = − 4 Log10 3 7 × 12 25 . . 16.45 × 10 6 Solving for F, we get F = 19.09 Using this value, the revised ﬂow rate is found by proportion as Q = 202.28 ×

19.09 = 193.08 MMSCFD 20

Repeating the calculation of Re and F, we get Re = 16.45 ×

193.08 = 15.7 million 202.28

and 700 × 10 −6 1.255F + F = − 4 Log10 6 3.7 × 12.25 15.7 × 10 Therefore, F = 19.08. This is fairly close to the previous value of F = 19.09; therefore, we will use this value and calculate the ﬂow rate as Q = 202.28 ×

19.08 = 192.98 MMSCFD 20

Comparing this result using the General Flow equation with that calculated using the Weymouth equation, we see that the latter equation is quite conservative. Example 14 A natural gas transmission line transports 30 million m3/day of gas from a processing plant to a compressor station site 100 km away. The pipeline can be assumed to be along a ﬂat terrain. Calculate the minimum pipe diameter required such that the maximum pipe operating pressure is limited to 8500 kPa. The delivery pressure desired at the end of the pipeline is a minimum of 5500 kPa. Assume a pipeline efﬁciency of 0.95. The gas gravity is 0.65, and the gas temperature is 18°C. Use the Weymouth equation, considering a base temperature = 15°C and base pressure 101 kPa. The gas compressibility factor Z = 0.92.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 64 Friday, April 1, 2005 3:23 PM

64

GAS PIPELINE HYDRAULICS

Solution The base temperature = 15 + 273 = 288 K The gas ﬂowing temperature = 18 + 273 = 291 K We will assume that given pressures are absolute values. Upstream pressure = 8500 kPa (absolute) Downstream pressure = 5500 kPa (absolute) Using the Weymouth Equation 2.52 and substituting given values, we get 288 8500 2 − 5500 2 30 × 10 6 = 3.7435 × 10 −3 × 0.95 × 101 0.65 × 291 × 100 × 0.92

0.5

× D 2.667

Solving for diameter, D, we get D = 826.1 mm Therefore, the minimum diameter required will be DN 850 with 10 mm wall thickness.

2.15 PANHANDLE A EQUATION The Panhandle A Equation was developed for use in natural gas pipelines, incorporating an efﬁciency factor for Reynolds numbers in the range of 5 to 11 million. In this equation, the pipe roughness is not used. The general form of the Panhandle A equation is expressed in USCS units as follows: T Q = 435.87 E b P b

where Q = E = Pb = Tb = P1 = P2 = G = Tf =

1.0788

P 2 − es P 2 1 2 0.8539 T f Le Z G

0.5394

D 2.6182

volume ﬂow rate, standard ft3/day (SCFD) pipeline efﬁciency, a decimal value less than 1.0 base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂow temperature, °R (460 + °F)

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units) (2.55)

2785_C002.fm Page 65 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

65

Le = equivalent length of pipe segment, mi Z = gas compressibility factor, dimensionless D = pipe inside diameter, in. Other symbols are as deﬁned previously. In SI units, the Panhandle A equation is T Q = 4.5965 × 10 E b P −3

b

where Q = E = Tb = Pb = Tf = P1 = P2 = Le =

1.0788

P 2 − es P 2 1 2 0.8539 T f Le Z G

0.5394

D 2.6182

(SI units)

(2.56)

gas ﬂow rate, standard m3/day pipeline efﬁciency, a decimal value less than 1.0 base temperature, K (273 + °C) base pressure, kPa average gas ﬂow temperature, K (273 + °C) upstream pressure, kPa (absolute) downstream pressure, kPa (absolute) equivalent length of pipe segment, km

Other symbols are as deﬁned previously. Due to the exponents involved in this equation, all pressures must be in kPa. By comparing the Panhandle A equation with the General Flow equation, we can calculate an equivalent transmission factor in USCS units as follows: QG F = 7.2111E D

0.07305

(USCS)

(2.57)

(SI)

(2.58)

and in SI units, it is QG F = 11.85E D

0.07305

Sometimes the transmission factor is used to compare the results of calculations using the General Flow equation and the Panhandle A equation. Example 15 Using the Panhandle A equation, calculate the outlet pressure in a natural gas pipeline, NPS 16 with 0.250 in. wall thickness, 15 miles long. The gas ﬂow rate is 100 MMSCFD at an inlet pressure of 1000 psia. The gas gravity = 0.6 and viscosity = 0.000008 lb/ft-sec. The average gas temperature is 80°F. Assume base pressure = 14.73 psia and base temperature = 60°F. For compressibility factor Z, use the CNGA method. Assume pipeline efﬁciency of 0.92.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 66 Friday, April 1, 2005 3:23 PM

66

GAS PIPELINE HYDRAULICS

Solution The average pressure, Pavg, needs to be calculated before the compressibility factor Z can be determined. Since the inlet pressure P1 = 1,000 psia, and the outlet pressure P2 is unknown, we will have to assume a value of P2 (such as 800 psia) and calculate Pavg and then calculate the value of Z. Once Z is known, from the Panhandle A equation we can calculate the outlet pressure P2. Using this value of P2, a better approximation for Z is calculated from a new Pavg. This process is repeated until successive values of P2 are within allowable limits, such as 0.5 psia. Assume P2 = 800 psia. The average pressure from Equation 2.14 is Pavg =

2 1000 × 800 = 903.7 psia 1000 + 800 − 3 1000 + 800

Next, we calculate the compressibility factor Z using the CNGA method. From Equation 1.34, Z=

1 5 1.785× 0.6 1 + (903.7−14.73)×3.444×310.825×(10) (80+ 460)

or Z = 0.8869 From Panhandle A Equation 2.55, substituting given values, neglecting elevations, we get 60 + 460 100 × 10 6 = 435.87 × 0.92 14.73

1.0788

1000 2 − P22 0.8539 (540 × 15 × 0.8869) (0.6)

5394 0.5

(15.5)2.6182

Solving for P2, we get P2 = 968.02 psia Since this is different from the assumed value of P2 = 800, we recalculate the average pressure and Z using P2 = 968.02 psia The revised average pressure is

Pavg =

Copyright 2005 by Taylor & Francis Group, LLC

2 1000 × 968.02 = 984.10 psia 1000 + 968.02 − 3 1000 + 968.02

2785_C002.fm Page 67 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

67

Using this value of Pavg, we recalculate Z as Z=

1 1.785× 0.6

×(10) 1 + (984.10−14.73)×3.444×310 .825 5

(80+ 460)

or Z = 0.8780 Recalculating P2 from the Panhandle A Equation 2.55, we get 60 + 460 100 × 10 = 435.87 × 0.92 14.73

1.0788

6

1000 2 − P22 0.8539 (540 × 15 × 0.8780) (0.6)

0.55394

(15.5)2.6182

Solving for P2, we get P2 = 968.35 psia This is within 0.5 psi of the previously calculated value. Hence, we will not continue the iteration any further. Therefore, the outlet pressure is 968.35 psia.

Example 16 Using the Panhandle A equation, calculate the inlet pressure required in a natural gas pipeline, DN 300 with 6 mm wall thickness, 24 km long, for a gas ﬂow rate of 3.5 Mm3/day. The gas gravity = 0.6 and viscosity = 0.000119 Poise. The average gas temperature is 20°C. The delivery pressure is 6000 kPa (absolute). Assume base pressure = 101 kPa, base temperature = 15°C, and compressibility factor Z = 0.90, with a pipeline efﬁciency of 0.92. Solution Pipe inside diameter D = 300 – 2 × 6 = 288 mm Gas ﬂow temperature = 20 + 273 = 293 K Using Panhandle A Equation 2.56 and neglecting elevation effect, we substitute 15 + 273 3.5 × 10 6 = 4.5965 × 10 −3 × 0.92 101

Copyright 2005 by Taylor & Francis Group, LLC

1.0788

P12 − 6000 2 0.8539 (293 × 24 × 0.9) (0.6)

0.55394

(288)2.6182

2785_C002.fm Page 68 Friday, April 1, 2005 3:23 PM

68

GAS PIPELINE HYDRAULICS

Solving for inlet pressure, we get P 21 – (6000)2 = 19,812,783 or P1 = 7471 kPa (absolute)

2.16 PANHANDLE B EQUATION The Panhandle B equation, also known as the revised Panhandle equation, is used in large diameter, high pressure transmission lines. In fully turbulent ﬂow, it is found to be accurate for values of Reynolds number in the range of 4 to 40 million. This equation in USCS units is as follows: T Q = 737 E b P b

where Q = E = Pb = Tb = P1 = P2 = G = Tf = Le = Z = D =

1.02

P 2 − es P 2 1 2 0.961 G T L f eZ

0.51

D 2.53

(USCS units)

(2.59)

volume ﬂow rate, standard ft3/day (SCFD) pipeline efﬁciency, a decimal value less than 1.0 base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂow temperature, °R (460 + °F) equivalent length of pipe segment, mi gas compressibility factor, dimensionless pipe inside diameter, in.

Other symbols are as deﬁned previously. In SI units, the Panhandle B equation is T Q = 1.002 × 10 E b P −2

b

where Q = E = Tb = Pb = Tf =

1.02

P 2 − es P 2 1 2 0.961 G T f Le Z

0.51

D 2.53

gas ﬂow rate, standard m3/day pipeline efﬁciency, a decimal value less than 1.0 base temperature, K (273 + °C) base pressure, kPa average gas ﬂow temperature, K (273 + °C)

Copyright 2005 by Taylor & Francis Group, LLC

(SI units)

(2.60)

2785_C002.fm Page 69 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

P1 P2 Le Z

= = = =

69

upstream pressure, kPa (absolute) downstream pressure, kPa (absolute) equivalent length of pipe segment, km gas compressibility factor at the ﬂowing temperature, dimensionless

Other symbols are as deﬁned previously. The equivalent transmission factor for the Panhandle B equation in USCS is given by QG F = 16.7 E D

0.01961

(USCS units)

(2.61)

In SI units, it is QG F = 19.08 E D

0.01961

(SI units)

(2.62)

Example 17 Using the Panhandle B equation, calculate the outlet pressure in a natural gas pipeline, NPS 16 with 0.250 in. wall thickness, 15 miles long. The gas ﬂow rate is 100 MMSCFD at 1000 psia inlet pressure. The gas gravity = 0.6 and viscosity = 0.000008 lb/ft-sec. The average gas temperature is 80°F. Assume base pressure = 14.73 psia and base temperature = 60°F. The compressibility factor Z = 0.90 and pipeline efﬁciency is 0.92. Solution Inside diameter of pipe = 16 – 2 × 0.25 = 15.5 in. Gas ﬂow temperature = 80 + 460 = 540°R Using Panhandle B Equation 2.59, substituting the given values, we get 60 + 460 100 × 10 6 = 737 × 0.92 14.73

1.02

1000 2 − P22 0.961 (0.6) (540 × 15 × 0.90)

0.51

15.52.53

Solving for P2, we get 10002 – P22 = 60,778 P2 = 969.13 psia Compare this with the results of Panhandle A equation in Example 15, where the outlet pressure P2 = 968.35 psia. Therefore, the Panhandle B equation gives a slightly lower pressure drop compared to that from the Panhandle A equation. In other words, Panhandle A is more conservative and will give a lower ﬂow rate for the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 70 Friday, April 1, 2005 3:23 PM

70

GAS PIPELINE HYDRAULICS

same pressures compared to Panhandle B. In this example, we use the constant value of Z = 0.9, whereas in example 15, Z was calculated using the CNGA equation as Z = 0.8780. If we factor this in, the result for the outlet pressure in this example will be 969.9 psia, which is not too different from the calculated value of 969.13 psia. Example 18 Using the Panhandle B equation, calculate the inlet pressure in a natural gas pipeline, DN 300 with 6 mm wall thickness, 24 km long. The gas ﬂow rate is 3.5 Mm3/day, gas gravity = 0.6, and viscosity = 0.000119 Poise. The average gas temperature is 20°C, and the delivery pressure is 6,000 kPa (absolute). Assume base pressure = 101 kPa, base temperature = 15°C, and compressibility factor Z = 0.90. The pipeline efﬁciency is 0.92. Solution Inside diameter of pipe = 300 – 2 × 6 = 288 mm Gas ﬂow temperature = 20 + 273 = 293 K Neglecting elevations, using Panhandle B Equation 2.60, we get 15 + 273 3.5 × 10 6 = 1.002 × 10 −2 × 0.92 101

1.02

P12 − 6000 2 0.961 (0.6) (293 × 24 × 0.9)

0.51

2882.53

Solving for the inlet pressure P1, we get P21 – (6000) 2 = 19,945,469 P1 = 7480 kPa (absolute) Compare this with the results of the Panhandle A equation in Example 16, where the inlet pressure P1 = 7471 kPa (absolute). Again, we see that the Panhandle B equation gives a slightly lower pressure drop compared to that obtained from the Panhandle A equation.

2.17 INSTITUTE OF GAS TECHNOLOGY (IGT) EQUATION The IGT equation proposed by the Institute of Gas Technology is also known as the IGT distribution equation and is stated as follows for USCS units: T P 2 − e s P22 Q = 136.9 E b 01.8 Pb G T f Le µ 0.2

0.555

D 2.667

where Q = volume ﬂow rate, standard ft3/day (SCFD) E = pipeline efﬁciency, a decimal value less than 1.0

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(2.63)

2785_C002.fm Page 71 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

Pb Tb P1 P2 G Tf Le Z D m

= = = = = = = = = =

71

base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂow temperature, °R (460 + °F) equivalent length of pipe segment, mi gas compressibility factor, dimensionless pipe inside diameter, in. gas viscosity, lb/ft-s

Other symbols are as deﬁned previously. In SI units, the IGT equation is expressed as follows: T P 2 − e s P22 Q = 1.2822 × 10 E b 01.8 Pb G T f Le µ 0.2 −3

where Q = E = Tb = Pb = Tf = P1 = P2 = Le = m =

0.555

D 2.667

(SI units)

(2.64)

gas ﬂow rate, standard m3/day pipeline efﬁciency, a decimal value less than 1.0 base temperature, K (273 + °C) base pressure, kPa average gas ﬂow temperature, K (273 + °C) upstream pressure, kPa (absolute) downstream pressure, kPa (absolute) equivalent length of pipe segment, km gas viscosity, Poise

Other symbols are as deﬁned previously. Example 19 Using the IGT equation, calculate the ﬂow rate in a natural gas pipeline, NPS 16 with 0.250 in. wall thickness, 15 mi long. The inlet and outlet pressure are 1000 psig and 800 psig, respectively. The gas gravity = 0.6 and viscosity = 0.000008 lb/ft-s. The average gas temperature is 80°F, base pressure = 14.7 psia, and base temperature = 60°F. The compressibility factor Z = 0.90, and the pipeline efﬁciency is 0.95. Solution Inside diameter of pipe = 16 – 2 × 0.25 = 15.5 in. The pressures given are in psig, and they must be converted to absolute pressures.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 72 Friday, April 1, 2005 3:23 PM

72

GAS PIPELINE HYDRAULICS

Therefore, P1 = 1000 + 14.7 = 1014.7 psia P2 = 800 + 14.7 = 814.7 psia Tb = 60 + 460 = 520°R Tf = 80 + 460 = 540°R Substituting in IGT Equation 2.63, we get 520 1014.72 − 814.72 Q = 136.9 × 0.95 0.8 14.7 (0.6) × 540 × 15 × (8 × 10 −6 )0.2

0.555

15.52.667

Q = 263.1 × 106 ft3/day = 263.1 MMSCFD Therefore, the ﬂow rate is 263.1 MMSCFD. Example 20 A natural gas pipeline, DN 400 with 6 mm wall thickness, 24 km long, is used to transport gas at an inlet pressure of 7000 kPa (gauge) and an outlet pressure of 5500 kPa (gauge). The gas gravity = 0.6 and viscosity = 0.000119 Poise. The average gas temperature is 20°C. Assume base pressure = 101 kPa and base temperature = 15°C. The compressibility factor Z = 0.90 and pipeline efﬁciency is 0.95. a) Calculate the ﬂow rate using the IGT equation. b) What are the gas velocities at inlet and outlet? c) If the velocity must be limited to 10 m/s, what should the minimum pipe size be, assuming the ﬂow rate and inlet pressure remain constant? Solution Inside diameter of pipe D = 400 – 2 × 6 = 388 mm All pressures are given in gauge values and must be converted to absolute values. Inlet pressure P1 = 7000 + 101 = 7101 kPa (absolute) Outlet pressure P2 = 5500 + 101 = 5601 kPa (absolute) Base temperature Tb = 15 + 273 = 288 K Flowing temperature Tf = 20 + 273 = 293 K From IGT Equation 2.64, we get the ﬂow rate in m3/day as 288 71012 − 56012 Q = 1.2822 × 10 × 0.95 0.8 101 (0.6) × 293 × 24 × (1.19 × 10 −4 )0.2 −3

Copyright 2005 by Taylor & Francis Group, LLC

0.555

(388)2.667

2785_C002.fm Page 73 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

73

or Q = 7,665,328 m3/day = 7.67 Mm3/day (a) Therefore, the ﬂow rate is 7.67 Mm3/day. (b) Using Equation 2.29, we calculate the average velocity of the gas at the inlet pressure as 7.67 × 10 6 101 0.9 × 293 Inlet velocity u1 = 14.7349 288 7101 = 9.78 m/s 2 388 In the preceding, we assumed a constant compressibility factor, Z = 0.9. Similarly, at the outlet pressure, the average gas velocity is 7.67 × 10 6 101 0.9 × 293 Outlet velocity u2 = 14.7349 288 5601 = 12.4 m/s 2 388 (c) Since the velocity must be limited to 10 m/s, the pipe diameter must be increased. Increasing the pipe diameter will also increase the outlet pressure if we keep both the ﬂow rate and inlet pressure the same as before. The increased outlet pressure will also reduce the gas velocity as can be seen from Equation 2.29. We will try a DN 450 pipe with 10 mm wall thickness. Inside diameter of pipe D = 450 – 2 × 10 = 430 mm Assuming P1 and Q are the same as before, we calculate the new outlet pressure P2 from IGT Equation 2.64 as 288 71012 − P22 7.67 × 10 6 = 1.2822 × 10 −3 × 0.95 −4 0.2 0.8 101 (0.6) × 293 × 24 × (1.19 × 10 )

0.555

(430)2.667

Solving for P2, we get P2 = 6228 kPa The new velocity at the outlet will be 7.67 × 10 6 101 0.9 × 293 u2 = 14.7349 = 9.08 m/s 430 2 288 6228 Since this is less than the 10 m/s speciﬁed, the DN 450 pipe is satisfactory. In the preceding calculations we assumed the same compressibility factor for both inlet and outlet pressures. Actually, a more nearly correct solution would be to calculate Z using the CNGA equation at both inlet and outlet conditions and using these values in the calculation of gas velocities. This is left as an exercise for the reader.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 74 Friday, April 1, 2005 3:23 PM

74

GAS PIPELINE HYDRAULICS

2.18 SPITZGLASS EQUATION The Spitzglass equation has been around for many years and originally was used in fuel gas piping calculations. There are two versions of the Spitzglass equation. One equation is for low pressure (less than or equal to 1 psig) and another is for high pressure (more than 1 psig). These equations have been modiﬁed to include a pipeline efﬁciency and compressibility factor. The low-pressure (less than or equal to 1 psig) version of the Spitzglass equation in USCS units is T P1 − P2 Q = 3.839 × 10 E b Pb GT f Le Z 1 + 3D.6 + 0.03D 3

(

)

0.5

D 2.5

(USCS units) (2.65)

where Q = E = Pb = Tb = P1 = P2 = G = Tf = Le = D = Z =

volume ﬂow rate, standard ft3/day (SCFD) pipeline efﬁciency, a decimal value less than 1.0 base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂow temperature, °R (460 + °F) equivalent length of pipe segment, mi pipe inside diameter, in. gas compressibility factor, dimensionless

Other symbols are as deﬁned previously. The low-pressure (less than 6.9 kPa) version of the Spitzglass equation in SI units is T P1 − P2 Q = 5.69 × 10 E b 91.44 P + + 0.0012 D GT L Z 1 b D f e −2

(

)

0.5

D 2.5

(SI units) (2.66)

where Q = E = Tb = Pb = P1 = P2 = G = Tf =

gas ﬂow rate, standard m3/day pipeline efﬁciency, a decimal value less than 1.0 base temperature, K (273 + °C) base pressure, kPa upstream pressure, kPa (absolute) downstream pressure, kPa (absolute) gas gravity (air = 1.00) average gas ﬂow temperature, K (273 + °C)

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 75 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

75

Le = equivalent length of pipe segment, km Z = gas compressibility factor, dimensionless Other symbols are as deﬁned previously. The high-pressure (more than 1 psig) version in USCS units is as follows. T P12 − e s P22 Q = 729.6087 E b Pb GT f Le Z 1 + 3D.6 + 0.03D

(

)

0.5

D 2.5

(USCS units) (2.67)

where Q = E = Pb = Tb = P1 = P2 = G = Tf = Le = D = Z =

volume ﬂow rate, standard ft3/day (SCFD) pipeline efﬁciency, a decimal value less than 1.0 base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂow temperature, °R (460 + °F) equivalent length of pipe segment, mi pipe inside diameter, in. gas compressibility factor, dimensionless

Other symbols are as deﬁned previously. In SI units, the high-pressure (more than 6.9 kPa) version of the Spitzglass equation is T P12 − e s P22 Q = 1.0815 × 10 E b 91 . 44 Pb GT f Le Z 1 + D + 0.0012 D −2

(

)

0.5

D 2.5

(SI units) (2.68)

where Q = E = Tb = Pb = P1 = P2 = G = Tf = Le = Z =

gas ﬂow rate, standard m3/day pipeline efﬁciency, a decimal value less than 1.0 base temperature, K (273 + °C) base pressure, kPa upstream pressure, kPa (absolute) downstream pressure, kPa (absolute) gas gravity (air = 1.00) average gas ﬂow temperature, K (273 + °C) equivalent length of pipe segment, km gas compressibility factor, dimensionless

Other symbols are as deﬁned previously.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 76 Friday, April 1, 2005 3:23 PM

76

GAS PIPELINE HYDRAULICS

Example 21 Calculate the fuel gas capacity of an NPS 6 pipe, with an inside diameter of 6.065 in. and a total equivalent length of 180 ft. The ﬂowing temperature of fuel gas is 60°F, and the inlet pressure is 1.0 psig. Consider a pressure drop of 0.7 in the water column and the speciﬁc gravity of gas = 0.6. Assume pipeline efﬁciency E = 1.0 and compressibility factor Z = 1.0. The base pressure and base temperature are 14.7 psia and 60°F, respectively. Solution Base temperature = 60 + 460 = 520°R Gas ﬂowing temperature = 60 + 460 = 520°R Pressure drop (P1 – P2) = 1.0 −

0.7 × 0.433 = 0.9747 psi 12

Since this is low pressure, using Spitzglass Equation 2.65, we get 520 0.9747 Q = 3.839 × 10 × 1.0 × 180 × 1.0 1 + 3.6 + 0.03 × 6.065 14.7 0.6 × 520 × 5280 6.065 3

(

)

0.5

6.0652.5

Q = 2,794,842 SCFD = 2.79 MMSCFD Therefore, the fuel gas capacity is 2.79 MMSCFD.

2.19 MUELLER EQUATION The Mueller equation is another form of the ﬂow rate vs. pressure relationship in gas pipelines. In USCS units, it is expressed as follows: T P12 − e s P22 Q = 85.7368 E b 0.7391 0.2609 Pb G T f Le µ where Q = E = Pb = Tb = P1 = P2 = G =

0.575

D 2.725

volume ﬂow rate, standard ft3/day (SCFD) pipeline efﬁciency, a decimal value less than 1.0 base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00)

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(2.69)

2785_C002.fm Page 77 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

Tf Le D m

= = = =

77

average gas ﬂow temperature, °R (460 + °F) equivalent length of pipe segment, mi pipe inside diameter, in. gas viscosity, lb/ft-s

Other symbols are as deﬁned previously. In SI units, the Mueller equation is as follows: T P12 − e s P22 Q = 3.0398 × 10 −2 E b 0.7391 0.2609 Pb G T f Le µ where Q = E = Tb = Pb = P1 = P2 = G = Tf = Le = m =

0.575

D 2.725

(SI units)

(2.70)

gas ﬂow rate, standard m3/day pipeline efﬁciency, a decimal value less than 1.0 base temperature, K (273 + °C) base pressure, kPa upstream pressure, kPa (absolute) downstream pressure, kPa (absolute) gas gravity (air = 1.00) average gas ﬂow temperature, K (273 + °C) equivalent length of pipe segment, km gas viscosity, cP

Other symbols are as deﬁned previously.

2.20 FRITZSCHE EQUATION The Fritzsche formula, developed in Germany in 1908, has found extensive use in compressed air and gas piping. In USCS units, it is expressed as follows: T P2 − P2 Q = 410.1688 E b 01.8587 2 Pb G T f Le

0.538

D 2.69

(USCS units)

(2.71)

D 2.69

(SI units)

(2.72)

All symbols are as deﬁned before. In SI units, T P 2 − es P 2 Q = 2.827 E b 10.8587 2 Pb G T f Le All symbols are as deﬁned before.

Copyright 2005 by Taylor & Francis Group, LLC

0.538

2785_C002.fm Page 78 Friday, April 1, 2005 3:23 PM

78

GAS PIPELINE HYDRAULICS

2.21 EFFECT OF PIPE ROUGHNESS In the preceding sections, we used the pipe roughness as a parameter in the friction factor and transmission factor calculations. Both the AGA and Colebrook-White equations use the pipe roughness, whereas the Panhandle and Weymouth equations do not use the pipe roughness directly in the calculations. Instead, these equations use a pipeline efﬁciency to compensate for the internal conditions and age of the pipe. Therefore, when comparing the predicted ﬂow rates or pressures using the AGA or Colebrook-White equations with the Panhandle or Weymouth equations, we can adjust the pipeline efﬁciency to correlate with the pipe roughness used in the former equations. Since most gas pipelines operate in the turbulent zone, the laminar ﬂow friction factor, which is independent of pipe roughness, is of little interest to us. Concentrating, therefore, on turbulent ﬂow, we see that Colebrook-White Equation 2.45 is affected by variation in pipe internal roughness. For example, suppose we want to compare an internally coated pipeline with an uncoated pipeline. The internal roughness of the coated pipe might be in the range of 100 to 200 µin., whereas the uncoated pipe might have a roughness of 600 to 800 µin. or more. If the pipe is NPS 20 with a 0.500 in. wall thickness, the relative roughness using the lower roughness value is as follows: For coated pipe, e 100 × 10 −6 = = 5.263 × 10 −6 D 19 and For uncoated pipe, e 600 × 10 −6 = = 3.1579 × 10 −5 D 19 Substituting these values of relative roughness in Equation 2.45 and using a Reynolds number of 10 million, we calculate the following transmission factors: F = 21.54 for coated pipe and F = 19.83 for uncoated pipe Since the ﬂow rate is directly proportional to the transmission factor F, from General Flow Equation 2.4 we see that the coated pipe will be able to transport 21.54 −19.83 = 0.086 = 8.6% more ﬂow rate than the uncoated pipe, if all other parameters 19.83 remain the same. This is true in the fully turbulent zone where the Reynolds number has little effect on the friction factor f and the transmission factor F. However, in the smooth

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 79 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

79

pipe zone, pipe roughness has less effect on the friction factor and the transmission factor. This is evident from the Moody diagram in Figure 2.3. Using a Reynolds number of 106, we ﬁnd from the Moody diagram in Figure 2.3, for coated pipe, that f = 0.0118

and F = 18.41

f = 0.0122

and F = 18.10

and, for the uncoated pipe,

Therefore, the increase in ﬂow rate in this case will be 18.41 − 18.10 = 0.017 = 1.7% 18.10 Thus, the impact of pipe roughness is less in the smooth pipe zone or for a lower Reynolds number. A similar comparison can be made using the AGA equation. Figure 2.4 shows the effect of pipe roughness on the pipeline ﬂow rate considering the AGA and Colebrook-White equations. The graph is based on NPS 20 pipe, 0.500 in. wall thickness, 120 miles long, with 1200 psig upstream pressure and 800 psig downstream pressure. The ﬂowing temperature of gas is 70°F. It can be seen that as the pipe roughness is increased from 200 to 800 µin., the ﬂow rate decreases from 224 MMSCFD to 206 MMSCFD for the Colebrook-White equation and from 220 MMSCFD to 196 MMSCFD for the AGA equation.

240

Flow rate, MMSCFD

230 220 Colebrook-White 210 AGA 200 190 180 200

400

600 Roughness, microinches

Figure 2.4

Effect of pipe roughness.

Copyright 2005 by Taylor & Francis Group, LLC

800

1000

2785_C002.fm Page 80 Friday, April 1, 2005 3:23 PM

80

GAS PIPELINE HYDRAULICS

We can therefore conclude that decreasing the pipe roughness directly results in a throughput increase in a pipeline. However, the cost of internally coating a pipe to reduce the pipe roughness must be weighed against the revenue increase due to enhanced ﬂow rate. We will revisit this issue in Chapter 10, when we discuss pipeline economics.

2.22 COMPARISON OF FLOW EQUATIONS In the preceding sections, we calculated the ﬂow rates and pressures in gas pipelines using the various ﬂow equations. Each equation is slightly different from the other, and some equations consider the pipeline efﬁciency while others use an internal pipe roughness value. How do these equations compare when predicting ﬂow rates through a given pipe size when the upstream or downstream pressure is held constant? Obviously, some equations will predict higher ﬂow rates for the same pressures than others. Similarly, if we start with a ﬁxed upstream pressure in a pipe segment at a given ﬂow rate, these equations will predict different downstream pressures. This indicates that some equations calculate higher pressure drops for the same ﬂow rate than others. Figure 2.5 and Figure 2.6 show some of these comparisons when using the AGA, Colebrook-White, Panhandle, and Weymouth equations. In Figure 2.5, we consider a pipeline 100 mi long, NPS 16 with 0.250 in. wall thickness, operating at a ﬂow rate of 100 MMSCFD. The gas ﬂowing temperature is 80°F. With the upstream pressure ﬁxed at 1400 psig, the downstream pressure was calculated using the different ﬂow equations. By examining Figure 2.5, it is clear that the highest pressure drop is predicted by the Weymouth equation and the lowest pressure drop is predicted by the Panhandle B equation. It must be noted that we

1450

Pressure, psig

1400 1350 1300 Panhandle B Panhandle A Colebrook-White AGA

1250 1200

Weymouth 1150 0.00

100.00 Distance, mi

Figure 2.5

Comparison of flow equations.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 81 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

81

Upstream pressure, psig

1200

Weymouth AGA Colebrook-White Panhandle B Panhandle A

1100

1000

900

800 200

300

400

500

600

Flow rates, MMSCFD Figure 2.6

Upstream pressures for various flow equations.

used a pipe roughness of 700 µin. for both the AGA and Colebrook equations, whereas a pipeline efﬁciency of 0.95 was used in the Panhandle and Weymouth equations. Figure 2.6 shows a comparison of the ﬂow equations from a different perspective. In this case, we calculated the upstream pressure required for an NPS 30 pipeline, 100 miles long, holding the delivery pressure constant at 800 psig. The upstream pressure required for various ﬂow rates, ranging from 200 to 600 MMSCFD, was calculated using the ﬁve ﬂow equations. Again it can be seen that the Weymouth equation predicts the highest upstream pressure at any ﬂow rate, whereas the Panhandle A equation calculates the least pressure. We therefore conclude that the most conservative ﬂow equation that predicts the highest pressure drop is the Weymouth equation and the least conservative ﬂow equation is Panhandle A.

2.23 SUMMARY In this chapter we introduced the various methods of calculating the pressure drop in a pipeline transporting gas and gas mixtures. The more commonly used equations for pressure drop vs. ﬂow rate and pipe size were discussed and illustrated using example problems. The effect of elevation changes was explained, and the concepts of the Reynolds number, friction factor, and transmission factor were introduced. The importance of the Moody diagram and how to calculate the friction factor for laminar and turbulent ﬂow were explained. We compared the more commonly used pressure drop equations, such as AGA, Colebrook-White, Weymouth, and Panhandle equations. The use of a pipeline efﬁciency in comparing various equations was illustrated using an example. The average velocity of gas ﬂow was introduced, and the limiting value of erosional velocity was discussed.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 82 Friday, April 1, 2005 3:23 PM

82

GAS PIPELINE HYDRAULICS

PROBLEMS 1. A gas pipeline, NPS 18 with 0.375 in. wall thickness, transports natural gas (speciﬁc gravity = 0.6) at a ﬂow rate of 160 MMSCFD at an inlet temperature of 60°F. Assuming isothermal ﬂow, calculate the velocity of gas at the inlet and outlet of the pipe if the inlet pressure is 1200 psig and the outlet pressure is 700 psig. The base pressure and base temperature are 14.7 psia and 60°F, respectively. Assume the compressibility factor Z = 0.95. What is the pipe length for these pressures, if elevations are neglected? 2. A natural gas pipeline, DN 400 with 10 mm wall thickness, transports 3.2 Mm3/ day. The speciﬁc gravity of gas is 0.6 and viscosity is 0.00012 Poise. Calculate the value of the Reynolds number. Assume the base temperature and base pressure are 15 C and 101 kPa, respectively. 3. A natural gas pipeline, NPS 20 with 0.500 in. wall thickness, 50 miles long, transports 220 MMSCFD. The speciﬁc gravity of gas is 0.6 and viscosity is 0.000008 lb/ft-s. Calculate the friction factor using the Colebrook equation. Assume absolute pipe roughness = 750 µin. The base temperature and base pressure are 60°F and 14.7 psia, respectively. What is the upstream pressure for an outlet pressure of 800 psig? 4. For a gas pipeline ﬂowing 3.5 Mm3/day gas of speciﬁc gravity 0.6 and viscosity of 0.000119 Poise, calculate the friction factor and transmission factor, assuming a DN 400 pipeline, 10 mm wall thickness, and internal roughness of 0.015 mm. The base temperature and base pressure are 15°C and 101 kPa, respectively. If the ﬂow rate is increased by 50%, what is the impact on the friction factor and transmission factor? If the pipe length is 48 km, what is the outlet pressure for an inlet pressure of 9000 kPa? 5. A gas pipeline ﬂows 110 MMSCFD gas of speciﬁc gravity 0.65 and viscosity of 0.000008 lb/ft-s. Calculate, using the modiﬁed Colebrook-White equation, the friction factor and transmission factor, assuming an NPS 20 pipeline, 0.375 in. wall thickness, and internal roughness of 700 µin. The base temperature and base pressure are 60°F and 14.7 psia, respectively. 6. Using the AGA method, calculate the transmission factor and friction factor for gas ﬂow in an NPS 20 pipeline with 0.375 in. wall thickness. The ﬂow rate is 250 MMSCFD, gas gravity = 0.6, and viscosity = 0.000008 lb/ft-sec. The absolute pipe roughness is 600 µin. Assume a bend index of 60°, base pressure = 14.73 psia, and base temperature = 60°F. If the ﬂow rate is doubled, what pipe size is needed to keep both inlet and outlet pressures the same as that at the original ﬂow rate? 7. A natural gas transmission line transports 4 million m3/day of gas from a processing plant to a compressor station site 100 km away. The pipeline can be assumed to be along a ﬂat terrain. Calculate the minimum pipe diameter required such that the maximum pipe operating pressure is limited to 8500 kPa. The delivery pressure desired at the end of the pipeline is a minimum of 5500 kPa. Assume a pipeline efﬁciency of 0.92. The gas gravity is 0.60, and the gas temperature is 18°C. Use the Weymouth equation, considering a base temperature = 15°C and base pressure = 101 kPa. The gas compressibility factor Z = 0.90.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 83 Friday, April 1, 2005 3:23 PM

PRESSURE DROP DUE TO FRICTION

83

8. Using the Panhandle B equation, calculate the outlet pressure in a natural gas pipeline, NPS 16 with 0.250 in. wall thickness, 25 miles long. The gas ﬂow rate is 120 MMSCFD at 1200 psia inlet pressure. The gas gravity = 0.6 and viscosity = 0.000008 lb/ft-sec. The average gas temperature is 80°F. Assume the base pressure = 14.73 psia and base temperature = 60°F. The compressibility factor Z = 0.90 and pipeline efﬁciency is 0.95.

REFERENCES 1. Nayyar, M.L., Piping Handbook, 7th ed., McGraw-Hill, New York, 2000. 2. Baumeister, T., Ed., Standard Handbook for Mechanical Engineers, 7th ed., McGrawHill, New York, 1967. 3. Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003. 4. Westaway, C.R. and Loomis, A.W., Cameron Hydraulic Data, 16th ed., IngersollRand, Montvale, NJ, 1981. 5. Flow of Fluids through Valves, Fittings and Pipe, Crane Company, New York, 1976. 6. McCain, W.D. Jr., The Properties of Petroleum Fluids, Petroleum Publishing Company, Tulsa, OK, 1973. 7. Holman, J.P., Thermodynamics, 2nd ed., McGraw-Hill, New York, 1974. 8. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 9. Engineering Data Book, 10th ed., Gas Processors Suppliers Association, Tulsa, OK, 1994. 10. Pipeline Design for Hydrocarbon Gases and Liquids, American Society of Civil Engineers, New York, 1975. 11. Steady Flow in Gas Pipelines, Contract Report No. 10, July 1965, Pipeline Research Council International, Houston, TX, 1965. 12. Katz, D.L. et al., Handbook of Natural Gas Engineering, McGraw-Hill, New York, 1959.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C002.fm Page 84 Friday, April 1, 2005 3:23 PM

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 85 Friday, April 1, 2005 3:09 PM

CHAPTER

3

Pressure Required to Transport In this chapter we will extend the use of the concepts of pressure drop calculations developed in Chapter 2 to determine the total pressure required for transporting gas in a pipeline under various conﬁgurations, such as series and parallel pipelines. We will identify the various components that make up this total pressure and analyze their impact on gas pipeline pressures. The effect of intermediate delivery volumes and injection rates along a gas pipeline, the impact of contract delivery pressures, and the necessity of regulating pressures using a control valve or pressure regulators will also be analyzed. Thermal effects due to heat transfer between the gas and the surrounding soil in a buried pipeline, soil temperatures and thermal conductivities, and the JouleThompson effect will be introduced with reference to commercial hydraulic simulation models. Equivalent lengths in series piping and equivalent diameters in parallel piping will be explained. We will compare different pipe looping scenarios to improve pipeline throughput and review the concept of the hydraulic pressure gradient. Calculation methodology for line pack in a gas pipeline will also be discussed.

3.1 TOTAL PRESSURE DROP REQUIRED In the ﬂow of incompressible ﬂuids such as water, the pressure required to transport a speciﬁed volume of ﬂuid from point A to point B will consist of the following components: 1. Frictional component 2. Elevation component 3. Pipe delivery pressure

In addition, in some cases where the pipeline elevation differences are drastic, we must also take into account the minimum pressure in a pipeline such that vaporization of liquid does not occur. The latter results in two-phase ﬂow in the pipeline, which causes higher pressure drop and, therefore, more pumping power requirement in addition to possible damage to pumping equipment. Thus, single-phase incompressible 85

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 86 Friday, April 1, 2005 3:09 PM

86

GAS PIPELINE HYDRAULICS

ﬂuids must be pumped such that the pressure at any point in the pipeline does not drop below the vapor pressure of the liquid. When pumping gases, which are compressible ﬂuids, the three components listed in the preceding section also contribute to the total pressure required. Even though the relationship between the total pressure required and the pipeline elevation is not straightforward (as in liquid ﬂow), the dependency still exists and will be demonstrated using an example problem. Going back to the case of a liquid pipeline, suppose the total pressure required to pump a given volume is 1000 psig and it is composed of the following components: 1. Frictional component = 600 psig 2. Elevation component = 300 psig 3. Delivery pressure = 100 psig

We will now discuss each of these components that make up the total pressure required by comparing the situation between a liquid pipeline and a gas pipeline.

3.2 FRICTIONAL EFFECT The frictional effect results from the ﬂuid viscosity and pipe roughness. It is similar in liquid and gas ﬂow. The effect of friction was discussed in Chapter 2, where we introduced the internal roughness of pipe and how the friction factor and transmission factor were calculated using the Colebrook-White and AGA equations. We also discussed how the Weymouth and Panhandle equations took into account the internal conditions and age of the pipe by utilizing a pipeline efﬁciency factor rather than a friction factor. The magnitude of the pressure drop due to friction in a gas pipeline is generally held to smaller values in comparison with liquid pipelines. This is because efﬁcient gas pipeline transportation requires keeping the average gas pressure as high as possible. As pressure drops due to expansion of gas, there is loss in efﬁciency. The lower the pressure at the downstream end, the higher will be the compression ratio required (hence, the higher the HP) to boost the pressure for shipment downstream to the next compressor station in a long-distance gas pipeline. In this chapter we will continue to calculate the pressure drop due to friction in various pipe conﬁgurations that include ﬂow injection, deliveries, and series and parallel piping.

3.3 EFFECT OF PIPELINE ELEVATION The elevation component referred to in Section 3.1 is due to the difference in elevation along the pipeline that necessitates additional pressure for raising the ﬂuid in the pipeline from one point to another. Of course, a drop in elevation will have the opposite effect of a rise in elevation. The elevation component of 300 psig in the preceding example depends upon the static elevation difference between the beginning of the pipeline, A, and the delivery point, B, and the liquid speciﬁc gravity. In the case of a gas pipeline, the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 87 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

87

elevation component will depend upon the static elevation differences between A and B, as well as the gas gravity. However, the relationship between these parameters is more complex in a gas pipeline compared to a liquid pipeline. The rise and fall in elevations between the origin A and the terminus B have to be accounted for separately and summed up according to Section 2.4 in Chapter 2. Further, compared to a liquid, the gas gravity is several orders of magnitude lower and, hence, the inﬂuence of elevation is smaller in a pipeline that transports gas. Generally, if we were to break down the total pressure required in a gas pipeline into the three components discussed earlier, we would ﬁnd that the elevation component is very small. Let us illustrate this using an example. Example 1 A gas pipeline, NPS 16 with 0.250 in. wall thickness, 50 mi long, transports natural gas (speciﬁc gravity = 0.6 and viscosity = 0.000008 lb/ft-s) at a ﬂow rate of 100 MMSCFD at an inlet temperature of 60°F. Assuming isothermal ﬂow, calculate the inlet pressure required if the required delivery pressure at the pipeline terminus is 870 psig. The base pressure and base temperature are 14.7 psig and 60°F, respectively. Use the Colebrook equation with pipe roughness of 0.0007 in. Case A—Consider no elevation changes along the pipeline length. Case B—Consider elevation changes as follows: inlet elevation of 100 ft and elevation at delivery point of 450 ft, with elevation at the midpoint of 250 ft. Solution Inside diameter of pipe D = 16 – 2 × 0.250 = 15.5 in. First, we calculate the Reynolds number from Equation 2.34: 14.7 0.6 × 100 × 10 6 R = 0.0004778 = 6,535,664 60 + 460 0.000008 × 15.5 Next, using Colebrook Equation 2.39, we calculate the friction factor as 0.0007 2.51 = −2Log10 + f 3.7 × 15.5 6535664 f

1

Solving by trial and error, we get f = 0.0109 Therefore, the transmission factor is, using Equation 2.42, F=

Copyright 2005 by Taylor & Francis Group, LLC

2 0.0109

= 19.1954

2785_C003.fm Page 88 Friday, April 1, 2005 3:09 PM

88

GAS PIPELINE HYDRAULICS

To calculate the compressibility factor Z , the average pressure is required. Since the inlet pressure is unknown, we will calculate an approximate value of Z using a value of 110% of the delivery pressure for the average pressure. The average pressure is Pavg = 1.1 × (870 + 14.7) = 973.17 psia Using CNGA Equation 1.34, we calculate the value of the compressibility factor as Z=

1 +

(

1 (973.17−14.7)×344400 (10 )1.785× 0.6 5203.8825

)

= 0.8629

Case A Since there is no elevation difference between the beginning of the pipeline and the end of the pipeline, the elevation component in Equation 2.7 can be neglected, and es = 1. The outlet pressure is P2 = 870 + 14.7 = 884.7 psia. From General Flow Equation 2.4, substituting the given values, we get P12 − 884.72 520 100 × 10 = 38.77 × 19.1954 14.7 0.6 × 520 × 50 × 0.8629

0.5

6

(15.5)2.5

Therefore, the upstream pressure is P1 = 999.90 psia = 985.20 psig Using this value of P1, we calculate the new average pressure using Equation 2.14: Pavg =

999.9 × 884.7 2 999.9 + 884.7 − = 943.47 psia 999.9 + 884.7 3

compared to 973.17 we used for calculating Z. Recalculating Z using the new value of Pavg, we get Z=

Copyright 2005 by Taylor & Francis Group, LLC

1 +

(

1 (943.47−14.7)×344400 (10 )1.785× 0.6 5203.8825

)

= 0.8666

2785_C003.fm Page 89 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

89

This compares with 0.8629 we calculated earlier for Z . We will now recalculate the inlet pressure using this value of Z. From General Flow Equation 2.4, we get P12 − 884.72 520 100 × 10 = 38.77 × 19.1954 14.7 0.6 × 520 × 50 × 0.8666 6

0.5

(15.5)2.5

Solving for the upstream pressure, we get P1 = 1000.36 psia = 985.66 psig This is close enough to the previously calculated value 985.20 psig, and no further iteration is needed. Therefore, the pressure required at the beginning of the pipeline in case A is 985.66 psig when the elevation difference is zero. We will now calculate the pressure required, taking into account the given elevations at the beginning, midpoint, and end of the pipeline. Case B We will use Z = 0.8666 throughout, as in case A. Using Equation 2.10, the elevation adjustment factor is ﬁrst calculated for each of the two segments. For the ﬁrst segment, from milepost 0.0 to milepost 25.0, we get 250 − 100 s1 = 0.0375 × 0.6 = 0.0075 520 × 0.8666 Similarly, for the second segment, from milepost 25.0 to milepost 50.0, we get 450 − 100 s2 = 0.0375 × 0.6 = 0.0175 520 × 0.8666 Therefore, the adjustment for elevation is, using Equation 2.12,

j=

e 0.0075 − 1 = 1.0038 for the ﬁrst segment 0.0075

and

j=

Copyright 2005 by Taylor & Francis Group, LLC

e 0.0175 − 1 = 1.0088 for the second segment 0.0175

2785_C003.fm Page 90 Friday, April 1, 2005 3:09 PM

90

GAS PIPELINE HYDRAULICS

For the entire length, 450 − 100 s2 = 0.0375 × 0.6 = 0.0175 520 × 0.8666 The equivalent length from Equation 2.13 is then Le = 1.0038 × 25 + 1.0088 × 25 × e0.0075 = 50.5049 mi. Therefore, we see that the effect of the elevation is taken into account partly by increasing the pipe length from 50 mi to 50.50 mi, approximately. Substituting in Equation 2.7, we get P12 − e 0.0175 884.72 520 100 × 10 = 38.77 × 19.1954 . . . 14 7 0 6 520 50 50 0 8666 × × × . 6

0.5

15.52.5

Solving for the inlet pressure P1, P1 = 1008.34 psia = 993.64 psig Thus, the pressure required at the beginning of the pipeline in case B is 993.64 psig, taking into account elevation difference along the pipeline. Compare this with 985.66 calculated ignoring the elevation differences. For simplicity, we assume the same value of Z in the preceding calculations as in the previous case. To be correct, we should recalculate Z based on the average pressure and repeat calculations until the results are within 0.1 psi. This is left as an exercise for the reader. It can be seen from the preceding calculations that, due to an elevation difference of 350 ft (450 ft − 100 ft) between the delivery point and the beginning of the pipeline, the required pressure is approximately 8 psig (993.64 psig − 985.66 psig) more. In a liquid line, the effect of elevation would have been more. The elevation difference of 350 ft in a water line would result in an increased pressure of 350 × 0.433 = 152 psi, approximately, at the upstream end.

3.4 EFFECT OF CHANGING PIPE DELIVERY PRESSURE The delivery pressure component discussed in Section 3.1 is also similar to that between liquid and gas pipelines. The higher the pressure desired at the delivery end or terminus of the pipeline, the higher will be the total pressure required at the upstream end of the pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 91 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

91

The impact of changing the delivery pressure is not linear in the case of a compressible ﬂuid such as natural gas. For example, in a liquid pipeline, changing the delivery pressure from 100 to 200 psig will simply increase the required pressure at the pipe inlet by the same amount. Thus, suppose 1000 psig was the required inlet pressure in a liquid pipeline, at a certain ﬂow rate and at a delivery pressure of 100 psig. When the delivery pressure required is increased to 200 psig, the inlet pressure will increase to exactly 1100 psig. We will now explore the effect of changing the contract delivery pressure at the end of a gas pipeline. In liquid pipelines, an increase or decrease in the delivery pressure will proportionately increase or decrease the upstream pressure. In a gas pipeline, the increase and decrease in the upstream pressure will not be proportionate due to the nonlinear nature of the gas pressure drop. This will be explained in more detail in Section 3.9 in this chapter. Suppose that in the preceding Example 1 (neglecting elevation change), the delivery pressure required increases from 870 to 950 psig. If pressure variations were linear, as in a liquid pipeline, we would expect the required inlet pressure to increase from 985.66 to 985.66 + (950 − 870) = 1066 psig, approximately. However, this is incorrect because the pressure variation is not linear in gas pipelines. We will now calculate the required inlet pressure when the delivery pressure is increased from 870 to 950 psig. All parameters in case A are the same except for the delivery pressure. The increased delivery pressure will cause the compressibility factor to change slightly due to the change in average pressure. However, for simplicity, we will assume Z = 0.8666, as before. The new delivery pressure is P2 = 950 + 14.7 = 964.7 psia Substituting in General Flow Equation 2.4, we get 520 P12 − 964.72 100 × 10 6 = 38.77 × 19.1954 14.7 0.6 × 520 × 50 × 0.8666

0.5

15.52.5

Therefore, P1 = 1071.77 psia = 1057.07 psig Thus, the pressure required at the beginning of the pipeline is approximately 1057 psig. This compares with a value of 1066 psig we calculated if the pressure variation were linear. In general, for a gas pipeline, if the delivery pressure is increased by ∆P, the inlet pressure will increase by less than ∆P. Similarly, if the delivery pressure is decreased by ∆P, the inlet pressure will decrease by less than ∆P. We will illustrate this using the preceding example. Suppose in case A, the delivery pressure was decreased from 870 to 800 psig. If pressure variation were linear, we would expect the pipe inlet pressure to decrease

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 92 Friday, April 1, 2005 3:09 PM

92

GAS PIPELINE HYDRAULICS

by 70 psig to (985.66 – 70) = 916 psig, approximately. However, as indicated earlier, this is incorrect. We will now calculate the actual inlet pressure using the General Flow equation considering the reduced outlet pressure of 800 psig. All parameters in case A are the same except for the delivery pressure. The decreased delivery pressure will cause the compressibility factor to change slightly due to the change in average pressure. However, for simplicity, we will assume Z = 0.8666, as before. New delivery pressure P2 = 800 + 14.7 = 814.7 psia Substituting in General Flow Equation 2.4, we get 520 P12 − 814.72 100 × 10 6 = 38.77 × 19.1954 14.7 0.6 × 520 × 50 × 0.8666

0.5

15.52.5

Therefore, P1 = 939.03 psia = 924.33 psig Thus, the pressure required at the beginning of the pipeline is 924.33 psig. This compares with a value of 916 psig if the pressure variation were linear. Therefore, by decreasing the delivery pressure by 70 psig, the inlet pressure decreases by less than 70 psig. We can perform a similar analysis by changing the inlet pressure by a ﬁxed amount and calculating the effect on the pipe delivery pressure. As before, considering no elevation changes, an inlet pressure of 985.66 psig results in a delivery pressure of 870 psig. Suppose we decrease the inlet pressure to 900 psig (a reduction of 85.66 psig); if pressures were linear, we would expect the delivery pressure to drop to (870 – 85.66) = 784.34 psig. Actually, we will see that the delivery pressure would drop to a number lower than this. In other words, decreasing the inlet pressure by ∆P reduces the outlet pressure by more than ∆P. Following the previous methodology, we calculate the revised delivery pressure by assuming the same Z = 0.8666 for simplicity. New inlet pressure = 900 + 14.7 = 914.7 psia Substituting in General Flow Equation 2.4, we calculate the outlet pressure as 520 914.72 − P2 2 100 × 10 6 = 38.77 × 19.1954 14.7 0.6 × 520 × 50 × 0.8666 Therefore, P2 = 786.54 psia = 771.84 psig

Copyright 2005 by Taylor & Francis Group, LLC

0.5

15.52.5

2785_C003.fm Page 93 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

93

Thus, the delivery pressure is reduced by (870 – 771.84) = 98.16 psig, whereas the inlet pressure was reduced by only 85.66 psig. In general, if the inlet pressure is decreased by ∆P, the delivery pressure will decrease by more than ∆P. On the other hand, if the inlet pressure is increased by ∆P, the delivery pressure will increase by more than ∆P. Therefore, if the inlet pressure is increased from 985.66 to 1085.66 psig (an increase of 100 psig), the delivery pressure will increase from 870 psig to a number larger than 970 psig.

3.5 PIPELINE WITH INTERMEDIATE INJECTIONS AND DELIVERIES A pipeline in which gas enters at the beginning of the pipeline and the same volume exits at the end of the pipeline is a pipeline with no intermediate injection or deliveries. When portions of the inlet volume are delivered at various points along the pipeline and the remaining volume is delivered at the end of the pipeline, we call this system a pipeline with intermediate delivery points. A more complex case with gas ﬂow into the pipeline (injection) at various points along its length combined with deliveries at other points is shown in Figure 3.1. In such a pipeline system, the pressure required at the beginning point A will be calculated by considering the pipeline broken into segments AB, BC, etc. Another piping system can consist of gas ﬂow at the inlet of the pipeline along with multiple pipe branches making deliveries of gas, as shown in Figure 3.2. In this system, pipe AB has a certain volume, Q1, ﬂowing through it. At point B, another pipeline, CB, brings in additional volumes resulting in a volume of (Q1 + Q2) ﬂowing through section BD. At D, a branch pipe, DE, delivers a volume of Q3 to a customer location, E. The remaining volume (Q1 + Q2 – Q3) ﬂows from D to F through pipe segment DF to a customer location at F. In the subsequent sections, we will analyze pipelines with intermediate ﬂow deliveries, injections, as well as branch pipes, as shown in Figure 3.1 and Figure 3.2. The objectives in all cases will be to calculate the pressures and ﬂow rates through the various pipe sections and to determine pipe sizes required to limit pressure drop in certain pipe segments.

100 MMSCFD A

80 MMSCFD

20 MMSCFD Figure 3.1

C

B

NPS 16

30 MMSCFD

Pipeline with injection and deliveries.

Copyright 2005 by Taylor & Francis Group, LLC

90 MMSCFD

50 MMSCFD E

40 MMSCFD

D

2785_C003.fm Page 94 Friday, April 1, 2005 3:09 PM

94

GAS PIPELINE HYDRAULICS

C E

Q2

Q3

Q1

A

Q1 + Q2

10 in. – 10 mi

B

F

D Q1 + Q2 – Q3

Figure 3.2

Pipeline with branches.

Example 2 A 150 mi long natural gas pipeline consists of several injections and deliveries as shown in Figure 3.3. The pipeline is NPS 20, has 0.500 in. wall thickness, and has an inlet volume of 250 MMSCFD. At points B (milepost 20) and C (milepost 80), 50 MMSCFD and 70 MMSCFD, respectively, are delivered. At D (milepost 100), gas enters the pipeline at 60 MMSCFD. All streams of gas may be assumed to have a speciﬁc gravity of 0.65 and a viscosity of 8.0 × 10−6 lb/ft-s. The pipe is internally coated (to reduce friction), resulting in an absolute roughness of 150 µ in. Assume a constant gas ﬂow temperature of 60°F and base pressure and base temperature of 14.7 psia and 60°F, respectively. Use a constant compressibility factor of 0.85 throughout. Neglect elevation differences along the pipeline. a) Using the AGA equation, calculate the pressures along the pipeline at points A, B, C, and D for a minimum delivery pressure of 300 psig at the terminus E. Assume a drag factor = 0.96. b) What diameter pipe will be required for section DE if the required delivery pressure at E is increased to 500 psig? The inlet pressure at A remains the same as calculated above.

250 MMSCFD A

200 MMSCFD C

B

190 MMSCFD

130 MMSCFD D

NPS 20 0.500 in. wall

50 MMSCFD Figure 3.3

70 MMSCFD

Example pipeline with injection and deliveries.

Copyright 2005 by Taylor & Francis Group, LLC

E 300 psig

60 MMSCFD

2785_C003.fm Page 95 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

95

Solution We will start calculations beginning with the last segment DE. Pipe inside diameter D = 20 – 2 × 0.500 = 19.00 in. The ﬂow rate in pipe DE is 190 MMSCFD. Using Equation 2.34, the Reynolds number is 14.7 0.65 × 190 × 10 6 R = 0.0004778 = 10,974,469 520 8 × 10 −6 × 19 Next, calculate the two transmission factors required per AGA. 1) The fully turbulent transmission factor, using Equation 2.48, is 3.7 × 19 F = 4 Log10 = 22.68 150 × 10 −6 2) The smooth pipe zone Von Karman transmission factor, using Equation 2.50, is 10, 974, 469 Ft = 4 Log10 − 0.6 Ft Solving for Ft by iteration, we get Ft = 22.18 Therefore, for a partly turbulent ﬂow zone, the transmission factor, using Equation 2.49, is 10, 974, 469 F = 4 × 0.96Log10 = 21.29 1.4125 × 22.18 Using the smaller of the two values, the AGA transmission factor is F = 21.29 Next, we use General Flow Equation 2.4 to calculate the upstream pressure P1 at D, based on a given downstream pressure of 300 psig at E. P12 − 314.72 520 190 × 10 = 38.77 × 21.29 14.7 0.65 × 520 × 50 × 0.85 6

Copyright 2005 by Taylor & Francis Group, LLC

0.5

192.5

2785_C003.fm Page 96 Friday, April 1, 2005 3:09 PM

96

GAS PIPELINE HYDRAULICS

Solving for P1, we get the pressure at D as P1 = 587.11 psia = 572.41 psig Next, we consider the pipe segment CD, which has a ﬂow rate of 130 MMSCFD. We calculate the pressure at C using the downstream pressure at D calculated above. To simplify calculation, we will use the same AGA transmission factor we calculated for segment DE. A more nearly correct solution will be to calculate the Reynolds number and the two transmission factors as we did for the segment DE. However, for simplicity, we will use F = 21.29 for all pipe segments. Applying General Flow Equation 2.4, we calculate the pressure P1 at C as follows: P12 − 587.112 520 130 × 10 = 38.77 × 21.29 14.7 0.65 × 520 × 20 × 0.85

0.5

(19.0)2.5

6

Solving for P1, we get the pressure at C as P1 = 625.06 psia = 610.36 psig Similarly, we calculate the pressure at B, considering the pipe segment BC that ﬂows 200 MMSCFD. P12 − 625.062 520 200 × 10 = 38.77 × 21.29 14.7 0.65 × 520 × 60 × 0.85

0.5

(19.0)2.5

6

Solving for P1, we get the pressure at B as P1 = 846.95 psia = 832.25 psig Finally, for pipe segment AB that ﬂows 250 MMSCFD, we calculate the pressure P1 at A as follows: P12 − 846.952 520 250 × 10 = 38.77 × 21.29 14.7 0.65 × 520 × 20 × 0.85 6

0.5

(19.0)2.5

Solving for P1, we get the pressure at A as P1 = 942.04 psia = 927.34 psig If we maintain the same inlet pressure, 927.34 psig, at A and increase the delivery pressure at E to 500 psig, we can determine the pipe diameter required for section DE by considering the same upstream pressure of 572.41 psig at D, as we calculated before.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 97 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

97

Therefore, for segment DE, Upstream pressure P1 = 572.41 + 14.7 = 587.11 psia Downstream pressure P2 = 500 + 14.7 = 514.7 psia Using General Flow Equation 2.4, with the same AGA transmission factor as before, we get 520 587.112 − 514.72 190 × 10 = 38.77 × 21.29 14.7 0.65 × 520 × 50 × 0.85

0.5

6

( D)2.5

Solving for the inside diameter D of pipe DE, we get D = 23.79 in. The nearest standard pipe size is NPS 26 with 0.500 in. wall thickness. This will give an inside diameter of 25 in., which is slightly more than the required minimum of 23.79 in. calculated above. The wall thickness required for this pipe diameter and pressure will be dictated by the pipe material and is the subject of Chapter 6. Example 3 A pipeline 100 mi long transports natural gas from Corona to Beaumont. The gas has a speciﬁc gravity of 0.60 and a viscosity of 8 × 10−6 lb/ft-s. What is the minimum pipe diameter required to ﬂow 100 MMSCFD from Corona to Beaumont for a delivery pressure of 800 psig at Beaumont and inlet pressure of 1400 psig at Corona? The gas can be assumed constant at 60°F, and the base pressure and base temperature are 14.7 psia and 60°F, respectively. Use a constant value of 0.90 for the compressibility factor and a pipe roughness of 700 µ in. Compare results using the AGA, ColebrookWhite, Panhandle B, and Weymouth equations. Use 95% pipeline efﬁciency. Neglect elevation differences along the pipeline. How will the result change if the elevation at Corona is 100 ft and at Beaumont is 500 ft? Solution We will ﬁrst use the AGA equation to determine the pipe diameter. Since the transmission factor F depends on the Reynolds number, which depends on the unknown pipe diameter, we will ﬁrst assume a value of F = 20. From General Flow Equation 2.4, we get 520 1414.72 − 814.772 100 × 10 6 = 38.77 × 20.0 14.7 0.6 × 520 × 100 × 0.9

Copyright 2005 by Taylor & Francis Group, LLC

0.5

× ( D)2.5

2785_C003.fm Page 98 Friday, April 1, 2005 3:09 PM

98

GAS PIPELINE HYDRAULICS

Solving for diameter D, D = 12.28 in. or NPS 12 with a 0.250 in. wall thickness, approximately. Next, we will recalculate the transmission factor using this pipe size. Using NPS 12 with a 0.250 in. wall thickness, Inside pipe diameter D = 12.75 – 2 × 0.250 = 12.25 in. Calculating the Reynolds number from Equation 2.34, we get 14.7 0.6 × 100 × 10 6 = 8,269,615 R = 0.0004778 520 8 × 10 −6 × 12.25 The fully turbulent transmission factor, using Equation 2.48, is 3.7 × 12.25 = 19.25 F = 4 Log10 0.0007 For the smooth pipe zone, using Equation 2.50, the Von Karman transmission factor is 8, 269, 615 Ft = 4 Log10 − 0.6 Ft Solving for Ft by iteration, we get Ft = 21.72 Using a drag factor of 0.96, for partly turbulent ﬂow, the transmission factor is, from Equation 2.49, 8, 269, 615 F = 4 × 0.96 × Log10 = 20.85 1.4125 × 21.72 Using the lower of the two values, the AGA transmission factor is F = 19.25 Using this value of F, we recalculate the minimum pipe diameter from General Flow Equation 2.4 as follows: 520 1414.72 − 814.72 100 × 10 6 = 38.77 × 19.25 14.7 0.6 × 520 × 100 × 0.9

Copyright 2005 by Taylor & Francis Group, LLC

0.5

× ( D)2.5

2785_C003.fm Page 99 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

99

Solving for diameter D, D = 12.47 in. We will not continue iteration any further, since the new diameter will not change the value of F appreciably. Therefore, based on the AGA equation, the pipe inside diameter required is 12.47 in. Next, we calculate the transmission factor based on the Colebrook-White equation, assuming an inside diameter of 12.25 in. and the Reynolds number = 8,269,615, calculated earlier. Using Colebrook-White Equation 2.45, we get 0.0007 1.255F F = −4 Log10 + 3.7 × 12.25 8, 269, 615 Solving for F by successive iteration, we get the Colebrook-White transmission factor as F = 18.95 Using the General Flow equation with this Colebrook-White transmission factor, we calculate the diameter as follows: 520 1414.72 − 814.72 100 × 10 = 38.77 × 18.95 14.7 0.6 × 520 × 100 × 0.9 6

0.5

× ( D)2.5

Solving for diameter D, D = 12.55 in. Recalculating the Reynolds number and transmission factor using the pipe inside diameter of 12.55 in., we get R = 8,071,935 and F = 18.94 Therefore, the new diameter required is by proportions, using the General Flow equation, D 12.55

2.5

18.95 = 18.94

or D = 12.55, approximately. There is no appreciable change in the diameter required. Therefore, based on the Colebrook-White equation, the pipe inside diameter required is D = 12.55 in.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 100 Friday, April 1, 2005 3:09 PM

100

GAS PIPELINE HYDRAULICS

Next, we determine the diameter required using Panhandle B Equation 2.59 and a pipeline efﬁciency of 0.95: 520 100 × 10 6 = 737 × 0.95 14.7

1.02

1414.72 − 814.72 0.961 × 520 × 100 × 0.9 0.6

0.51

D 2.53

Solving for diameter D, we get D = 11.93 in. Therefore, based on the Panhandle B equation, the pipe inside diameter required is 11.93 in. Next, we calculate the diameter required, using Weymouth Equation 2.52 and a pipeline efﬁciency of 0.95: 520 1414.72 − 814.772 100 × 10 6 = 433.5 × 0.95 14.7 0.6 × 520 × 100 × 0.9

0.5

D 2.667

Solving for diameter D, we get D = 13.30 in. Therefore, based on the Weymouth equation, the pipe inside diameter required is 13.30 in. In summary, the minimum pipe inside diameter required based on the various ﬂow equations is as follows: AGA — D = 12.47 in. Colebrook-White — D = 12.55 in. Panhandle B — D = 11.93 in. Weymouth equation — D = 13.30 in. It can be seen that the Weymouth equation is the most conservative equation. The AGA and Colebrook-White equations predict almost the same pipe size, while Panhandle B predicts the smallest pipe size. To further illustrate the comparison of various pressure drop equations, refer to the discussion in Chapter 2 and Figure 2.5, which shows how the delivery pressure varies for a ﬁxed ﬂow rate and inlet pressure. Table 3.1 also summarizes the various pressure drop equations used in the gas pipeline industry. Considering elevation effects, with a single slope from Corona (100 ft) to Beaumont (500 ft), the elevation adjustment parameter is, from Equation 2.10, 500 − 100 s = 0.0375 × 0.6 = 0.0192 520 × 0.9

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 101 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

101

Table 3.1 Summary of Pressure Drop Equations Equation

Application

General Flow

Fundamental flow equation using friction or transmission factor; used with Colebrook-White friction factor or AGA transmission factor

Colebrook-White

Friction factor calculated for pipe roughness and Reynolds number; most popular equation for general gas transmission pipelines

Modified Colebrook-White

Modified equation based on U.S. Bureau of Mines experiments; gives higher pressure drop compared to original Colebrook equation

AGA

Transmission factor calculated for partially turbulent and fully turbulent flow considering roughness, bend index, and Reynolds number

Panhandle A Panhandle B

Panhandle equations do not consider pipe roughness; instead, an efficiency factor is used; less conservative than Colebrook or AGA

Weymouth

Does not consider pipe roughness; uses an efficiency factor used for high-pressure gas gathering systems; most conservative equation that gives highest pressure drop for given flow rate

IGT

Does not consider pipe roughness; uses an efficiency factor used on gas distribution piping

Therefore, the equivalent length from Equation 2.9 is Le = 100 ×

e 0.0192 − 1 = 100.97 mi 0.0192

We will apply the elevation correction factor for the extreme cases (Weymouth and Panhandle B equations) that produce the largest and the smallest diameter, respectively. From Weymouth Equation 2.52, we see that, keeping all other items the same, the diameter and pipe length are related by the following equation: D 2.667 L D 13.3

2.667

= Constant 100.97 = 100

0.5

Solving for the pipe inside diameter D, we get D = 13.32 in. This is not an appreciable change from the previous value of 13.30 in. Similarly, from Panhandle B Equation 2.59, we see that the pipe diameter and length are related by D 2.53 = Constant L0.51 D 11.93

Copyright 2005 by Taylor & Francis Group, LLC

2.53

100.97 = 100

0.51

2785_C003.fm Page 102 Friday, April 1, 2005 3:09 PM

102

GAS PIPELINE HYDRAULICS

Solving for pipe inside diameter D, we get D = 11.95 This is not an appreciable change from the previous value of 11.93 in. Therefore, considering elevation difference between Corona and Beaumont, the minimum pipe sizes required are as follows: Panhandle B — D = 11.95 in. Weymouth equation — D = 13.32 in. We thus see that even with a 400 ft elevation difference, the pipe diameter does not change appreciably. Example 4 A natural gas distribution piping system consists of NPS 12 with 0.250 in. wall thickness, 24 mi long, as shown in Figure 3.4. At Yale, an inlet ﬂow rate of 65 MMSCFD of natural gas enters the pipeline at 60°F. At the Compton terminus, gas must be supplied at a ﬂow rate of 30 MMSCFD at a minimum pressure of 600 psig. There are intermediate deliveries of 15 MMSCFD at milepost 10 and 20 MMSCFD at milepost 18. What is the required inlet pressure at Yale? Use a constant friction factor of 0.01 throughout. The compressibility factor can be assumed to be 0.94. The gas gravity and viscosity are 0.6 and 7 × 10 −6 lb/ft-s, respectively. Assume isothermal ﬂow at 60°F. The base temperature and base pressure are 60°F and 14.7 psia, respectively. If the delivery volume at B is increased to 30 MMSCFD and other deliveries remain the same, what increased pressure is required at Yale to maintain the same ﬂow rate and delivery pressure at Compton? Neglect elevation differences along the pipeline. Solution For each section of piping, such as AB, we must calculate the pressure drop due to friction at the appropriate ﬂow rate and then determine the total pressure drop for the entire pipeline. Inside diameter of pipe = 12.75 – 2 × 0.250 = 12.25 in. Friction factor f = 0.01

65 MMSCFD Yale

NPS 12 , 0.250 in. wall thickness B

m.p. 0.0

C m.p. 18.0

m.p. 10.0

A 20 MMSCFD

Yale to Compton gas distribution pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

600 psig

Compton m.p. 24.0 D

15 MMSCFD Figure 3.4

30 MMSCFD

2785_C003.fm Page 103 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

103

Therefore, the transmission factor, using Equation 2.42, is F=

2 0.01

= 20.00

Using General Flow Equation 2.7, for the last pipe segment from milepost 18 to milepost 24, we get PC2 − 614.72 520 30 × 10 = 38.77 × 20.0 14.7 0.6 × 520 × 6 × 0.94

0.5

6

× (12.25)2.5

Solving for the pressure at C, PC = 620.88 psia Next we will use this pressure PC to calculate the pressure PB for the 8 mi section of pipe segment BC ﬂowing 50 MMSCFD. Using General Flow Equation 2.7, PB2 − 620.882 520 50 × 10 = 38.77 × 20 14.7 0.6 × 520 × 8 × 0.94

0.5

× (12.25)2.5

6

Solving for PB, we get PB = 643.24 psia Finally, we calculate the pressure P1 at Yale by considering the 10 mi pipe segment from Yale to point B that ﬂows 65 MMSCFD. P12 − 643.24 2 520 65 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × 10 × 0.94

0.5

× (12.25)2.5

Solving for the pressure at Yale, we get P1 = 688.09 psia = 673.39 psig Therefore, the required inlet pressure at Yale is 673.39 psig. When the delivery volume at B is increased from 15 to 30 MMSCFD and all other delivery volumes remain the same, the inlet ﬂow rate at Yale will increase to 65 + 15 = 80 MMSCFD. If the delivery pressure at Compton is to remain the same as before, the pressures at B and C will also be the same as calculated before, since the ﬂow rate in BC and CD are the same as before. Therefore, we can recalculate the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 104 Friday, April 1, 2005 3:09 PM

104

GAS PIPELINE HYDRAULICS

inlet pressure for the pipe section from Yale to point B considering a ﬂow rate of 80 MMSCFD that causes a pressure of 643.24 psia at B. Using General Flow Equation 2.7, the pressure P1 at Yale is P12 − 643.24 2 520 80 × 10 6 = 38.77 × 20.0 14.7 0.6 × 520 × 10 × 0.94

0.5

× (12.25)2.5

Solving for the pressure at Yale, P1 = 710.07 psia = 695.37 psig Therefore, increasing the delivery volume at B by 15 MMSCFD causes the pressure at Yale to increase by approximately 22 psig.

3.6 SERIES PIPING In the preceding discussions we assumed the pipeline to have the same diameter throughout its length. There are situations where a gas pipeline can consist of different pipe diameters connected together in a series. This is especially true when the different pipe segments are required to transport different volumes of gas, as shown in Figure 3.5. In Figure 3.5, section AB with a diameter of 16 in. is used to transport a volume of 100 MMSCFD, and after making a delivery of 20 MMSCFD at B, the remainder of 80 MMSCFD ﬂows through the 14 in. diameter pipe BC. At C, a delivery of 30 MMSCFD is made, and the balance volume of 50 MMSCFD is delivered to the terminus D through a 12 in. pipeline CD. It is clear that the pipe section AB ﬂows the largest volume (100 MMSCFD), whereas the pipe segment CD transports the least volume (50 MMSCFD). Therefore, segments AB and CD, for reasons of economy, should be of different pipe diameters, as indicated in Figure 3.5. If we maintained the same pipe diameter of 16 in. from A to D, it would be a waste of pipe material and, therefore, cost. Constant diameter

100 MMSCFD

A

NPS 16

80 MMSCFD

B

20 MMSCFD Figure 3.5

Series piping.

Copyright 2005 by Taylor & Francis Group, LLC

NPS 14

50 MMSCFD

C

30 MMSCFD

NPS 12

D

2785_C003.fm Page 105 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

105

is used only when the same ﬂow that enters the pipeline is also delivered at the end of the pipeline, with no intermediate injections or deliveries. However, in reality, there is no way of determining ahead what the future delivery volumes will be along the pipeline. Hence, it is difﬁcult to determine initially the different pipe sizes for each segment. Therefore, in many cases you will ﬁnd that the same-diameter pipe is used throughout the entire length of the pipeline even though there are intermediate deliveries. Even with the same nominal pipe diameter, different pipe sections can have different wall thicknesses. Therefore, we have different pipe inside diameters for each pipe segment. Such wall thickness changes are made to compensate for varying pressures along the pipeline. The subject of pipe strength and its relation to pipe diameter and wall thickness is discussed in Chapter 6. The pressure required to transport gas in a series pipeline from point A to point D in Figure 3.5 is calculated by considering each pipe segment such as AB and BC and applying the appropriate ﬂow equation, such as the General Flow equation, for each segment, as illustrated in Example 5. Another approach to calculating the pressures in series piping systems is to use the equivalent length concept. This method can be applied when the same uniform ﬂow exists throughout the pipeline, with no intermediate deliveries or injections. We will explain this method of calculation for a series piping system with the same ﬂow rate Q through all pipe segments. Suppose the ﬁrst pipe segment has an inside diameter D1 and length L1, followed by the second segment of inside diameter D2 and length L2 and so on. We calculate the equivalent length of the second pipe segment based on the diameter D1 such that the pressure drop in the equivalent length matches that in the original pipe segment of diameter D2. The pressure drop in diameter D2 and length L2 equals the pressure drop in diameter D1 and equivalent length Le2. Thus, the second segment can be replaced with a piece of pipe of length Le2 and diameter D1. Similarly, the third pipe segment with diameter D3 and length L3 will be replaced with a piece of pipe of Le3 and diameter D1. Thus, we have converted the three segments of pipe in terms of diameter D1 as follows: Segment 1 — diameter D1 and length L1 Segment 2 — diameter D1 and length Le2 Segment 3 — diameter D1 and length Le3

For convenience, we picked the diameter D1 of segment 1 as the base diameter to use, to convert from the other pipe sizes. We now have the series piping system reduced to one constant-diameter (D1) pipe of total equivalent length given by Le = L1 + Le2 + Le3

(3.1)

The pressure required at the inlet of this series piping system can then be calculated based on diameter D1 and length Le. We will now explain how the equivalent length is calculated. Upon examining General Flow Equation 2.7, we see that for the same ﬂow rate and gas properties, neglecting elevation effects, the pressure difference (P12 – P22) is

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 106 Friday, April 1, 2005 3:09 PM

106

GAS PIPELINE HYDRAULICS

inversely proportional to the ﬁfth power of the pipe diameter and directly proportional to the pipe length. Therefore, we can state that, approximately, ∆Psq = where ∆Psq C L D

= = = =

CL D5

(3.2)

difference in the square of pressures (P12 – P22) for the pipe segment a constant pipe length pipe inside diameter

Actually, C depends on the ﬂow rate, gas properties, gas temperature, base pressure, and base temperature. Therefore, C will be the same for all pipe segments in a series pipeline with constant ﬂow rate. Hence, we regard C as a constant for all pipe segments. From Equation 3.2 we conclude that the equivalent length for the same pressure drop is proportional to the ﬁfth power of the diameter. Therefore, in the series piping discussed in the foregoing, the equivalent length of the second pipe segment of diameter D2 and length L2 is CL2 CLe2 = D25 D15

(3.3)

or D Le2 = L2 1 D2

5

(3.4)

Similarly, for the third pipe segment of diameter D3 and length L3 , the equivalent length is D Le3 = L3 1 D3

5

(3.5)

Therefore, the total equivalent length Le for all three pipe segments in terms of diameter D1 is 5

D D Le = L1 + L2 1 + L3 1 D2 D3

5

(3.6)

It can be seen from Equation 3.6 that if D1 = D2 = D3, the total equivalent length reduces to (L1 + L2 + L3), as expected. We can now calculate the pressure drop for the series piping system, considering a single pipe of length Le and uniform diameter D1 ﬂowing a constant volume Q. An example will illustrate the use of the equivalent length method.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 107 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

107

100 MMSCFD 100 MMSCFD

12 mi NPS 16

A

Figure 3.6

8 mi NPS 12

24 mi NPS 14

500 psig B

Example problem—series piping.

Example 5 A series piping system, shown in Figure 3.6, consists of 12 mi of NPS 16, 0.375 in. wall thickness connected to 24 mi of NPS 14, 0.250 in. wall thickness and 8 miles of NPS 12, 0.250 in. wall thickness pipes. Calculate the inlet pressure required at the origin A of this pipeline system for a gas ﬂow rate of 100 MMSCFD. Gas is delivered to the terminus B at a delivery pressure of 500 psig. The gas gravity and viscosity are 0.6 and 0.000008 lb/ft-s, respectively. The gas temperature is assumed constant at 60°F. Use a compressibility factor of 0.90 and the General Flow equation with Darcy friction factor = 0.02. The base temperature and base pressure are 60°F and 14.7 psia, respectively. Compare results using the equivalent length method and with the more detailed method of calculating pressure for each pipe segment separately. Solution Inside diameter of ﬁrst pipe segment = 16 – 2 × 0.375 = 15.25 in. Inside diameter of second pipe segment = 14 – 2 × 0.250 = 13.50 in. Inside diameter of third pipe segment = 12.75 – 2 × 0.250 = 12.25 in. Using Equation 3.6, we calculate the equivalent length of the pipeline, considering NPS 16 as the base diameter: 5

15.25 15.25 Le = 12 + 24 × +8× 13.5 12.25

5

or Le = 12 + 44.15 + 23.92 = 80.07 mi Therefore, we will calculate the inlet pressure P1 considering a single pipe from A to B having a length of 80.07 mi and inside diameter of 15.25 in. Outlet pressure = 500 + 14.7 = 514.7 psia

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 108 Friday, April 1, 2005 3:09 PM

108

GAS PIPELINE HYDRAULICS

Using General Flow Equation 2.2, neglecting elevation effects and substituting given values, we get

(

)

0.5

P12 − 514.72 1 520 2.5 100 × 10 = 77.54 14.7 0.6 × 520 × 80.07 × 0.9 15.25 0.02 6

or P12 – 514.72 = 724,642.99 Solving for the inlet pressure P1, we get P1 = 994.77 psia = 980.07 psig Next, we will compare the preceding result, using the equivalent length method, with the more detailed calculation of treating each pipe segment separately and adding the pressure drops. Consider the 8 mi pipe segment 3 ﬁrst, since we know the outlet pressure at B is 500 psig. Therefore, we can calculate the pressure at the beginning of segment 3 using General Flow Equation 2.2, as follows:

(

)

0.5

2 2 1 520 P1 − 514.7 12.252.5 100 × 10 = 77.54 0.02 14.7 0.6 × 520 × 8 × 0.9 6

Solving for the pressure P1, we get P1 = 693.83 psia = 679.13 psig This is the pressure at the beginning of the pipe segment 3, which is also the end of pipe segment 2. Next, consider pipe segment 2 (24 mi of NPS 14 pipe) and calculate the upstream pressure P1 required for a downstream pressure of 679.13 psig, calculated in the preceding section. Using General Flow Equation 2.2 for pipe segment 2, we get

(

)

0.5

2 2 1 520 P1 − 693.83 2.5 100 × 10 = 77.54 14.7 0.6 × 520 × 24 × 0.9 13.5 0.02 6

Solving for the pressure P1, we get P1 = 938.58 psia = 923.88 psig This is the pressure at the beginning of pipe segment 2, which is also the end of pipe segment 1.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 109 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

109

Next, we calculate the inlet pressure P1 of pipe segment 1 (12 mi of NPS 16 pipe) for an outlet pressure of 923.88 psig, just calculated. Using the General Flow equation for pipe segment 1, we get

(

)

0.5

2 2 1 520 P1 − 938.58 2.5 100 × 10 = 77.54 14.7 0.6 × 520 × 12 × 0.9 15.25 0.02 6

Solving for pressure P1, we get P1 = 994.75 psia = 980.05 psig This compares well with the pressure of 980.07 psig we calculated earlier using the equivalent length method. Example 6 A natural gas pipeline consists of three different pipe segments connected in series, pumping the same uniform ﬂow rate of 3.0 Mm3/day at 20°C. The ﬁrst segment, DN 500 with 12 mm wall thickness, is 20 km long. The second segment is DN 400, 10 mm wall thickness, and 25 km long. The last segment is DN 300, 6 mm wall thickness, and 10 km long. The inlet pressure is 8500 kPa. Assuming ﬂat terrain, calculate the delivery pressure, using the General Flow equation and the Colebrook friction factor of 0.02. The gas gravity = 0.65 and viscosity = 0.000119 Poise. The compressibility factor Z = 0.9. The base temperature = 15°C and base pressure = 101 kPa. Compare results using the equivalent length method as well as the method using individual pipe segment pressure drops. Solution Inside diameter of ﬁrst pipe segment = 500 – 2 × 12 = 476 mm Inside diameter of second pipe segment = 400 – 2 × 10 = 380 mm Inside diameter of last pipe segment = 300 – 2 × 6 = 288 mm Equivalent length method: Using Equation 3.6, we calculate the total equivalent length of the pipeline system based on the ﬁrst segment diameter DN 500 as follows: 5

500 − 2 × 12 500 − 2 × 12 Le = 20 + 25 × + 10 × 400 − 2 × 10 300 − 2 × 6

5

or Le = 20 + 77.10 + 123.33 = 220.43 km Thus, the given pipeline system can be considered equivalent to a single pipe DN 500, 12 mm wall thickness, 220.43 km long.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 110 Friday, April 1, 2005 3:09 PM

110

GAS PIPELINE HYDRAULICS

The outlet pressure P2 is calculated using General Flow Equation 2.3 as follows:

(

)

8500 2 − P22 15 + 273 3 × 10 6 = 1.1494 × 10 −3 101 0.65 × 293 × 0.9 × 0.02 × 220.43

0.5

(476)2.5

Solving for P2, we get 8500 2 – P22 = 25,908,801 or P2 = 6807 kPa (absolute) We have assumed that the given inlet pressure is in absolute value. Therefore, the delivery pressure is 6807 kPa (absolute). Next, we calculate the delivery pressure considering the three pipe segments treated separately. For the ﬁrst pipe segment 20 km long, we calculate the outlet pressure P2 at the end of the ﬁrst segment as follows. Using General Flow Equation 2.3, we get

(

)

8500 2 − P22 15 + 273 3 × 10 = 1.1494 × 10 101 0.65 × 293 × 0.9 × 0.02 × 20 6

−3

0.5

(476)2.5

Solving for P2, we get P2 = 8361 kPa (absolute) Thus, the pressure at the end of the ﬁrst pipe segment or the beginning of the second segment is 8361 kPa (absolute). Next, we repeat the calculation for the second pipe segment DN 400, 25 km long, using P1 = 8361 kPa (absolute), to calculate P2:

(

)

83612 − P22 15 + 273 3 × 10 = 1.1494 × 10 101 0.65 × 293 × 0.9 × 0.02 × 25 6

−3

0.5

(380)2.5

Solving for P2, we get P2 = 7800 kPa (absolute) This is the pressure at the end of the second pipe segment, which is also the inlet pressure for the third pipe segment.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 111 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

111

Finally, we calculate the outlet pressure of the last pipe segment (DN 300, 10 km) using P1 = 7800 kPa (absolute) as follows:

(

)

7800 2 − P22 15 + 273 3 × 10 6 = 1.1494 × 10 −3 101 0.65 × 293 × 0.9 × 0.02 × 10

0.5

(288)2.5

Solving for P2, we get P2 = 6808 kPa (absolute) Therefore, the delivery pressure is 6808 kPa (absolute). This compares favorably with the value of 6807 kPa we calculated earlier using the equivalent length approach.

3.7 PARALLEL PIPING Sometimes two or more pipes are connected such that the gas ﬂow splits among the branch pipes and eventually combines downstream into a single pipe, as illustrated in Figure 3.7. Such a piping system is referred to as parallel pipes. It is also called a looped piping system, where each parallel pipe is known as a loop. The reason for installing parallel pipes or loops is to reduce pressure drop in a certain section of the pipeline due to pipe pressure limitation or for increasing the ﬂow rate in a bottleneck section. By installing a pipe loop from B to E, in Figure 3.7 we are effectively reducing the overall pressure drop in the pipeline from A to F, since between B and E the ﬂow is split through two pipes. In Figure 3.7 we will assume that the entire pipeline system is in the horizontal plane with no changes in pipe elevations. Gas enters the pipeline at A and ﬂows through the pipe segment AB at a ﬂow rate of Q. At the junction B, the gas ﬂow splits into the two parallel pipe branches BCE and BDE at the ﬂow rates of Q1 and Q2, respectively. At E, the gas ﬂows recombine to equal the initial ﬂow rate Q and continue ﬂowing through the single pipe EF. In order to calculate the pressure drop due to friction in the parallel piping system, we follow two main principles of parallel pipes. The ﬁrst principle is that of conservation of ﬂow at any junction point. The second principle is that there is a common pressure across each parallel pipe. C

Q1

Q A

Q B

E D

Figure 3.7

Parallel piping.

Copyright 2005 by Taylor & Francis Group, LLC

Q2

F

2785_C003.fm Page 112 Friday, April 1, 2005 3:09 PM

112

GAS PIPELINE HYDRAULICS

Applying the principle of ﬂow conservation, at junction B, the incoming ﬂow into B must exactly equal the total outﬂow at B through the parallel pipes. Therefore, at junction B, Q = Q1 + Q2

(3.7)

where Q = inlet ﬂow at A Q1 = ﬂow through pipe branch BCE Q2 = ﬂow through pipe branch BDE According to the second principle of parallel pipes, the pressure drop in pipe branch BCE must equal the pressure drop in pipe branch BDE. This is due to the fact that both pipe branches have a common starting point (B) and common ending point (E). Therefore, the pressure drop in the branch pipe BCE and branch pipe BDE are each equal to (PB – PE ), where PB and PE are the pressures at junctions B and E, respectively. Therefore, we can write ∆PBCE = ∆PBDE = PB – PE

(3.8)

∆P represents pressure drop, and ∆PBCE is a function of the diameter and length of branch BCE and the ﬂow rate Q1. Similarly, ∆PBDE is a function of the diameter and length of branch BDE and the ﬂow rate Q2. In order to calculate the pressure drop in parallel pipes, we must ﬁrst determine the ﬂow split at junction B. From Equation 3.7, we know that the sum of the two ﬂow rates Q1 and Q2 must equal the given inlet ﬂow rate Q. If both pipe loops BCE and BDE are equal in length and pipe inside diameter, we can infer that the ﬂow rate will be split equally between the two branches. Thus, for identical pipe loops, Q1 = Q2 =

Q 2

(3.9)

In this case, the pressure drop from B to E can be calculated assuming a ﬂow rate of Q2 ﬂowing through one of the pipe loops. To illustrate this further, suppose we are interested in determining the pressure at A for the given ﬂow rate Q and a speciﬁed delivery pressure (PF) at the pipe terminus F. We start with the last pipe segment EF and calculate the pressure required at E for a ﬂow rate of Q in order to deliver gas at F at a pressure PF . We could use the General Flow equation for this and substitute PE for upstream pressure, P1, and PF for downstream pressure P2. Having calculated PE , we can now consider one of the pipe loops, such as BCE, and calculate the upstream pressure PB required for a ﬂow rate of Q2 through BCE for a downstream pressure of PE. In the General Flow equation, the upstream pressure P1 = PB and the downstream pressure P2 = PE. It must be noted that this is correct only for identical pipe loops. Otherwise, the ﬂow rate Q1 and Q2 through the pipe branches BCE and BDE will be unequal. From the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 113 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

113

calculated value of PB, we can now determine the pressure required at A by applying the General Flow equation to pipe segment AB that has a gas ﬂow rate of Q. The upstream pressure P1 will be calculated for a downstream pressure P2 = PB. Consider now a situation in which the pipe loops are not identical. This means that the pipe branches BCE and BDE can have different lengths and different diameters. In this case, we must determine the ﬂow split between these two branches by equating the pressure drops through each of the branches in accordance with Equation 3.8. Since Q1 and Q2 are two unknowns, we will use the ﬂow conservation principle and the common pressure drop principle to determine the values of Q1 and Q2. From the General Flow equation we can state the following: The pressure drop due to friction in branch BCE can be calculated from

(P

2 B

where K1 = L1 = D1 = Q1 =

)

− PE2 =

K1L1Q12 D15

(3.10)

a parameter that depends on gas properties, gas temperature, etc. length of pipe branch BCE inside diameter of pipe branch BCE ﬂow rate through pipe branch BCE

Other symbols are as deﬁned earlier. K1 is a parameter that depends on the gas properties, gas temperature, base pressure, and base temperature that will be the same for both pipe branches BCE and BDE in a parallel pipeline system. Hence, we regard this as a constant from branch to branch. Similarly, the pressure drop due to friction in branch BDE is calculated from

(P

2 B

where K2 = L2 = D2 = Q2 =

)

− PE2 =

K 2 L2Q22 D25

(3.11)

a constant like K1 length of pipe branch BDE inside diameter of pipe branch BDE ﬂow rate through pipe branch BDE

Other symbols are as deﬁned earlier. In Equation 3.10 and Equation 3.11, the constants K1 and K2 are equal, since they do not depend on the diameter or length of the branch pipes BCE and BDE. Combining both equations, we can state the following for common pressure drop through each branch: L1Q12 L2Q22 = D15 D25

Copyright 2005 by Taylor & Francis Group, LLC

(3.12)

2785_C003.fm Page 114 Friday, April 1, 2005 3:09 PM

114

GAS PIPELINE HYDRAULICS

Simplifying further, we get the following relationship between the two ﬂow rates Q1 and Q2: Q1 L2 = Q2 L1

0.5

D1 D 2

2.5

(3.13)

Combining Equation 3.13 with Equation 3.7, we can solve for the ﬂow rates Q1 and Q2. To illustrate this, consider the inlet ﬂow Q = 100 MMSCFD and the pipe branches as follows: L1 = 10 mi

D1 = 15.5 in. for branch BCE

L2 = 15 mi

D2 = 13.5 in. for branch BDE

From Equation 3.7, for ﬂow conservation, we get Q1 + Q2 = 100 From Equation 3.13, we get the ratio of ﬂow rates as Q1 15 = Q2 10

0.5

15.5 13.5

2.5

= 1.73

Solving these two equations in Q1 and Q2 , we get Q1 = 63.37 MMSCFD Q2 = 36.63 MMSCFD Once we know the values of Q1 and Q2, we can easily calculate the common pressure drop in the branch pipes BCE and BDE. An example problem (Example 7) will be used to illustrate this method. Another method of calculating pressure drops in parallel pipes is using the equivalent diameter. In this method, we replace the pipe loops BCE and BDE with a certain length of an equivalent diameter pipe that has the same pressure drop as one of the branch pipes. The equivalent diameter pipe can be calculated using the General Flow equation, as explained next. The equivalent pipe with the same ∆P that will replace both branches will have a diameter De and a length equal to one of the branch pipes, say L1. Since the pressure drop in the equivalent diameter pipe, which ﬂows the full volume Q, is the same as that in any of the branch pipes, from Equation 3.10, we can state the following:

(P

2 B

)

− PE2 =

Ke LeQ 2 De5

(3.14)

where Q = Q1 + Q2 from Equation 3.7 and Ke represents the constant for the equivalent diameter pipe of length Le ﬂowing the full volume Q.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 115 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

115

Equating the value of (PB2 – PE2) to the corresponding values, considering each branch separately, we get, using Equation 3.10, Equation 3.11, and Equation 3.14: K1L1Q12 K 2 L2Q22 Ke LeQ 2 = = D15 D25 De5

(3.15)

Also, setting K1 = K2 = Ke and Le = L1, we simplify Equation 3.15 as follows: L1Q12 L2Q22 L1Q 2 = = D15 D25 De5

(3.16)

Using Equation 3.16 in conjunction with Equation 3.7, we solve for the equivalent diameter De as 1/ 5

1 + Const1 2 De = D1 Const1

(3.17)

where 5

D L Const1 = 1 2 D2 L1

(3.18)

and the individual ﬂow rates Q1 and Q2 are calculated from Q1 =

QConst1 1 + Const1

(3.19)

Q2 =

Q 1 + Const1

(3.20)

and

To illustrate the equivalent diameter method, consider the inlet ﬂow Q = 100 MMSCFD and the pipe branches as follows: L1 = 10 mi

D1 = 15.5 in. for branch BCE

L2 = 15 mi

D2 = 13.5 in. for branch BDE

From Equation 3.18, 5

15.5 15 Const1 = = 1.73 13.5 10

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 116 Friday, April 1, 2005 3:09 PM

116

GAS PIPELINE HYDRAULICS

Using Equation 3.17, the equivalent diameter is 1/ 5

1 + 1.73 2 De = 15.5 1.73

= 18.60 in.

Thus, the NPS 16 and NPS 14 pipes in parallel can be replaced with an equivalent pipe having an inside diameter of 18.6 in. From Equation 3.19 and Equation 3.20, we get the ﬂow rates in the two branch pipes as follows: Q1 =

100 × 1.73 = 63.37 MMSCFD 1 + 1.73

and Q2 = 36.63 MMSCFD Having calculated an equivalent diameter De, we can now calculate the common pressure drop in the parallel branches by considering the entire ﬂow Q ﬂowing through the equivalent diameter pipe. An example problem will illustrate this method. Example 7 A gas pipeline consists of two parallel pipes, as shown in Figure 3.7. It is designed to operate at a ﬂow rate of 100 MMSCFD. The ﬁrst pipe segment AB is 12 miles long and consists of NPS 16, 0.250 in. wall thickness pipe. The loop BCE is 24 mi long and consists of NPS 14, 0.250 in. wall thickness pipe. The loop BDE is 16 miles long and consists of NPS 12, 0.250 in. wall thickness pipe. The last segment EF is 20 miles long and consists of NPS 16, 0.250 in. wall thickness pipe. Assuming a gas gravity of 0.6, calculate the outlet pressure at F and the pressures at the beginning and the end of the pipe loops and the ﬂow rates through them. The inlet pressure at A = 1200 psig. The gas ﬂowing temperature = 80°F, base temperature = 60°F, and base pressure = 14.73 psia. The compressibility factor Z = 0.92. Use the General Flow equation with Colebrook friction factor f = 0.015. Solution From Equation 3.13, the ratio of the ﬂow rates through the two pipe loops is given by Q1 16 = Q2 24

0.5

14 − 2 × 0.25 12.75 − 2 × 0.25

and from Equation 3.7 Q1 + Q2 = 100

Copyright 2005 by Taylor & Francis Group, LLC

2.5

= 1.041

2785_C003.fm Page 117 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

117

Solving for Q1 and Q2, we get Q1 = 51.0 MMSCFD

and Q2 = 49.0 MMSCFD

Next, considering the ﬁrst pipe segment AB, we will calculate the pressure at B based upon the inlet pressure of 1200 psig at A, using General Flow Equation 2.2, as follows:

(

)

0.5

2 2 1 520 1214.73 − P2 2.5 100 × 10 = 77.54 14.73 0.6 × 540 × 12 × 0.92 15.5 0.015 6

Solving for the pressure at B, we get P2 = 1181.33 psia = 1166.6 psig This is the pressure at the beginning of the looped section at B. Next we calculate the outlet pressure at E of pipe branch BCE, considering a ﬂow rate of 51 MMSCFD through the NPS 14 pipe, starting at a pressure of 1181.33 psia at B. Using the General Flow equation, we get

(

)

0.5

2 2 1 520 1181.333 − P2 2.5 51 × 10 = 77.54 14.73 0.6 × 540 × 24 × 0.92 13.5 0.015 6

Solving for the pressure at E, we get P2 = 1145.63 psia = 1130.9 psig Next, we use this pressure as the inlet pressure for the last pipe segment EF and calculate the outlet pressure at F using the General Flow equation, as follows:

(

)

0.5

2 2 1 520 1145.63 − P2 2.5 100 × 10 6 = 77.54 14.73 0.6 × 540 × 20 × 0.92 15.5 0.015

Solving for the outlet pressure at F, we get P2 = 1085.85 psia = 1071.12 psig In summary, the calculated results are as follows: Pressure at the beginning of pipe loops = 1166.6 psig Pressure at the end of pipe loops = 1130.9 psig Outlet pressure at the end of pipeline = 1071.12 psig Flow rate in NPS 14 loop = 51 MMSCFD Flow rate in NPS 12 loop = 49 MMSCFD

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 118 Friday, April 1, 2005 3:09 PM

118

GAS PIPELINE HYDRAULICS

We will now calculate the pressures using the equivalent diameter method. From Equation 3.18, 5

13.5 16 Const 1 = = 1.041 12.25 24 From Equation 3.17, the equivalent diameter is 1/ 5

1 + 1.041 2 De = 13.5 1.041

= 17.67 in.

Thus, we can replace the two branch pipes between B and E with a single piece of pipe 24 mi long, having an inside diameter of 17.67 in., ﬂowing 100 MMSCFD. The pressure at B was calculated earlier as PB = 1181.33 psia Using this pressure, we can calculate the downstream pressure at E for the equivalent pipe diameter as follows:

(

)

0.5

2 2 1 520 1181.33 − P2 17.672.5 100 × 10 = 77.54 0.015 14.73 0.6 × 540 × 24 × 0.92 6

Solving for the outlet pressure at E, we get P2 = 1145.60 psia, which is almost the same as what we calculated before. The pressure at F will therefore be the same as what we calculated before. Therefore, using the equivalent diameter method, the parallel pipes BCE and BDE can be replaced with a single pipe 24 mi long, having an inside diameter of 17.67 in. Example 8 A natural gas pipeline DN 500 with 12 mm wall thickness is 60 km long. The gas ﬂow rate is 5.0 Mm3/day at 20°C. Calculate the inlet pressure required for a delivery pressure of 4 MPa (absolute), using the General Flow equation with the modiﬁed Colebrook-White friction factor. The pipe roughness = 0.015 mm. In order to increase the ﬂow rate through the pipeline, the entire line is looped with a DN 500 pipeline, 12 mm wall thickness. Assuming the same delivery pressure, calculate the inlet pressure at the new ﬂow rate of 8 Mm3/day. The gas gravity = 0.65 and viscosity = 0.000119 Poise. The compressibility factor Z = 0.88. The base temperature = 15°C, and the base pressure = 101 kPa. If the inlet and outlet pressures are held the same as before, what length of the pipe should be looped to achieve the increased ﬂow?

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 119 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

119

Solution Pipe inside diameter D = 500 − 2 × 12 = 476 mm Flow rate Q = 5.0 × 106 m3/day Base temperature Tb = 15 + 273 = 288 K Gas ﬂow temperature Tf = 20 + 273 = 293 K Delivery pressure P2 = 4 MPa The Reynolds number, using Equation 2.35, is 101 0.65 × 5 × 10 6 R = 0.5134 = 10, 330, 330 288 0.000119 × 476 From the modiﬁed Colebrook-White Equation 2.47, the transmission factor is 0.015 1.4125F F = − 4 Log10 + 3.7 × 476 10, 330, 330 Solving by successive iteration, we get F = 19.80 Using General Flow Equation 2.8, the inlet pressure is calculated next: P12 − 4000 2 273 + 15 5 × 10 6 = 5.747 × 10 −4 × 19.80 101 0.65 × 293 × 60 × 0.88

0.5

× (476)2.5

Solving for the inlet pressure, we get P1 = 5077 kPa (absolute) = 5.08 MPa (absolute) Therefore, the inlet pressure required at 5 Mm3/day ﬂow rate is 5.08 MPa. Next, at 8 Mm3/day ﬂow rate, we calculate the new inlet pressure with the entire 60 km length looped with an identical DN 500 pipe. Since the loop is the same size as the main line, each parallel branch will carry half the total ﬂow rate or 4 Mm3/day. We calculate the Reynolds number for ﬂow through one of the loops using Equation 2.35: 101 0.65 × 4 × 10 6 R = 0.5134 = 8, 264, 264 288 0.000119 × 476

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 120 Friday, April 1, 2005 3:09 PM

120

GAS PIPELINE HYDRAULICS

From the modiﬁed Colebrook-White Equation 2.47, the transmission factor is 0.015 1.4125F F = − 4 Log10 + 3.7 × 476 8, 264, 264 Solving by successive iteration, we get F = 19.70 Keeping the delivery pressure the same as before (4 MPa), using General Flow Equation 2.8, we calculate the inlet pressure required as follows: P12 − 4000 2 273 + 15 4 × 10 = 5.747 × 10 × 19.70 101 . × × × . 0 65 293 60 0 88 6

0.5

−4

× (476)2.5

Solving for the inlet pressure, we get P1 = 4724 kPa (absolute) = 4.72 MPa (absolute) Therefore, for the fully looped pipeline at 8 Mm3/day ﬂow rate, the inlet pressure required is 4.72 MPa. Next, keeping the inlet and outlet pressures the same at 5077 kPa and 4000 kPa, respectively, at the new ﬂow rate of 8 Mm3/day, we assume L km of the pipe from the inlet is looped. We will calculate the value of L by ﬁrst calculating the pressure at the point where the loop ends. Since each parallel pipe carries 4 Mm3/day, we use the Reynolds number and transmission factor calculated earlier: R = 8,264,264

and F = 19.70

Using General Flow Equation 2.8, we calculate the outlet pressure at the end of the loop of length L km as follows: 50772 − P22 273 + 15 4 × 10 6 = 5.747 × 10 −4 × 19.70 101 0.65 × 293 × L × 0.88

0.5

× (476)2.5

Solving for pressure in terms of the loop length L, we get P22 = 50772 − 105, 291.13L

(3.21)

Next, we apply the General Flow equation for the pipe segment of length (60 – L) km that carries the full 8 Mm3/day ﬂow rate. The inlet pressure is P2 and the outlet pressure is 4000 kPa. The Reynolds number at 8 Mm3/day is 101 0.65 × 8 × 10 6 R = 0.5134 = 16, 528, 528 288 0.000119 × 476

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 121 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

121

From the modiﬁed Colebrook-White Equation 2.47, the transmission factor is 0.015 1.4125F F = − 4 Log10 + 3.7 × 476 16, 528, 528 Solving by successive iteration, we get F = 19.96 Using General Flow Equation 2.8, we calculate the inlet pressure for the pipe segment of length (60 − L) km as follows: P22 − 4000 2 273 + 15 8 × 10 6 = 5.747 × 10 −4 × 19.96 101 0.65 × 293 × (60 − L ) × 0.88

0.5

× (476)2.5

Simplifying, we get P22 = 4000 2 + 410, 263.77 (60 − L )

(3.22)

From Equation 3.21 and Equation 3.22, eliminating P2, we solve for L as follows: 50772 − 105, 291.13L = 4000 2 + 410, 263.77 (60 − L ) Therefore, L = 48.66 km Thus, 48.66 km out of the 60 km pipeline length will have to be looped starting at the pipe inlet so that at 8 Mm3/day both inlet and outlet pressures will be the same as before at 5 Mm3/day. What will be the effect if the loop was installed starting at the downstream end of the pipeline and proceeding toward the upstream end? Will the results be the same? In the next section we will explore the best location to install the pipe loop.

3.8 LOCATING PIPE LOOP In the preceding example, we looked at looping an entire pipeline to reduce pressure drop and increase the ﬂow rate. We also explored looping a portion of the pipe, beginning at the upstream end. How do we determine where the loop should be placed for optimum results? Should it be located upstream, downstream, or in a midsection of the pipe? We will analyze this in this section. Three looping scenarios are presented in Figure 3.8. In case (a), a pipeline of length L is shown looped with X miles of pipe, beginning at the upstream end A. In case (b), the same length X of pipe is looped, but it is located on the downstream end B. Case (c) shows the midsection of the pipeline being looped. For most practical purposes, we can say that the cost of all three loops will be the same as long as the loop length is the same.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 122 Friday, April 1, 2005 3:09 PM

122

GAS PIPELINE HYDRAULICS

Flow A

B

C (a) Upstream loop

Flow A

B

C (b) Downstream loop

Flow A

C

D

B

(c) Midsection loop Figure 3.8

Different looping scenarios.

In order to determine which of these cases is optimum, we must analyze how the pressure drop in the pipeline varies with distance from the pipe inlet to outlet. It is found that if the gas temperature is constant throughout, at locations near the upstream end, the pressure drops at a slower rate than at the downstream end. Therefore, there is more pressure drop in the downstream section compared to that in the upstream section. Hence, to reduce the overall pressure drop, the loop must be installed toward the downstream end of the pipe. This argument is valid only if the gas temperature is constant throughout the pipeline. In reality, due to heat transfer between the ﬂowing gas and the surrounding soil (buried pipe) or the outside air (above-ground pipe), the gas temperature will change along the length of the pipeline. If the gas temperature at the pipe inlet is higher than that of the surrounding soil (buried pipe), the gas will lose heat to the soil and the temperature will drop from the pipe inlet to the pipe outlet. If the gas is compressed at the inlet using a compressor, then the gas temperature will be much higher than that of the soil immediately downstream of the compressor. The hotter gas will cause higher pressure drops (examine the General Flow equation and see how the pressure varies with the gas ﬂow temperature). Hence, in this case the upstream segment will have a larger pressure drop compared to the downstream segment. Therefore, considering heat transfer effects, the pipe loop should be installed in the upstream portion for maximum beneﬁt. The installation of the pipe loop in the midsection of the pipeline, as in case (c) in Figure 3.8, will not be the optimum location, based on the preceding discussion. It can therefore be concluded that if the gas temperature is fairly constant along the pipeline, the loop should be installed toward the downstream end, as in case (b). If heat transfer is taken into account and the gas temperature varies along the pipeline, with the hotter gas being upstream, the better location for the pipe loop will be on the upstream end, as in case (a). Looping pipes will be explored more in Chapter 5 and in Chapter 10, where we discuss pipeline economics.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 123 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

123

Pressure

1000 psig

800 psig

Flow B

A

Distance

Figure 3.9

Hydraulic pressure gradient for uniform flow.

3.9 HYDRAULIC PRESSURE GRADIENT

Pressure

The hydraulic pressure gradient is a graphical representation of the gas pressures along the pipeline, as shown in Figure 3.9. The horizontal axis shows the distance along the pipeline starting at the upstream end. The vertical axis depicts the pipeline pressures. Since pressure in a gas pipeline is nonlinear compared to liquid pipelines, the hydraulic gradient for a gas pipeline appears to be a slightly curved line instead of a straight line. The slope of the hydraulic gradient at any point represents the pressure loss due to friction per unit length of pipe. As discussed earlier, this slope is more pronounced as we move toward the downstream end of the pipeline, since the pressure drop is larger toward the end of the pipeline. If the ﬂow rate through the pipeline is a constant value (no intermediate injections or deliveries) and pipe size is uniform throughout, the hydraulic gradient appears to be a slightly curved line, as shown in Figure 3.9, with no appreciable breaks. If there are intermediate deliveries or injections along the pipeline, the hydraulic gradient will be a series of broken lines, as indicated in Figure 3.10.

Q A Q1

Figure 3.10

B

Distance Q2

Hydraulic pressure gradient for deliveries and injections.

Copyright 2005 by Taylor & Francis Group, LLC

Q3

2785_C003.fm Page 124 Friday, April 1, 2005 3:09 PM

124

GAS PIPELINE HYDRAULICS

A similar broken hydraulic gradient can also be seen in the case of a pipeline with variable pipe diameters and wall thicknesses, even if the ﬂow rate is constant. Unlike liquid pipelines, the breaks in hydraulic pressure gradient are not as conspicuous in gas pipelines. In a long-distance gas pipeline, due to limitations of pipe pressure, intermediate compressor stations will be installed to boost the gas pressure to the required value so the gas can be delivered at the contract delivery pressure at the end of the pipeline. For example, consider a 200 mi long, NPS 16 pipeline that transports 150 MMSCFD of gas from Compton to a delivery point at Beaumont, as shown in Figure 3.11. Suppose calculations show that 1600 psig pressure is required at Compton in order to deliver gas to Beaumont at 800 psig. If the maximum operating pressure (MOP) of this pipeline is limited to 1350 psig, obviously we will need more than one compressor station. The ﬁrst compressor station will be located at Compton and will provide a pressure of 1350 psig. As gas ﬂows from Compton toward some intermediate location, such as Sheridan, the gas pressure will drop to some value such as 900 psig. At Sheridan, a second compressor station will boost the gas pressure to 1350 psig on its way to the terminus at Beaumont. By installing the second compressor station at Sheridan, pipeline pressures are maintained within MOP limits. The actual location of the intermediate compressor station at Sheridan will depend upon many factors, including pipeline elevation proﬁle, the gas pressure at Sheridan, and the delivery pressure required at Beaumont. The hydraulic pressure gradient in this case is as shown in Figure 3.11. In the preceding discussion, we picked an arbitrary pressure of 900 psig at Sheridan. This gives us an approximate compression ratio of 1350 + 14.7 = 1.492 914.7 which is a reasonable number for centrifugal compressors used in gas pipelines. In reference to Figure 3.11, we will now outline the method of locating the intermediate 1600 psig 1350 psig

1350 psig

900 psig 800 psig

Flow A Compton

Distance

Sheridan

NPS 16 pipeline 200 mi long Figure 3.11

Compton to Beaumont pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

B Beaumont

2785_C003.fm Page 125 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

125

compressor station at Sheridan. In Chapter 4, we will further analyze gas pipelines with multiple compressor stations. Starting at Compton, with an inlet pressure P1 = 1350 psig, we calculate the pipe length L that will cause the pressure to drop to 900 psig, using the General Flow equation. Assuming a ﬂow rate of 150 MMSCFD and a friction factor f = 0.01, we get 1 520 1364.72 − 914.72 150 × 10 = 77.54 0.01 14.7 0.6 × 520 × L × 0.9

0.5

6

× (15.5)2.5

Solving for L, we get L = 109.28 mi Thus, the approximate location of the second compressor station at Sheridan is 109.28 mi from Compton. If we allow the compressor at Sheridan to boost the gas pressure to 1350 psig, the compression ratio is r=

1350 + 14.7 = 1.492 914.7

which is a reasonable ratio for a centrifugal compressor. Therefore, starting at 1350 psig at Sheridan, we calculate the delivery pressure at Beaumont using the General Flow equation for (200 – 109.28) = 90.7 mi of pipe as follows: 1 520 1364.72 − P2 2 150 × 10 = 77.54 14.7 0.6 × 520 × 90.7 × 0.9 0.01 6

0.5

× (15.5)2.5

Solving for P2, we get P2 = 1005.5 psia = 990.82 psig This is more than the required delivery pressure at Beaumont of 800 psig. We could go back and repeat the above calculations, considering slightly lower pressure at Compton, say 1300 psig, in order to get the correct delivery pressure of 800 psig at Beaumont. This is left as an exercise for the reader. Alternatively, we could start with the required delivery pressure of 800 psig at Beaumont and work backward to determine the distance at which the upstream pressure reaches 1350 psig. That would be the location for the Sheridan compressor station. Next, we would determine the pressure at Sheridan, beginning with the 1350 psig upstream pressure at Compton. This would establish the suction pressure of the Sheridan compressor station. Knowing the suction pressure and the discharge pressure at Sheridan, we could calculate the compression ratio required. In Chapter 4 we will discuss multiple compressor stations in more detail.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 126 Friday, April 1, 2005 3:09 PM

126

GAS PIPELINE HYDRAULICS

3.10 PRESSURE REGULATORS AND RELIEF VALVES In a long-distance gas pipeline with intermediate delivery points, there may be a need to regulate the gas pressure at certain delivery points in order to satisfy the customer requirements. Suppose the pressure at the delivery point B in Figure 3.11 is 800 psig, whereas the customer requirement is only 500 psig. Obviously, some means of reducing the gas pressure must be provided so that the customer can utilize the gas for his or her requirements at the correct pressure. This is achieved by means of a pressure regulator that will ensure a constant pressure downstream of the delivery point, regardless of the pressure on the upstream side of the pressure regulator. This concept is further illustrated using the example pipeline shown in Figure 3.12. The main pipeline from A to C is shown along with a branch pipe BE. The ﬂow rate from A to B is 100 MMSCFD, with an inlet pressure of 1200 psig at A. At B, gas is delivered into a branch line BE at the rate of 30 MMSCFD. The remaining volume of 70 MMSCFD is delivered to the pipeline terminus C at a delivery pressure of 600 psig. Based on the delivery pressure requirement of 600 psig at C and a takeoff of 30 MMSCFD at point B, the calculated pressure at B is 900 psig. Starting with 900 psig on the branch line at B, at 30 MMSCFD, gas is delivered to point E at 600 psig. If the actual requirement at E is only 400 psig, a pressure regulator will be installed at E to reduce the delivery pressure by 200 psig. It can be seen from Figure 3.12 that at point D immediately upstream of the pressure regulator, the gas pressure is approximately 600 psig and is regulated to 400 psig downstream at E. If the mainline ﬂow rate changes from 100 MMSCFD to 90 MMSCFD and the delivery at B is maintained at 30 MMSCFD, the gas pressure at B will reduce to a value below 900 psig. Accordingly, the pressure at point D in the branch pipe BE will also reduce to some value below 600 psig. Regardless, due to the pressure regulator, the pressure at E will be maintained at the required 400 psig. However, if for some reason the pressure upstream of the regulator at D falls below 400 psig, the downstream pressure at E cannot be maintained at the original value of 400 psig. The pressure regulator can only reduce the pressure downstream to the required value. It cannot increase the pressure beyond the pressure on the upstream

400 psig E

D 30 MMSCFD 100 MMSCFD

70 MMSCFD C 600 psig

A 1200 psig Figure 3.12

B

Pressure regulation.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 127 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

127

side. If the pressure at D drops to 300 psig, the pressure regulator is ineffective and will remain fully open, and the delivery pressure at E will be 300 psig as well. Example 9 A natural gas pipeline, NPS 16, 0.250 in. wall thickness, 50 mi long, with a branch pipe (NPS 8, 0.250 in. wall thickness, 15 mi long), as shown in Figure 3.13, is used to transport 100 MMSCFD gas (gravity = 0.6 and viscosity = 0.000008 lb/ft-s) from A to B. At B (milepost 20), a delivery of 30 MMSCFD occurs into the branch pipe BE. The delivery pressure at E must be maintained at 300 psig. The remaining volume of 70 MMSCFD is shipped to the terminus C at a delivery pressure of 600 psig. Assume a constant gas temperature of 60°F and a pipeline efﬁciency of 0.95. The base temperature and base pressure are 60°F and 14.7 psia, respectively. The compressibility factor Z = 0.88. a) Using the Panhandle A equation, calculate the inlet pressure required at A. b) Is a pressure regulator required at E? c) If the inlet ﬂow at A drops to 60 MMSCFD, what is the impact in the branch pipeline BE if the ﬂow rate of 30 MMSCFD is maintained? Solution Pipe inside diameter for pipe segment AB and BC = 16 – 2 × 0.25 = 15.5 in. First, we will consider the pipe segment BC and calculate the pressure P1 at B for a 70 MMSCFD ﬂow rate to deliver gas at 600 psig at C. Using Panhandle A Equation 2.55, neglecting elevation effects, 60 + 460 70 × 10 6 = 435.87 × 0.95 14.7

1.0788

P12 − 614.72 0.8539 × 520 × 30 × 0.88 0.6

0.5394

(15.5)2.6182

300 psig E

D NPS 8

30 MMSCFD

100 MMSCFD

A 1200 psig Figure 3.13

NPS 16

70 MMSCFD C 600 psig B

Example problem—pressure regulation.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 128 Friday, April 1, 2005 3:09 PM

128

GAS PIPELINE HYDRAULICS

Solving for pressure at B, we get P1 = 660.39 psia = 645.69 psig Next, considering pipe segment AB, ﬂowing 100 MMSCFD, we calculate the inlet pressure P1 at A using the outlet pressure 660.39 psia we calculated at B. From Panhandle A Equation 2.55, 520 100 × 10 = 435.87 × 0.95 14.7

1.0788

6

P12 − 660.392 0.8539 × 520 × 20 × 0.88 0.6

0.5394

(15.5)2.6182

Solving for the pressure at A, we get P1 = 715.08 psia = 700.38 psig Next, using the pressure 660.39 psia at B, we calculate the outlet pressure of branch BE that ﬂows 30 MMSCFD through the 15 mi NPS 8 pipe. Using Panhandle A Equation 2.55, 520 30 × 10 6 = 435.87 × 0.95 14.7

1.0788

660.392 − P22 0.8539 × 520 × 15 × 0.88 0.6

0.5394

(8.125)2.6182

Solving for the pressure at E, P2 = 544.90 psia = 530.2 psig Since the required delivery pressure at E is 300 psig, a pressure regulator will be required at E. If the ﬂow rate at A drops to 60 MMSCFD and the branch BE ﬂow rate is maintained at 30 MMSCFD, we will calculate the junction pressure at B by using Panhandle A Equation 2.55 for the pipe segment BC, considering a ﬂow rate of 30 MMSCFD and a delivery pressure of 600 psig at C. 520 30 × 10 = 435.87 × 0.95 14.7

1.0788

6

P12 − 614.72 0.8539 × 520 × 30 × 0.88 0.6

0.5394

(15.5)2.6182

Solving for the pressure at B, we get P1 = 624.47 psia = 609.77 psig Using the pressure at B, we calculate the outlet pressure at E on branch BE for a 30 MMSCFD ﬂow rate using Panhandle A Equation 2.55: 520 30 × 10 6 = 435.87 × 0.95 14.7

Copyright 2005 by Taylor & Francis Group, LLC

1.0788

624.472 − P22 0.8539 × 520 × 15 × 0.88 0.6

0.5394

(8.125)2.6182

2785_C003.fm Page 129 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

129

Solving for the pressure P2, we get P2 = 500.76 psia = 486.06 psig This is the new pressure at E. Comparing this pressure with the previously calculated pressure of 530.2 psig, we see that the delivery pressure at E has dropped by 44 psig, approximately. To maintain the delivery pressure of 300 psig at E, a pressure regulator is still required. Therefore, the answers are a) Inlet pressure at A = 700.38 psig. b) A pressure regulator is required at E to reduce the pressure from 530.2 psig to 300 psig. c) Finally, a pressure regulator is required at E to reduce the pressure from 486.1 psig to 300 psig.

3.11 TEMPERATURE VARIATION AND GAS PIPELINE MODELING

Gas temperature

In the preceding sections we assumed the gas temperature to be constant (isothermal) along the length of the pipeline. By assuming isothermal ﬂow, we were able to calculate the pressure drop using constant gas properties such as the compressibility factor. In reality, the temperature of gas in a buried pipeline varies along the length of the pipeline due to heat transfer between the gas and the surrounding soil. If the inlet temperature of the gas is 80°F and the surrounding soil is 60°F, the temperature difference will cause heat loss from the gas to the soil. Additionally, in a longdistance pipeline, the soil temperature can vary along the pipeline. This will cause the gas temperature to vary, as shown in Figure 3.14.

T Ts

Ground temperature

x Distance

Figure 3.14

Temperature variation in a gas pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 130 Friday, April 1, 2005 3:09 PM

130

GAS PIPELINE HYDRAULICS

Generally, if the pipeline is fairly long, the gas temperature will ultimately equal the soil temperature as gas approaches the delivery point. Due to such variation in gas temperature, calculation of pressure drop must be made by considering short lengths of pipe that make up the total pipeline. For example, if the pipeline is 50 mi long, we will subdivide the pipeline into short segments of 1- or 2-mile lengths and apply the General Flow equation for each pipe segment. Starting with the upstream pressure of segment 1, the downstream pressure will be calculated, assuming an average temperature for segment 1. Next, using the calculated downstream pressure as the upstream pressure for segment 2, we calculate the downstream pressure for segment 2. The process is continued until all segments of the pipeline are covered. It must be noted that the variation of temperature from segment to segment must be taken into account to calculate the compressibility factor to be used in the General Flow equation. The calculation of gas temperature at any point along the pipeline by taking into account the heat transfer between the gas and surrounding soil is quite complicated. It does not lend itself easily to manual calculations. The method of calculation will be discussed brieﬂy in this section for information only. To accurately take into account the temperature variations, a suitable gas pipeline hydraulics simulation program must be used, since, as indicated earlier, manual calculation is quite laborious and time consuming. Several commercial simulation programs are available to model steady-state gas pipeline hydraulics. These programs calculate the gas temperature and pressures by taking into consideration variations of soil temperature, pipe burial depth, and thermal conductivities of pipe, insulation, and soil. One such software program is GASMOD, marketed by SYSTEK Technologies, Inc. (www.systek.us). Appendix A includes a sample simulation of a gas pipeline using the GASMOD software. Even though manual calculation of the temperature variation and corresponding pressure drop in a gas pipeline is quite tedious, we will present here the basic equations for reference. Consider a buried pipeline transporting gas from point A to point B. We will analyze a short segment of length ∆L of this pipe as shown in Figure 3.15 and apply the principles of heat transfer to determine how the gas temperature varies along the pipeline. The upstream end of the pipe segment of length ∆L is at a temperature T1 and the downstream end at temperature T2. The average gas temperature in this segment is

Soil temperature Ts

A

Q

T1

T

T2

L 1

Figure 3.15

Analysis of temperature variation.

Copyright 2005 by Taylor & Francis Group, LLC

2

B

2785_C003.fm Page 131 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

131

represented by T. The outside soil temperature at this location is Ts. Assume steadystate conditions and the mass ﬂow rate of gas to be m. The gas ﬂow from the upstream end to the downstream end of the segment causes a temperature drop of ∆T. The heat loss from the gas can be represented by ∆H = −mCp∆T where ∆H m Cp ∆T

= = = =

(3.23)

heat transfer rate, Btu/h mass ﬂow rate of gas, lb/h average speciﬁc heat of gas, Btu/lb/°F temperature difference = T1 – T2 °F

The negative sign in Equation 3.23 indicates loss of heat from upstream temperature T1 to downstream temperature T2. Next, we consider the heat transfer from the gas to the surrounding soil in terms of the overall heat transfer coefﬁcient U and the difference in temperature between the gas and surrounding soil, represented by (T – Ts). Therefore, we can write the following equation for heat transfer: ∆H = U∆A (T – Ts) where U = ∆A = T = Ts = D =

(3.24)

overall heat transfer coefﬁcient, Btu/h/ft2/ °F surface area of pipe for heat transfer = pD∆L average gas temperature in pipe segment, °F average soil temperature surrounding pipe segment, °F pipe inside diameter, ft

Equating the two values of heat transfer rate ∆H from Equation 3.23 and Equation 3.24, we get −mCp∆T = U∆A(T – Ts) Simplifying, we get πUD ∆T ∆L = − T − Ts mCp

(3.25)

Rewriting Equation 3.25 in differential form and integrating, we get

∫

2

1

dT = T − Ts

πUD

∫ − mCp dL 1

(3.26)

2

Integrating and simplifying, we get T2 − Ts = e −θ T1 − Ts

Copyright 2005 by Taylor & Francis Group, LLC

(3.27)

2785_C003.fm Page 132 Friday, April 1, 2005 3:09 PM

GAS PIPELINE HYDRAULICS

Gas temperature

132

Ts

Ground temperature

Distance Figure 3.16

Joule-Thompson effect in gas pipeline.

where

θ=

πUD∆L mCp

(3.28)

Simplifying Equation 3.27 further, we get the downstream temperature of the pipe segment of length ∆L as T2 = Ts + (T1 − Ts )e −θ

(3.29)

It can be seen from Equation 3.29 that as the pipe length increases, the term e −θ approaches zero and the temperature, T2, becomes equal to soil temperature, Ts. Therefore, in a long gas pipeline, the gas temperature ultimately equals the surrounding soil temperature. This is illustrated in Figure 3.14. In the preceding analysis, we made several simplifying assumptions. We assumed that the soil temperature and the overall heat transfer coefﬁcient remained constant and ignored the Joule-Thompson effect as gas expands through a pipeline. In a long pipeline, the soil temperature can actually vary along the pipeline and, therefore, must be taken into account in these calculations. One approach would be to subdivide the pipeline into segments that have constant soil temperatures and perform calculations for each segment separately. The Joule-Thompson effect causes the gas to cool slightly due to expansion. Therefore, in a long pipeline, the gas temperature at the delivery point may fall below that of the ground or soil temperature, as indicated in Figure 3.16.

3.12 LINE PACK As gas ﬂows through a pipeline from point A to point B, the pressures and temperatures vary along the pipeline length. The volume of gas contained in a given length of pipeline is simply the physical volume of the pipe segment. For example, a 1-mile

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 133 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

133

section of NPS 16 pipe can have a physical volume of 7000 ft3. Therefore, this volume represents the volume of the gas in this 1-mile section at the actual gas temperature and pressure. The quantity of gas contained within the pipeline under pressure, measured at standard conditions (generally 14.7 psia and 60°F), is termed the line pack volume. Consider a segment of pipe, of length L, with upstream pressure and temperature of P1 and T1 and downstream values of P2 and T2, respectively. We can calculate the line pack using the gas laws discussed in Chapter 1. Suppose the inside diameter of the pipe is D; then the physical volume Vp of the pipe section is Vp =

π 2 DL 4

(3.30)

This volume is the gas volume at pressures and temperatures ranging from P1, T1 at the upstream end to P2, T2 at the downstream end of the pipe length L. In order to convert this volume to standard conditions of pressure, Pb, and temperature, Tb, we apply the gas law Equation 1.16 as follows: P V PbVb = avg p Z bTb ZavgTavg where Pavg Tavg Zavg Zb

= = = =

(3.31)

average gas pressure in pipe segment average gas temperature in pipe segment average gas compressibility factor at Tavg and Pavg compressibility factor at base conditions = 1.00, approximately

The average pressure, Pavg, is calculated from the upstream and downstream pressures P1 and P2 using Equation 2.14. The average temperature can be taken as the arithmetic mean of the upstream and downstream temperatures T1 and T2. This approach for average temperature will be accurate only if we consider short segments of pipe. From Equation 3.31, solving for line pack Vb at standard conditions, we get T P V Vb = b avg p Pb ZavgTavg

(3.32)

Substituting the value of Vp from Equation 3.30 and simplifying, we get T P Vb = 0.7854 b avg ( D 2 L ) Pb ZavgTavg where Vb = line pack in pipe segment, standard ft3 D = pipe inside diameter, ft L = pipe segment length, ft Other symbols are as deﬁned before.

Copyright 2005 by Taylor & Francis Group, LLC

(3.33)

2785_C003.fm Page 134 Friday, April 1, 2005 3:09 PM

134

GAS PIPELINE HYDRAULICS

Equation 3.33 is modiﬁed in terms of commonly used pipeline units as follows: T P Vb = 28.798 b avg ( D 2 L ) Pb ZavgTavg

(USCS units)

(3.34)

where Vb = line pack in pipe segment, standard ft3 D = pipe inside diameter, in. L = pipe segment length, mi Other symbols are as deﬁned before. The corresponding equation in SI units is T P Vb = 7.855 × 10 −4 b avg ( D 2 L) Pb ZavgTavg

(SI units)

(3.35)

where Vb = line pack in pipe segment, standard m3 D = pipe inside diameter, mm L = pipe segment length, km Other symbols are as deﬁned before. Since the pressure and temperature in a gas pipeline vary along the length, to improve the accuracy of calculations, the line pack volume Vb is calculated for short segments of pipe and summed to obtain the line pack of the entire pipeline. Example 10 A natural gas pipeline is 10 mi long and has an inlet pressure of 1000 psig and outlet pressure of 900 psig when transporting 100 MMSCFD. The base pressure and base temperature are 14.7 psia and 60°F, respectively. If the pipe is NPS 16, 0.250 in. wall thickness, calculate the line pack assuming an average gas temperature of 78°F. Use an average compressibility factor of 0.90. Solution Pipe inside diameter = 16 – 2 × 0.250 = 15.5 in. The average pressure is calculated from Equation 2.14 as follows:

Pavg =

Copyright 2005 by Taylor & Francis Group, LLC

1014.7 × 914.7 2 1014.7 + 914.7 − = 965.56 psia 1014.7 + 914.77 3

2785_C003.fm Page 135 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

135

Using Equation 3.34, we calculate the line pack as follows: 60 + 460 965.56 15.52 × 10 = 4,880,521 standard ft3 Vb = 28.798 14.7 78 + 460 0.90 Therefore, the line pack is 4,880,521 standard ft3. Example 11 A natural gas pipeline is 20 km long and has an inlet pressure of 8000 kPa (gauge) and outlet pressure of 5000 kPa (gauge) when transporting 5 Mm3/day. The base pressure and base temperature are 101 kPa and 15°C, respectively. If the pipe is DN 500, 12 mm wall thickness, calculate the line pack assuming an average gas temperature of 20°C. Use an average compressibility factor of 0.90. Solution Pipe inside diameter = 500 – 2 × 12 = 476 mm The average pressure is calculated from Equation 2.14 as follows: Pavg =

8101 × 5101 2 8101 + 5101 − = 6714.62 kPa (absolute) 8101 + 5101 3

Using Equation 3.35, we calculate the line pack as follows: 15 + 273 6714.62 4762 × 20 Vb = 7.855 × 10 −4 = 258,448 standard m3 0.9 101 20 + 273 Therefore, the line pack is 258,448 standard m3.

3.13 SUMMARY In this chapter we continued to look at the application of the pressure drop equations introduced in Chapter 2. Several piping conﬁgurations, such as pipes in series, pipes in parallel, and gas pipelines with injections and deliveries, were analyzed to determine pressures required and pipe size needed to satisfy certain requirements. The concepts of equivalent length in series piping and equivalent diameter in pipe loops were explained and illustrated using example problems. The hydraulic pressure gradient and the need for intermediate compressor stations to transport given volumes of gas without exceeding allowable pipeline pressures were also covered. The importance of temperature variation in gas pipelines and how it is taken into account in calculating pipeline pressures were introduced with reference to commercial hydraulic simulation models. The method of calculating the line pack volume in a gas pipeline was also explained. In the next chapter, we will discuss compressor stations, compressor performance, and horsepower requirements.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 136 Friday, April 1, 2005 3:09 PM

136

GAS PIPELINE HYDRAULICS

PROBLEMS 1. A pipeline, NPS 14 with 0.250 in. wall thickness, 40 mi long, transports natural gas (speciﬁc gravity = 0.6 and viscosity = 0.000008 lb/ft-s) at a ﬂow rate of 80 MMSCFD at an inlet temperature of 60°F. Assuming isothermal ﬂow and neglecting elevation changes, calculate the inlet pressure required for a delivery pressure of 800 psig. The base pressure and base temperature are 14.7 psia and 60°F, respectively. Use the Colebrook equation with pipe roughness of 0.0007 in. 2. A 100 mi long natural gas pipeline consists of several injections and deliveries. The pipeline is NPS 18, 0.375 in. wall thickness and has an inlet volume of 150 MMSCFD. At points B (milepost 25) and C (milepost 70), 60 MMSCFD and 50 MMSCFD, respectively, are delivered. At D (milepost 90), gas enters the pipeline at 40 MMSCFD. All streams of gas can be assumed to have a speciﬁc gravity of 0.60 and a viscosity of 7.5 × 10−6 lb/ft-s. The pipe is inter-nally coated such that the absolute roughness is 200 µ in. Assume a constant gas ﬂow temperature of 80°F and base pressure and base temperature of 14.7 psia and 60°F, respectively. Use a constant compressibility factor of 0.88 throughout. Neglect elevation differences along the pipeline. a) Using the modiﬁed Colebrook equation, calculate the pressures along the pipeline at points A, B, C, and D for a minimum delivery pressure of 400 psig at the terminus E. b) What diameter pipe will be required for section DE if the required delivery pressure at E is increased to 600 psig? 3. A natural gas pipeline, 210 km long, consists of an inlet stream at A and deliveries at B and C. The pipeline is DN 400, 10 mm wall thickness. At A, the gas enters at a ﬂow rate of 3.5 Mm3/day. At points B (km 20) and C (km 100), gas is delivered at 0.5 Mm3/day and 1.0 Mm3/day, respectively. At D (km 150), gas enters a branch pipe DF (DN 200, 6 mm wall thickness, 10 km long) at a ﬂow rate of 1.0 Mm3/day. The remaining gas volume of 1.0 Mm3/day is delivered to the pipe terminus E. All streams of gas can be assumed to have a speciﬁc gravity of 0.58 and a viscosity of 0.00012 Poise. The pipe’s absolute roughness is 0.015 mm throughout. Assume a constant gas ﬂow temperature of 15°C and base pressure and base temperature of 101 kPa and 15°C, respectively. Use a pipeline efﬁciency of 0.95 and constant compressibility factor of 0.88 throughout. Neglect elevation differences along the pipeline. a) Using the Panhandle B equation, calculate the pressures along the pipeline at A, B, C, and D for a minimum delivery pressure of 30 Bar at terminus E. b) What is the delivery pressure of gas at the end of the branch DF? c) What pipe diameter is needed for the branch DF if the delivery pressure required at F is 40 Bar? 4. A series piping system consists of 10 mi of NPS 16, 0.250 in. wall thickness, connected to 20 mi of NPS 14, 0.250 in. wall thickness and 10 miles of NPS 12, 0.250 in. wall thickness pipes. Calculate the inlet pressure required at the beginning A for a gas ﬂow rate of 85 MMSCFD. Gas is delivered to the terminus B at

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 137 Friday, April 1, 2005 3:09 PM

PRESSURE REQUIRED TO TRANSPORT

137

a delivery pressure of 600 psig. The gas gravity and viscosity are 0.6 and 0.000008 lb/ft-s, respectively. The gas temperature is assumed constant at 60°F. Use a compressibility factor of 0.85 and the General Flow equation with Colebrook friction factor of 0.015. The base temperature and base pressure are 60°F and 14.7 psia, respectively. 5. A gas pipeline consists of two single pipes with a couple of parallel pipes in the middle. The inlet ﬂow rate is 120 MMSCFD. The ﬁrst pipe segment AB is 10 miles long and consists of NPS 16, 0.250 in. wall thickness pipe. The loop BCE is 20 mi long and consists of NPS 14, 0.250 in. wall thickness pipe. The loop BDE is 15 miles long and consists of NPS 12, 0.250 in. wall thickness pipe. The last segment EF is 18 miles long, NPS 16, 0.250 in. wall thickness pipe. Assuming a gas gravity of 0.6, calculate the outlet pressure at F and the pressures at the beginning and the end of the pipe loops and the ﬂow rates through them. The inlet pressure at A = 1000 psig. The gas ﬂowing temperature = 60°F, base temperature = 60°F, and base pressure = 14.73 psia. The compressibility factor Z = 0.90. Use the AGA fully turbulent equation throughout. 6. A natural gas pipeline is 60 km long. The gas ﬂow rate is 5.0 Mm3/day at 20°C. Calculate the minimum diameter required for an inlet and delivery pressure of 8.5 MPa (absolute) and 5 MPa (absolute), respectively. Use the General Flow equation with the modiﬁed Colebrook-White friction factor. The pipe roughness = 0.020 mm. In order to increase the ﬂow rate through the pipeline, the entire line is looped with an identical-diameter pipeline. Assuming the same delivery pressure, calculate the inlet pressure at the new ﬂow rate of 8 Mm3/day. The gas gravity = 0.60 and viscosity = 0.000119 Poise. The compressibility factor Z = 0.90, base temperature = 15°C, and base pressure = 101 kPa. 7. A natural gas pipeline is 50 mi long and has an inlet pressure of 1200 psig and outlet pressure of 890 psig when transporting 120 MMSCFD. The base pressure and base temperature are 14.7 psia and 60°F, respectively. If the pipe is NPS 16, 0.375 in. wall thickness, calculate the line pack assuming an average gas temperature of 75°F. Use an average compressibility factor of 0.85.

REFERENCES 1. Nayyar, M.L., Piping Handbook, 7th ed., McGraw-Hill, New York, 2000. 2. Baumeister, T., Ed., Standard Handbook for Mechanical Engineers, 7th ed., McGrawHill, New York, 1967. 3. Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003. 4. Westaway, C.R. and Loomis, A.W., Cameron Hydraulic Data, 16th ed., IngersollRand, Montvale, NJ, 1981. 5. Flow of Fluids through Valves, Fittings and Pipe, Crane Company, New York, 1976. 6. McCain, W.D. Jr., The Properties of Petroleum Fluids, Petroleum Publishing Company, Tulsa, OK, 1973. 7. Holman, J.P., Thermodynamics, 2nd ed., McGraw-Hill, New York, 1974.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C003.fm Page 138 Friday, April 1, 2005 3:09 PM

138

GAS PIPELINE HYDRAULICS

8. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 9. Engineering Data Book, 10th ed., Gas Processors Suppliers Association, Tulsa, OK, 1994. 10. Pipeline Design for Hydrocarbon Gases and Liquids, American Society of Civil Engineers, New York, 1975. 11. Steady Flow in Gas Pipelines, Contract Report No. 10, July 1965, Pipeline Research Council International, Houston, TX, 1965.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 139 Saturday, April 2, 2005 9:48 AM

CHAPTER

4

Compressor Stations In this chapter we will discuss compressor stations that are needed to transport gas in a pipeline. The optimum locations and pressures at which compressor stations operate will be analyzed in relation to pipeline throughputs, allowable pipe pressures, and pipeline topography. Centrifugal and positive displacement compressors used in natural gas transportation will be compared with reference to their performance characteristics and cost. Typical compressor station design and equipment used will be covered. Isothermal, adiabatic, and polytropic compression processes and horsepower required will be discussed with illustrative examples. The discharge temperature of compressed gas, its impact on pipeline throughput and gas cooling will be explained.

4.1 COMPRESSOR STATION LOCATIONS Compressor stations are installed on gas pipelines to provide the pressure needed to transport gas from one location to another. Due to limitations of pipeline pressures, multiple compressor stations may be needed to transport a given volume through a long-distance pipeline. The locations and pressures at which these compressor stations operate are determined by allowable pipe pressures, power available, and environmental and geotechnical factors. Consider a pipeline that is designed to transport 100 MMSCFD of natural gas from Dover to a power plant at Leeds, 50 miles away. According to methods outlined in Chapter 3, we would calculate the pressure required at Dover to ensure delivery of the gas at a pressure of 500 psig at Leeds. This calculated pressure at Dover may be more or less than the maximum allowable pipe pressures. Suppose the maximum allowable operating pressure (MAOP) of the pipeline is 1200 psig, whereas the pressure at Dover is calculated to be 1050 psig. It is clear that there is no violation of pressures and, hence, a single compressor station at Dover would suffice to deliver gas to Leeds at the required delivery pressure. If the pipeline length were 100 miles

139

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 140 Saturday, April 2, 2005 9:48 AM

140

GAS PIPELINE HYDRAULICS

instead, calculations would show that in order to deliver the same quantity of gas to Leeds at the same terminus pressure, the pressure required at Dover would have to be 1580 psig. Obviously, since this is greater than the MAOP, we would need more than one compressor station. As a first step, we will assume that an intermediate compressor station will be needed in addition to the one at Dover. The next question is where would this compressor station be located? A logical location would be the midpoint between Dover and Leeds. For simplicity, assume the pipeline elevation profile is fairly flat and, therefore, elevation differences can be ignored. Having selected the location of the intermediate pump station at the midpoint, Kent, as shown in Figure 4.1, we will proceed to determine the pressures at the compressor stations. Since the MAOP is limited to 1200 psig, assume that the compressor at Dover discharges at this pressure. Due to friction, the gas pressure drops as it travels through the pipeline from Dover to Kent, as indicated in Figure 4.1. Suppose the gas pressure reaches 900 psig at Kent and is boosted to 1200 psig by the compressor at Kent. Therefore, the compressor station at Kent is said to have a suction pressure of 900 psig and a discharge pressure of 1200 psig. The gas continues to move from Kent to Leeds, starting at 1200 psig at Kent. As the gas reaches Leeds, the pressure may or may not be equal to the desired pressure of 500 psig. Therefore, if the desired terminus pressure at Leeds is maintained, the pressure at the discharge of the Kent compressor stations may have to be adjusted. Alternatively, Kent could discharge at the same 1200 psig, but its location along the pipeline may have to be adjusted. We selected the 900 psig suction pressure at the Kent compressor station quite arbitrarily. It could have been 700 psig or 1000 psig. The actual number depends upon the “compression ratio” desired. The compression ratio is simply the

1580 psig 1200 psig

1200 psig

900 psig

500 psig Flow Dover

Distance

Kent

NPS 16 pipeline 100 mi long Figure 4.1

Gas pipeline with two compressor stations.

Copyright 2005 by Taylor & Francis Group, LLC

Leeds

2785_C004.fm Page 141 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

141

ratio of the compressor discharge pressure to its suction pressure, both pressures being expressed in absolute units. Compression ratio r =

Pd Ps

(4.1)

where the suction and discharge pressures Ps and Pd are in absolute units. In the present case, the compression ratio for Kent is r=

1200 + 14.7 = 1.33 900 + 14.7

In the above calculation, we assumed the base pressure to be 14.7 psia. If we had chosen a suction pressure of 700 psig, the compression ratio would be r=

1200 + 14.7 = 1.7 700 + 14.7

An acceptable compression ratio for centrifugal compressors is about 1.5. A larger number requires more compressor horsepower, whereas a smaller compression ratio means less horsepower required. In gas pipelines, it is desirable to keep the average pipeline pressure as high as possible to reduce compression power. Therefore, if the suction pressure at Kent is allowed to fall to 700 psig or lower, the average pressure in the pipeline would be lower than if we chose 900 psig for the suction pressure at Kent. Obviously, there is a tradeoff between the number of compressor stations, the suction pressure, and the compression horsepower required. We will discuss this in more detail later in this chapter. Going back to the example problem above, we concluded that we may have to adjust the location of the Kent compressor station or adjust its discharge pressure to ensure the 500 psig delivery pressure at Leeds. Alternatively, we could leave the intermediate compressor station at the halfway point and discharge at 1200 psig, eventually delivering gas to Leeds at 600 psig. If calculations show that by discharging out of Kent results in 600 psig at Leeds, we have satisfied the minimum pressure requirement at Leeds. However, there is extra energy associated with the extra 100 psig delivery pressure. If the power plant can use this extra energy, then there is no waste. On the other hand, if the power plant requirement is 500 psig maximum, then some means of regulating the pressure must be present at the delivery point. This would mean that 100 psig would be reduced through a pressure regulator or control valve at Leeds and energy would be wasted. Another option would be to keep the Kent compressor at the midpoint but reduce the discharge pressure to a number that would result in the requisite 500 psig at Leeds. Since pressure drop in gas pipelines is nonlinear, remembering our discussion in Chapter 3, Kent discharge pressure may have to be reduced by less than 100 psig to provide the fixed 500 psig delivery pressure at Leeds. This would mean that Dover would operate at 1200 psig discharge, whereas Kent would discharge at 1150 psig. This violates our premise of keeping the average pressure in a gas pipeline as high as possible. However, this is

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 142 Saturday, April 2, 2005 9:48 AM

142

GAS PIPELINE HYDRAULICS

still a solution, and in order to pick the best option, we must compare two or more alternative approaches, factoring in the total horsepower required as well as the cost involved. Moving the Kent compressor station slightly upstream or downstream would change the suction and discharge pressures and, hence, the horsepower required. From a cost standpoint, the change would not be significant. However, the horsepower variation would result in change in energy cost and, therefore, in annual operating cost. We must therefore take into account the capital cost and annual operating cost in order to come up with the optimum solution. An example will illustrate this method. In Chapter 10, we will cover several different cost scenarios when dealing with pipeline economics. Example 1 A natural gas pipeline, 140 miles long from Dover to Leeds, is constructed of NPS 16, 0.250 in. wall thickness pipe, with an MOP of 1200 psig. The gas specific gravity and viscosity are 0.6 and 8 × 10−6 lb/ft-s, respectively. The pipe roughness can be assumed to be 700 µ in., and the base pressure and base temperature are 14.7 psia and 60°F, respectively. The gas flow rate is 175 MMSCFD at 80°F, and the delivery pressure required at Leeds is 800 psig. Determine the number and locations of compressor stations required, neglecting elevation difference along the pipeline. Assume Z = 0.85. Solution We will use the Colebrook-White equation to calculate the pressure drop. The Reynolds number is calculated from Equation 2.34 as follows: R=

0.0004778 × 175 × 10 6 × 0.6 × 14.7 = 11,437,412 15.5 × 8 × 10 −6 × 520

Relative roughness =

700 × 10 −6 = 4.5161 × 10−5 15.5

Using Colebrook-White Equation 2.39, we get the friction factor 4.516 × 10 −5 2.51 = −2 Log10 + 3.7 11, 437, 412 f f

1

Solving for f by successive iteration, we get f = 0.0107 Using General Flow Equation 2.2, we calculate the pressure required at Dover as, neglecting elevation effects, 520 P12 − 814.72 1 175 × 10 6 = 77.54 14.7 0.6 × 540 × 140 × 0.85 0.0107

Copyright 2005 by Taylor & Francis Group, LLC

0.5

× (15.5)2.5

2785_C004.fm Page 143 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

143

1579 psig

Pressure

MOP = 1200 psig

800 psig

175 MMSCFD NPS 16 pipeline 140 mi long Leeds

Dover

Figure 4.2

Distance

Pipeline with 1200 psig MOP.

Solving for the pressure at Dover, we get P1 = 1594 psia = 1579.3 psig It can be seen from Figure 4.2 that since the MOP is 1200 psig, we cannot discharge at 1579.3 psig at Dover. We will need to reduce the discharge pressure at Dover to 1200 psig and install an additional compressor station at some point between Dover and Leeds, as shown in Figure 4.3.

1579 psig MOP = 1200 psig

1200 psig

800 psig

175 MMSCFD

Dover

Distance

Kent

NPS 16 pipeline 140 mi long Figure 4.3

Dover to Leeds pipeline with one compressor station.

Copyright 2005 by Taylor & Francis Group, LLC

Leeds

2785_C004.fm Page 144 Saturday, April 2, 2005 9:48 AM

144

GAS PIPELINE HYDRAULICS

We will initially assume that the intermediate compressor station will be located at Kent, halfway between Dover and Leeds. For the pipe segment from Dover to Kent, we will calculate the suction pressure at the Kent compressor station as follows. Using General Flow Equation 2.2, 520 1214.72 − P22 1 175 × 10 = 77.54 14.7 0.6 × 540 × 70 × 0.85 0.0107

0.5

× (15.5)2.5

6

Solving for the pressure at Kent (suction pressure), P2 = 733 psia = 718 psig At Kent, if we boost the gas pressure from 718 psig to 1200 psig (MOP), the .7 compression ratio at Kent is 1214 = 1.66. This is a reasonable compression ratio for 733 a centrifugal compressor. Next, we will see if the 1200 psig pressure at Kent will give the desired 800 psig delivery pressure at Leeds. Considering the 70 mi segment from Kent to Leeds, using the General Flow equation we get 520 1214.72 − P22 1 175 × 10 = 77.54 14.7 0.6 × 540 × 70 × 0.85 0.0107

0.5

6

× (15.5)2.5

resulting in a pressure at Leeds of P2 = 733 psia = 718 psig This is less than the 800 psig desired. Hence, we must move the location of the Kent compressor station slightly toward Leeds so that the 800 psig delivery pressure can be achieved. We will calculate the distance L required between Kent and Leeds. To achieve this, using General Flow Equation 2.2 520 1214.72 − 814.72 1 175 × 10 6 = 77.54 0.0107 14.7 0.6 × 540 × L × 0.85

0.5

× (15.5)2.5

Solving for length L, we get 1214.72 − 814.72 1214.72 − 7332 = L 70 L = 60.57 miles Therefore, Kent must be located approximately 61 miles from Leeds. We must now recalculate the suction pressure at the Kent compressor station based on the pipe

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 145 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

145

length of 79.43 (140 − 60.57) miles between Dover and Kent. From this suction pressure, we must also check the compression ratio. Using General Flow Equation 2.2 for the pipe segment between Dover and Kent, we get 520 1214.72 − P22 1 175 × 10 6 = 77.54 14.7 0.6 × 540 × 79.43 × 0.85 0.0107

0.5

× (15.5)2.5

Solving for P2, we get 1214.72 − P2 2 1214.72 − 7332 = 79.43 70 or P2 = 645.49 psia = 630.79 psig Therefore, the suction pressure at Kent = 630.79 psig. The compression ratio at Kent = 1214.7 = 1.88. 645.49 The compression ratio is slightly more than the 1.5 we would like to see. However, for now, we will go ahead with this compression ratio. Figure 4.4 shows the revised configuration with the new location of the Kent compressor station.

1579 psig MOP = 1200 psig

1200 psig

631 psig

800 psig

175 MMSCFD

Dover

Figure 4.4

Distance

61 miles Kent NPS 16 pipeline 140 mi long

Dover to Leeds pipeline with relocated Kent compressor station.

Copyright 2005 by Taylor & Francis Group, LLC

Leeds

2785_C004.fm Page 146 Saturday, April 2, 2005 9:48 AM

146

GAS PIPELINE HYDRAULICS

4.2 HYDRAULIC BALANCE In the preceding discussions, we considered each compressor station operating at the same discharge pressure and also considered the same compression ratio. Recalling the definition of compression ratio from Equation 4.1, we can state that each compressor station operates at the same suction and discharge pressures. If there are no intermediate injections or deliveries along the pipeline, as in Example 1, each compressor station is required to compress the same amount of gas. Therefore, with pressures and flow rates being the same, each compressor station will require the same amount of horsepower. This is known as hydraulic balance. In a long pipeline with multiple compressor stations, in which each compressor station adds the same amount of energy to the gas, we say that this is a hydraulically balanced pipeline. One of the advantages of a hydraulically balanced pipeline is that all compression equipment can be identical, which will reduce inventory of spare parts and minimize maintenance. It is much easier and cheaper to maintain five identical compressor stations of 5000 horsepower each than to maintain two 6000 HP and three 5000 HP compressors. Also, in order to pump the same volume through a pipeline, hydraulically balanced compressor stations will require less total horsepower than if the stations were not located for hydraulic balance. We will now discuss the different processes by which gas is compressed, such as isothermal, adiabatic (isentropic), and polytropic compression. After that, we will outline the method for calculating horsepower for a compressor station in the subsequent sections.

4.3 ISOTHERMAL COMPRESSION The isothermal compression process is one in which the gas pressure and volume compressed vary in a way that the temperature remains constant. Isothermal compression requires the least amount of work compared to other forms of compression. This process is of theoretical interest since, in reality, maintaining the temperature constant in a gas compressor is virtually impossible. Figure 4.5 shows the pressure volume diagram for isothermal compression. Point 1 represents the inlet conditions of pressure (P1), volume (V1), and at temperature (T1). Point 2 represents the final compressed conditions of pressure (P2), volume (V2), and at constant temperature (T1). The relationship between pressure, P, and volume, V, in an isothermal process is as follows: PV = C

(4.2)

P1V1 = P2V2

(4.3)

where C is a constant. Therefore, we can state that

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 147 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

147

2 Iso

P2

the

al

Pressure, P

rm

P1V1 = P2V2 pr

oc

es

sP

V=

P1

con

sta nt

1 V2

V1 Volume, V

Figure 4.5

Isothermal compression.

Considering 1 lb of natural gas compressed isothermally, the work done is calculated as follows:

Wi =

where Wi G T1 P1 P2 Loge

= = = = = =

P 53.28 T1 Loge 2 G P1

(USCS units)

(4.4)

isothermal work done, ft-lb/lb of gas gas gravity, dimensionless suction temperature of gas, °R suction pressure of gas, psia discharge pressure of gas, psia natural logarithm to base e (e = 2.718)

( )

The ratio PP21 is also called the compression ratio. In SI units, the work done in isothermal compression of 1 kg of gas is Wi = where Wi T1 P1 P2

= = = =

P 286.76 T1 Loge 2 G P1

isothermal work done, J/kg of gas suction temperature of gas, K suction pressure of gas, kPa absolute discharge pressure of gas, kPa absolute

Other symbols are as defined earlier.

Copyright 2005 by Taylor & Francis Group, LLC

(SI units)

(4.5)

2785_C004.fm Page 148 Saturday, April 2, 2005 9:48 AM

148

GAS PIPELINE HYDRAULICS

Example 2 Natural gas is compressed isothermally at 60°F from an initial pressure of 500 psig to a pressure of 1000 psig. The gas gravity is 0.6. Calculate the work done in compressing 5 lb of gas. Use 14.7 psia and 60°F for the base pressure and temperature, respectively. Solution Using Equation 4.4, the work done per lb of gas is Wi =

1000 + 14.7 53.28 = 31,343 ft-lb/lb (60 + 460) Loge 0.6 500 + 14.7

The total work done in compressing 5 lb of gas is WT = 31,343 × 5 = 156,715 ft-lb

Example 3 Calculate the work done in compressing 2 kg of gas (gravity = 0.65) isothermally at 20°C from 700 kPa to 2000 kPa. Use 101 kPa and 15°C for the base pressure and temperature, respectively. Solution Using Equation 4.5, the work done in isothermal compression of 1 kg of gas is Wi =

2000 + 101 286.76 (273 + 20) Loge = 124,649 J/kg 0.65 700 + 101

Therefore, the total work done in compressing 2 kg of gas is WT = 124,649 × 2 = 249,298 J

4.4 ADIABATIC COMPRESSION The adiabatic compression process is characterized by zero heat transfer between the gas and the surroundings. The terms adiabatic and isentropic are used synonymously, although isentropic really means “constant entropy.” An adiabatic process that is also frictionless is referred to as isentropic. In an adiabatic compression process, the relationship between pressure and volume is as follows: PV γ = C

Copyright 2005 by Taylor & Francis Group, LLC

(4.6)

2785_C004.fm Page 149 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

149

where C g = ratio of specific heats of gas, Cvp Cp = specific heats of gas at constant pressure Cv = specific heats of gas at constant volume C = a constant, different from the one for isothermal compression in Equation 4.2 g is also known as the adiabatic or isentropic exponent for the gas. It ranges in value from 1.2 to 1.4. Therefore, we can state that γ PV = P2V2γ 1 1

(4.7)

Figure 4.6 shows adiabatic compression similar to the P-V diagram for isothermal compression. Considering 1 lb of natural gas compressed adiabatically, the work done is calculated as follows: γ −1 53.28 γ P2 γ − 1 Wa = T1 G γ − 1 P1

where Wa G T1 g P1 P2

= = = = = =

(USCS units)

adiabatic work done, ft-lb/lb of gas gas gravity, dimensionless suction temperature of gas, °R ratio of specific heats of gas, dimensionless suction pressure of gas, psia discharge pressure of gas, psia

P2

2 ati iab Ad

P1V1γ = P2V2γ

Pressure, P

ro cp

ce

P1

ss

PV γ =c ons

tant 1

V2

V1 Volume, V

Figure 4.6

Adiabatic compression.

Copyright 2005 by Taylor & Francis Group, LLC

(4.8)

2785_C004.fm Page 150 Saturday, April 2, 2005 9:48 AM

150

GAS PIPELINE HYDRAULICS

In SI units, the work done in adiabatic compression of 1 kg of gas is γ −1 286.76 γ P2 γ − 1 Wa = T1 G γ − 1 P1

where Wa T1 P1 P2

= = = =

(SI units)

(4.9)

adiabatic work done, J/kg of gas suction temperature of gas, K suction pressure of gas, kPa absolute discharge pressure of gas, kPa absolute

Other symbols are as defined earlier. Example 4 Natural gas is compressed adiabatically from an initial temperature and pressure of 60°F and 500 psig, respectively, to a final pressure of 1000 psig. The gas gravity is 0.6 and the ratio of specific heat is 1.3. Calculate the work done in compressing 5 lb of gas. Use 14.7 psia and 60°F for the base pressure and temperature, respectively. Solution Using Equation 4.8, the work done in adiabatic compression is

Wa =

0.3 1.3 1014.7 1.3 53.28 (60 + 460) 1 − = 33,931 ft-lb/lb 0.6 0.3 514.7

Therefore, the total work done in compressing 5 lb of gas is WT = 33,931 × 5 = 169,655 ft-lb Example 5 Calculate the work done in compressing 2 kg of gas (gravity = 0.65) adiabatically from an initial temperature of 20°C and pressure of 700 kPa to a final pressure of 2000 kPa. The specific heat ratio of gas is 1.4 and the base pressure and base temperature are 101 kPa and 15°C, respectively. Solution Using Equation 4.9, the work done in adiabatic compression of 1 kg of gas is 0.4 1.4 2000 + 101 1.4 286.76 1 Wa = (20 + 273) − = 143,512 J/kg 0.65 0.4 700 + 101

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 151 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

151

Therefore, the total work done in compressing 2 kg of gas is WT = 143,512 × 2 = 287,024 J

4.5 POLYTROPIC COMPRESSION Polytropic compression is similar to adiabatic compression, but there is no requirement of zero heat transfer as in adiabatic compression. In a polytropic process, the relationship between pressure and volume is as follows: PV n = C

(4.10)

where n = polytropic exponent C = a constant, different from the one for isothermal or adiabatic compression in Equation 4.2 and Equation 4.6 Therefore, we can state that n n PV 1 1 = P2V2

(4.11)

Since polytropic compression is similar to adiabatic compression, we can easily calculate the work done in polytropic compression by substituting n for g in Equation 4.8 and Equation 4.9. Example 6 Natural gas is compressed polytropically from an initial temperature and pressure of 60°F and 500 psig, respectively, to a final pressure of 1000 psig. The gas gravity is 0.6 and the base pressure and base temperature are 14.7 psia and 60°F, respectively. Calculate the work done in compressing 5 lb of gas using a polytropic exponent of 1.5. Solution Polytropic compression is similar to adiabatic compression, and, therefore, the same equation can be used for work done, substituting the polytropic exponent n for the adiabatic exponent g (the ratio of specific heat). Using Equation 4.8, the work done in polytropic compression of 1 lb of gas is

Wp =

0.5 1.5 1014.7 1.5 53.28 (60 + 460) 1 − = 35,168 ft-lb/lb 0.6 0.5 514.7

Therefore, the total work done in compressing 5 lb of gas is WT = 35,168 × 5 = 175,840 ft-lb

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 152 Saturday, April 2, 2005 9:48 AM

152

GAS PIPELINE HYDRAULICS

Example 7 Calculate the work done in compressing 2 kg of gas (gravity = 0.65) polytropically from an initial temperature of 20°C and pressure of 700 kPa to a final pressure of 2000 kPa. Use a polytropic exponent of 1.5. The base pressure and base temperature are 101 kPa and 15°C, respectively. Solution Using Equation 4.9, and substituting the polytropic exponent 1.5 in place of g, the work done in polytropic compression is

Wp =

0.5 1.5 2000 + 101 1.5 286.76 1 (20 + 273) − = 146,996 J/kg 0.65 0.5 700 + 101

Therefore, the total work done in compressing 2 kg of gas is WT = 146,996 × 2 = 293,992 J

4.6 DISCHARGE TEMPERATURE OF COMPRESSED GAS In adiabatic or polytropic compression of natural gas, we can determine the final temperature of the gas knowing the initial temperature and initial and final pressures. Using Equation 4.6 for adiabatic compression and the perfect gas law, by eliminating the volume, V, we can write the following: T2 Z1 P2 T = Z P 1 2 1 where T1 = T2 = Z1 = Z2 =

γ −1 γ

(4.12)

suction temperature of gas, °R discharge temperature of gas, °R gas compressibility factor at suction, dimensionless gas compressibility factor at discharge, dimensionless

Other symbols are as defined earlier. Similarly, for polytropic compression, the discharge temperature can be calculated from the following equation: T2 Z1 P2 T = Z P 1 2 1 where all symbols are as defined earlier.

Copyright 2005 by Taylor & Francis Group, LLC

n −1 n

(4.13)

2785_C004.fm Page 153 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

153

Example 8 Gas is compressed adiabatically (g = 1.4) from 60°F suction temperature and a compression ratio of 2.0. Calculate the discharge temperature, assuming Z1 = 0.99 and Z2 = 0.85. Solution Using Equation 4.12, 0.4 T2 0.99 1.4 60 + 460 = 0.85 (2.0) = 1.4198

T2 = 1.4198 × 520 = 738.3°R = 278.3°F

4.7 HORSEPOWER REQUIRED The amount of energy input to the gas by the compressors is dependent upon the pressure of the gas and flow rate. The horsepower (HP), which represents the energy per unit time, also depends upon the gas pressure and the flow rate. As the flow rate increases, the pressure also increases and, hence, the horsepower needed will also increase. Since energy is defined as work done by a force, we can state the power required in terms of the gas flow rate and the discharge pressure of the compressor station. Suppose the gas flow rate is Q measured in standard ft3 per day (SCFD), and the suction and discharge pressures of the compressor station are Ps and Pd , respectively. The compressor station adds the differential pressure of (Pd – Ps) psia to the gas flowing at Q SCFD. Therefore, the rate at which energy is supplied to the gas is (Pd – Ps) × Q × Const1, where Const1 is a constant depending upon the units employed. This is a very simplistic approach, since the gas properties vary with temperature and pressure. Also, the compressibility factor and the type of gas compression (adiabatic or polytropic) must be taken into account. Therefore, the calculation for HP will be approached from another angle in what follows. The head developed by the compressor is defined as the amount of energy supplied to the gas per unit mass of gas. Therefore, by multiplying the mass flow rate of gas by the compressor head, we can calculate the total energy supplied to the gas. Dividing this by compressor efficiency, we will get the horsepower required to compress the gas. The equation for horsepower can be expressed as follows: HP = where HP = compressor horsepower M = mass flow rate of gas, lb/min

Copyright 2005 by Taylor & Francis Group, LLC

M × ∆H η

(4.14)

2785_C004.fm Page 154 Saturday, April 2, 2005 9:48 AM

154

GAS PIPELINE HYDRAULICS

∆H = compressor head, ft-lb/lb h = compressor efficiency, decimal value Another more commonly used formula for compressor horsepower that takes into account the compressibility of gas is as follows: γ −1 Z1 + Z 2 1 P2 γ γ HP = 0.0857 QT − 1 γ − 1 1 2 ηa P1

where HP g Q T1 P1 P2 Z1 Z2 ha

= = = = = = = = =

(4.15)

compressor horsepower ratio of specific heats of gas, dimensionless gas flow rate, MMSCFD suction temperature of gas, °R suction pressure of gas, psia discharge pressure of gas, psia compressibility of gas at suction conditions, dimensionless compressibility of gas at discharge conditions, dimensionless compressor adiabatic (isentropic) efficiency, decimal value

In SI units, the Power equation is as follows: γ −1 Z1 + Z 2 1 P2 γ γ Power = 4.0639 QT − 1 γ − 1 1 2 ηa P1

where Power g Q T1 P1 P2 Z1 Z2 ha

= = = = = = = = =

(4.16)

compression Power, kW ratio of specific heats of gas, dimensionless gas flow rate, Mm3/day suction temperature of gas, K suction pressure of gas, kPa discharge pressure of gas, kPa compressibility of gas at suction conditions, dimensionless compressibility of gas at discharge conditions, dimensionless compressor adiabatic (isentropic) efficiency, decimal value

The adiabatic efficiency ha generally ranges from 0.75 to 0.85. By considering a mechanical efficiency hm of the compressor driver, we can calculate the brake horsepower (BHP) required to run the compressor as follows: BHP =

Copyright 2005 by Taylor & Francis Group, LLC

HP ηm

(4.17)

2785_C004.fm Page 155 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

155

where HP is the horsepower calculated from the preceding equations, taking into account the adiabatic efficiency ha of the compressor. The mechanical efficiency hm of the driver can range from 0.95 to 0.98. The overall efficiency, ho, is defined as the product of the adiabatic efficiency, ha , and the mechanical efficiency, hm: ho = ha × hm

(4.18)

From the adiabatic compression Equation 4.6, eliminating the volume V, the discharge temperature of the gas is related to the suction temperature and the compression ratio by means of the following equation: T2 P2 T = P 1 1

γ −1 γ

(4.19)

The adiabatic efficiency, ha, can also be defined as the ratio of the adiabatic temperature rise to the actual temperature rise. Thus, if the gas temperature due to compression increases from T1 to T2, the actual temperature rise is (T2 – T1). The theoretical adiabatic temperature rise is obtained from the adiabatic pressure– temperature relationship as follows, considering the gas compressibility factors similar to Equation 4.12: T2 Z1 P2 T = Z P 1 2 1

γ −1 γ

(4.20)

or Z P T2 = T1 1 2 Z 2 P1

γ −1 γ

(4.21)

Therefore, the theoretical adiabatic temperature rise is Z P T1 1 2 Z 2 P1

γ −1 γ

− T1

Therefore, the adiabatic efficiency is

ηa =

T1

( )( ) Z1 Z2

P2 P1

γ −1 γ

− T1

T2 − T1

where T2 is the actual discharge temperature of the gas.

Copyright 2005 by Taylor & Francis Group, LLC

(4.22)

2785_C004.fm Page 156 Saturday, April 2, 2005 9:48 AM

156

GAS PIPELINE HYDRAULICS

Simplifying, we get γ −1 T1 Z1 P2 γ − ηa = 1 Z P T − T 2 1 2 1

(4.23)

For example, if the inlet gas temperature is 80°F and the suction and discharge pressures are 800 psia and 1400 psia, respectively, we can calculate the adiabatic efficiency if the outlet temperature is given as 200°F. Using g = 1.4, and from Equation 4.23, the adiabatic efficiency is, assuming compressibility factors to be equal to 1.0, 1.4 −1 80 + 460 1400 1.4 = 0.7802 − ηa = 1 200 − 80 800

(4.24)

Thus, the adiabatic compression efficiency is 0.7802. Example 9 Calculate the compressor horsepower required for an adiabatic compression of 106 MMSCFD gas with inlet temperature of 68°F and 725 psia pressure. The discharge pressure is 1305 psia. Assume the compressibility factors at suction and discharge conditions to be Z1 = 1.0 and Z2 = 0.85, respectively, and the adiabatic exponent g = 1.4, with the adiabatic efficiency ha = 0.8. If the mechanical efficiency of the compressor driver is 0.95, what BHP is required? Calculate the outlet temperature of the gas. Solution From Equation 4.15, the horsepower required is 0.40 1.40 1 + 0.85 1 1305 1.40 ( ) HP = 0.0857 × 106 68 460 + 1 − 2 0.8 725 = 3550 0.40

Using Equation 4.17, we calculate the driver horsepower required based on a mechanical efficiency of 0.95: BHP required =

3550 = 3737 0.95

The outlet temperature of the gas is found from Equation 4.23 after transposing as follows: T2 = (68 + 460) ×

( 0.185 )( 1305 725 ) 0.8

0.4 1.4

− 1 + (68 + 460) = 786.46°R = 326.46°F

The discharge temperature of the gas is 326.46°F.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 157 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

157

Example 10 Natural gas at 3 Mm3/day and 20°C is compressed isentropically (g = 1.4) from a suction pressure of 5 MPa absolute to a discharge pressure of 9 MPa absolute in a centrifugal compressor with an isentropic efficiency of 0.80. Calculate the compressor power required, assuming the compressibility factors at suction and discharge conditions to be Z1 = 0.95 and Z2 = 0.85, respectively. If the mechanical efficiency of the compressor driver is 0.95, what is the driver power required? Calculate the outlet temperature of the gas. Solution From Equation 4.16, the power required is 0.40 0.95 + 0.885 1 9 1.40 1.40 1 Power = 4.0639 × 3 ( ) 20 273 + − 0.8 5 = 2572 kW 2 0.40

Power = 2572 kW Using Equation 4.17, we calculate the driver power required as follows: Driver power required =

2572 = 2708 kW 0.95

The outlet temperature of the gas is found from Equation 4.23 as follows: 0.4 20 + 273 0.95 9 1.4 × T2 = − 1 + (20 + 273) = 410.94 K = 137.94°C 0.8 0.85 5

4.8 OPTIMUM COMPRESSOR LOCATIONS In the foregoing discussion, we looked at a two-compressor station configuration for gas deliveries from Dover to the Leeds power plant. In this section, we will consider various locations of the intermediate compressor stations on a long-distance gas transmission pipeline to arrive at the optimum locations, taking into account the overall horsepower required. In Section 4.2, we discussed hydraulic balance. The advantage of locating the intermediate compressor station, such that the same amount of energy is added to the gas at each compressor station, was pointed out. In the next example, we will analyze optimum compressor locations by considering both hydraulically balanced and unbalanced compressor station locations. Example 11 A gas transmission pipeline is 240 mi long, NPS 30, 0.500 in. wall thickness, with an origin compressor station at Payson and two intermediate compressor stations

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 158 Saturday, April 2, 2005 9:48 AM

158

GAS PIPELINE HYDRAULICS

MOP = 1400 psig

1400 psig

600 psig

900 MMSCFD

Payson m.p. 0

Williams m.p. 80

Snowflake m.p. 160

Douglas m.p. 240

NPS 30 pipeline 240 mi long Figure 4.7

Gas pipeline with three compressor stations.

tentatively located at Williams (milepost 80) and Snowflake (milepost 160), as shown in Figure 4.7. There are no intermediate flow deliveries or injections, and the inlet flow rate of 900 MMSCFD at Payson equals the delivery flow rate at Douglas. The delivery pressure required at Douglas is 600 psig and the MOP of the pipeline is 1400 psig throughout. Neglect the effects of elevation and assume constant gas flow temperature of 80°F and constant values of transmission factor F = 20 and compressibility factor Z = 0.85 throughout the pipeline. The gas gravity = 0.6, base pressure = 14.7 psia, and base temperature = 60°F. Use a polytropic compression coefficient of 1.38 and a compression efficiency of 0.9. Solution Neglecting the effects of elevation, we could calculate for each of the three segments— Payson to Williams, Williams to Snowflake, and Snowflake to Douglas—the downstream pressure starting with an upstream pressure of 1400 psig. Thus, using the General Flow equation for the Payson to Williams segment, we would calculate the downstream pressure at Williams starting with a pressure of 1400 psig at Payson. This downstream pressure is actually the suction pressure at the Williams compressor station. Next, in a similar fashion, we would calculate the downstream pressure at Snowflake, for the second segment from Williams to Snowflake, based on an upstream pressure of 1400 psig at Williams. This downstream pressure is actually the suction pressure at the Snowflake compressor station. Finally, we would calculate the downstream pressure at Douglas, for the third segment from Snowflake to Douglas, based on an upstream pressure of 1400 psig at Snowflake. This final pressure is the delivery pressure at the Douglas terminus. We have thus calculated the suction pressures at each of the two intermediate compressor stations at Williams and Snowflake and also calculated the final delivery pressure at Douglas. This pressure calculated at Douglas may or may not be equal to the desired delivery pressure of 600 psig, since we performed a forward calculation going from Payson to Douglas. Therefore, since the delivery pressure is usually a desired or contracted value, we will have to adjust the location of the last compressor station at Snowflake to achieve the desired delivery pressure at Douglas.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 159 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

159

Another approach would be to perform a backward calculation starting at Douglas and proceeding toward Payson. In this case, we will start with segment 3 and calculate the location of the Snowflake compressor station that will result in an upstream pressure of 1400 psig at Snowflake. Thus, we locate the Snowflake compressor station that will cause a discharge pressure of 1400 psig at Snowflake and a delivery pressure of 600 psig at Douglas. Having located the Snowflake compressor station, we can now recalculate the suction pressure at Snowflake by considering the pipe segment 2 and using an upstream pressure of 1400 psig at Williams. We will not have to repeat calculations for segment 1, since the location of Williams has not changed and, therefore, the suction pressure at Williams will remain the same as the previously calculated value. We have thus been able to determine the pressures along the pipeline with the given threecompressor-station configuration such that the desired delivery pressure at Douglas has been achieved and each compressor station discharges at an MOP value of 1400 psig. But are these the optimum locations of the intermediate compressor stations Williams and Snowflake? And are all compressor stations in hydraulic balance? We can state that these compressor stations are optimized and are in hydraulic balance only if each compressor station operates at the same compression ratio and, therefore, adds the same amount of horsepower to the gas at each compressor station. The locations of Williams and Snowflake may not result in the same suction pressures even though the discharge pressures are the same. Therefore, chances are that Williams might be operating at a lower compression ratio than Snowflake or Payson, or vice versa, which will not result in hydraulic balance. However, if the compression ratios are close enough that the required compressor sizes are the same, we could still be in hydraulic balance and the stations could be at optimum locations. Next, perform the actual calculations and determine how much tweaking of the compressor station locations is required to optimize these stations. First, we will perform the backward calculations for segment 3, starting with a downstream pressure of 600 psig at Douglas and an upstream pressure of 1400 psig at Snowflake. With these constraints, we will calculate the pipe length, L, miles between Snowflake and Douglas. Using General Flow Equation 2.4, neglecting elevations, 520 1414.72 − 614.772 900 × 10 = 38.77 × 20.0 14.7 0.6 × 540 × L × 0.85 6

0.5

(29)2.5

Solving for pipe length, we get L = 112.31 mi Therefore, in order to discharge at 1400 psig at Snowflake and deliver gas at 600 psig at Douglas, the Snowflake compressor station will be located at a distance of 112.31 mi upstream of Douglas—or at milepost (240 – 112.31) = 127.69 measured from Payson. Next, keeping the location of the Williams compressor at milepost 80, we will calculate the downstream pressure at Snowflake for pipe segment 2 starting at 1400 psig at Williams. This calculated pressure will be the suction pressure of the Snowflake compressor station.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 160 Saturday, April 2, 2005 9:48 AM

160

GAS PIPELINE HYDRAULICS

Using General Flow Equation 2.4, neglecting elevations, 1414.72 − P2 2 520 900 × 10 = 38.77 × 20.0 14.7 0.6 × 540 × 47.69 × 0.85

0.5

6

(29)2.5

where the pipeline segment length between Williams and Snowflake was calculated as 127.69 – 80 = 47.69 mi Solving for suction pressure at Snowflake, we get P2 = 1145.42 psia = 1130.72 psig Therefore, the compression ratio at Snowflake is

1414.7 = 1.24. 1145.42

Next, for pipe segment 1 between Payson and Williams, we will calculate the downstream pressure at Williams, starting at 1400 psig at Payson. This calculated pressure will be the suction pressure of the Williams compressor station. Using General Flow Equation 2.4, neglecting elevations, 1414.72 − P2 2 520 900 × 10 = 38.77 × 20.0 14.7 0.6 × 540 × 80 × 0.85

0.5

6

(29)2.5

Solving for suction pressure at Williams, we get P2 = 919.20 psia = 904.5 psig Therefore, the compression ratio at Williams =

1414.7 = 1.54 919.2

Therefore, from the foregoing calculations, the compressor station at Williams requires a compression ratio r = 1.54, whereas the compressor station at Snowflake requires a compression ratio r = 1.24. Obviously, this is not a hydraulically balanced compressor station system. Further, we do not know what the suction pressure is at the Payson compressor station. If we assume that Payson receives gas at approximately the same suction pressure as Williams (905 psig), both the Payson and Williams compressor stations will have the same compression ratio of 1.54. In this case, the Snowflake compressor station will be the odd one, operating at a compression ratio of 1.24. How do we balance these compressor stations? One way would be to obtain the same compression ratios for all three compressor stations by simply relocating the Snowflake compressor station toward Douglas such that its suction pressure will drop from 1131 psig to 905 psig while keeping the discharge at Snowflake at 1400 psig. This will then ensure that all three compressor stations will be operating at the following suction and discharge pressures and compression ratios: Suction pressure Ps = 904.5 psig Discharge pressure Pd = 1400 psig Compression ratio r =

Copyright 2005 by Taylor & Francis Group, LLC

1400 + 14.7 = 1.54 904.5 + 14.7

2785_C004.fm Page 161 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

161

MOP = 1400 psig

1400 psig

905 psig ∆P 600 psig 900 MMSCFD

Payson m.p. 0

Williams m.p. 80

Snowflake m.p. 160

Douglas m.p. 240

NPS 30 pipeline 240 mi long Figure 4.8

Pressure regulation at Douglas.

However, because the Snowflake compressor station is now located closer to Douglas than before (127.69), the discharge pressure of 1400 psig at Snowflake will result in a higher delivery pressure at Douglas than the required 600 psig, as shown in Figure 4.8. If the additional pressure at Douglas can be tolerated by the customer, then there will be no problem. But if the customer requires no more than 600 psig, we have to reduce the delivery pressure to 600 psig by installing a pressure regulator at Douglas, as shown in Figure 4.8. Therefore, by balancing the compressor station locations, we have also created a problem of getting rid of the extra pressure at the delivery point. Pressure regulation means wasted horsepower. The advantage of the balanced compressor stations vs. the negative aspect of the pressure regulation must be factored into the decision process. To illustrate this pressure regulation scenario, we will now determine the revised location of the Snowflake compressor station for hydraulic balance. We will calculate the length of pipe segment 2 by assuming 1400 psig discharge pressure at Williams and a suction pressure of 904.5 psig at Snowflake. Using General Flow Equation 2.4, neglecting elevations, 520 1414.72 − 919.222 900 × 10 6 = 38.77 × 20.0 14.7 0.6 × 540 × L × 0.85

0.5

(29)2.5

Solving for pipe length for segment 2, we get L = 80 mi Therefore, the Snowflake compressor station should be located at a distance of 80 mi from Williams or at milepost 160. We could have arrived at this without the above calculations, since elevations are neglected and the Payson to Williams pressure

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 162 Saturday, April 2, 2005 9:48 AM

162

GAS PIPELINE HYDRAULICS

profile will be the same as the pressure profile from Williams to Snowflake. With the Snowflake compressor station located at milepost 160, and discharging at 1400 psig, we conclude that the delivery pressure at Douglas will also be 904.5 psig, since all three pipe segments are hydraulically the same. We see that the delivery pressure at Douglas is approximately 305 psig more than the desired pressure. As indicated earlier, a pressure regulator will be required at Douglas to reduce the delivery pressure to 600 psig. We can compare the hydraulically balanced scenario with the previously calculated case where Payson and Williams operate at a compression ratio of 1.54 and Snowflake operates at lower compression ratio of 1.24. By applying approximate cost per installed horsepower, we can compare these two cases. First, using Equation 4.15, calculate the horsepower required at each compressor station, assuming polytropic compression and a compression ratio of 1.54 for a balanced compressor station: 0.38 1 + 0.85 1 1.38 (80 + 460) HP = 0.0857 × 900 × (1.54) 1.38 − 1 = 19,627 2 0.9 0.38

Therefore, the total horsepower required in the hydraulically balanced case is Total HP = 3 × 19,627 = 58,881 At a cost of $2000 per installed HP, Total HP cost = $2000 × 58,881 = $117.76 million In the hydraulically unbalanced case, the Payson and Williams compressor stations will operate at a compression ratio of 1.54 each, whereas the Snowflake compressor station will require a compression ratio of 1.24. Using Equation 4.15, the horsepower required at the Snowflake compressor station is 0.38 1 + 0.85 1 1.38 1.38 − 1 = 9487 ( ) ( . ) HP = 0.0857 × 900 × 80 460 + 1 24 2 0.9 0.38

Therefore, the total horsepower required in the hydraulically unbalanced case is Total HP = (2 × 19,627) + 9487 = 48,741 At a cost of $2000 per installed HP, Total HP cost = $2000 × 48,741 = $97.48 million The hydraulically balanced case requires 10,140 (58,881 − 48,741) HP more and will cost approximately $20.28 ($117.76 − $97.48) million more. In addition to the extra HP cost, the hydraulically balanced case will require a pressure regulator that will waste energy and result in extra equipment cost. Therefore, the advantages of using identical components, by reducing spare parts and inventory in the hydraulically balanced case, must be weighed against the additional cost. It may not be worth spending the extra $20 million to obtain this benefit. The preferred solution in this case is for the Payson and Williams compressor stations to be identical (compression ratio = 1.54) and the Snowflake compressor station to be a smaller one (compression ratio = 1.24), requiring the lower compression ratio and horsepower, to provide the required 600 psig delivery pressure at Douglas.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 163 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

163

1296 psia

1080 psia

900 psia

Q

Flow = Q Compression ratio = 1.2 Compression ratio = 1.2 Overall compression ratio = 1.44 Figure 4.9

Compressors in series.

4.9 COMPRESSORS IN SERIES AND PARALLEL When compressors operate in series, each unit compresses the same amount of gas but at different compression ratios, such that the overall pressure increase of the gas is achieved in stages, as shown in Figure 4.9. It can be seen from Figure 4.9 that the first compressor compresses gas from a suction pressure of 900 psia to 1080 psia at a compression ratio of 1.2. The second compressor takes the same volume and compresses it from 1080 psia to a discharge pressure of 1080 × 1.2 = 1296 psia. Thus, the overall compression ratio of the two identical compressors in series is 1296/900 = 1.44. We have thus achieved the increase in pressure in two stages. At the end of each compression cycle, the gas temperature rises to some value calculated in accordance with Equation 4.19. Therefore, with multiple stages of compression, unless the gas is cooled between stages, the final gas temperature may be too high. High gas temperatures are not desirable, since the throughput capability of a gas pipeline decreases with gas flow temperature. Therefore, with compressors in series, the gas is cooled to the original suction temperature between each stage of compression, such that the final temperature at the end of all compressors in series is not exceedingly high. Suppose the calculated discharge temperature of a compressor is 232°F, starting at a 70°F suction temperature and with a compression ratio of 1.4. If two of these compressors were in series and there were no cooling between compressions, the final gas temperature would reach approximately (232 + 460)(232 + 460) = 903.5°R = 443.5°F 70 + 460 This is too high a temperature for pipeline transportation. On the other hand, if we cool the gas back to 70°F before compressing it through the second compressor, the final temperature of the gas coming out of the second compressor will be 232°F, approximately. We will discuss compressors in series in more detail in the subsequent section.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 164 Saturday, April 2, 2005 9:48 AM

164

GAS PIPELINE HYDRAULICS

1260 psia

300 MMSCFD

900 MMSCFD

300 MMSCFD

900 psia 900 MMSCFD

300 MMSCFD

Compression ratio = 1.4 Figure 4.10

Compressors in parallel.

Compressors are installed in parallel so that large volumes necessary can be provided by multiple compressors, each producing the same compression ratio. Three identical compressors with compression ratio of 1.4 can be used to provide a 900 MMSCFD gas flow from a suction pressure of 900 psia. In this example, each compressor will compress 300 MMSCFD from 900 psia to a discharge pressure of P2 = 900 × 1.4 = 1260 psia This is illustrated schematically in Figure 4.10. Unlike compressors in series, the discharge temperature of the gas coming out of the parallel bank of compressors will not be high, since the gas does not undergo multiple compression ratios. The gas temperature on the discharge side of each parallel compressor will be the same as that of a single compressor with the same compression ratio. Therefore, three parallel compressors, each compressing the same volume of gas at a compression ratio of 1.4, will have a final discharge temperature of 232°F, starting from a suction temperature of 70°F. Gas cooling is required at these temperatures in order to achieve efficient gas transportation and also operate at temperatures not exceeding the limits of the pipe coating material. Generally, pipe coating requires the gas temperature not to exceed 140 to 150°F. The compression ratio was defined earlier as the ratio of the discharge pressure to the suction pressure. The higher the compression ratio, the higher will be the gas discharge temperature, in accordance with Equation 4.19. Consider a suction temperature of 80°F and the suction and discharge pressures of 900 psia and 1400 psia, respectively. The compression ratio is 1400/900 = 1.56. Using Equation 4.19, the discharge temperature will be T2 1400 80 + 460 = 900

1.3−1 1.3

T2 = 598.36°R or 138.36°F If the compression ratio is increased to 2.0, the discharge temperature will become 173.67°F. It can be seen that the discharge temperature of the gas increases

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 165 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

165

considerably with the compression ratio. Since the throughput capacity of a gas pipeline decreases with gas temperature, we must find a way to reduce the high gas temperature resulting from gas compression. In previous chapters, we solved many problems with a constant gas inlet temperature of 80°F. In order to maintain throughput, cooling should be provided on the discharge of the compressor. We prefer centrifugal compressors used in gas pipeline applications to have a compression ratio of 1.5 to 2.0; there may be instances in which higher compression ratios are required due to lower gas receipt pressures and higher pipeline discharge pressures to enable a given volume of gas to be transported through a pipeline. Reciprocating compressors are designed to provide higher compression ratios. However, manufacturers limit maximum compression ratios to a range of 4 to 6. This is due to high forces that are exerted on the compressor components, which cause expensive material requirements as well as complicated safety needs. Suppose a compressor is required to provide gas at 1500 psia from gas that is received at 200 psia. This requires an overall compression ratio of 7.5. Since this is beyond the acceptable range of compression ratios, we will have to provide this compression in stages. If we provide the necessary pressure by using two compressors in series, each compressor will require to be at a compression ratio of 7.5 , or approximately 2.74. The first compressor raises the pressure from 200 psia to 200 × 2.74 = 548 psia. The second compressor will then boost the gas pressure from 548 psia to 548 × 2.74 = 1500 psia, approximately. In general, if n compressors are installed in series to achieve the required compression ratio r, we can state that each compressor will operate at a compression ratio of 1

r = (rt ) n

(4.25)

where r = compression ratio, dimensionless rt = overall compression ratio, dimensionless n = number of compressors in series It has been found that by providing the overall compression ratio by means of identical compressors in series, power requirements will be minimized. Thus, in the preceding example, we assumed that two identical compressors in series, each providing a compression ratio of 2.74 resulting in an overall compression ratio of 7.5, will be a better option than if we had a compressor with a compression ratio of 3.0 in series with another compressor with a compression ratio of 2.5. To illustrate this further, if an overall compression ratio of 20 were required and we were to use three compressors in series, the most economical option would be to use identical compres1 sors, each with a compression ratio of (20) 3 = 2.71. Example 12 A compressor station with multiple compressors in series is to provide a gas discharge pressure of 1500 psia. The gas inlet pressure and temperature are 100 psia and 80°F, respectively. How many compressors in series will be required if the discharge temperature is limited to 250°F? The ratio of specific heats g = 1.4.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 166 Saturday, April 2, 2005 9:48 AM

166

GAS PIPELINE HYDRAULICS

Solution The overall compression ratio is r=

1500 = 15.0 100

Since this is more than the maximum recommended compression ratio of 4 to 6, we need two or more compressors in series. Initially, consider two compressors in series. The compression ratio for each compressor is 1

r = (15) 2 = 3.873 This is acceptable, but the discharge temperature needs to be checked. From Equation 4.19, the discharge temperature for the first compressor is 0.4

T2 = (80 + 460)(3.8) 1.4 = 790.76°R or 330.76°F. This temperature is higher than the 250°F allowable. Therefore, we will need to consider three stages of compression. Using three compressors in series, the compression ratio is 1

r = (15) 3 = 2.466 Therefore, the discharge temperature for each compressor is 0.4

T2 = (80 + 460)(2.466) 1.4 = 698.86°R or 239°F. Since this is less than 250°F allowed, the three compressors in series are the choice. However, the gas must be cooled to the initial inlet temperature of 80°F between each compressor to limit discharge temperatures to 239°F.

4.10 TYPES OF COMPRESSORS—CENTRIFUGAL AND POSITIVE DISPLACEMENT Compressors used in natural gas transportation systems are either positive displacement (PD) type or centrifugal (CF) type. Positive displacement compressors generate the pressure required by trapping a certain volume of gas within the compressor and increasing the pressure by reduction of volume. The high-pressure gas is then released through the discharge valve into the pipeline. Piston-operated reciprocating compressors fall within the category of positive displacement compressors. These compressors

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 167 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

167

have a fixed volume and are able to produce high compression ratios. Centrifugal compressors, on the other hand, develop the pressure required by the centrifugal force due to rotation of the compressor wheel that translates the kinetic energy into pressure energy of the gas. Centrifugal compressors are more commonly used in gas transmission systems due to their flexibility. Centrifugal compressors have lower installed cost and lower maintenance expenses. They can handle larger volumes within a small area compared to positive displacement compressors. They also operate at high speeds and are of balanced construction. However, centrifugal compressors have less efficiency than positive displacement compressors. Positive displacement compressors have flexibility in pressure range, have higher efficiency, and can deliver compressed gas at a wide range of pressures. They are also not very sensitive to the composition of the gas. Positive displacement compressors have pressure ranges up to 30,000 psi and range from very low HP to more than 20,000 HP per unit. Positive displacement compressors can be single stage or multistage, depending upon the compression ratio required. The compression ratio per stage for positive displacement compressors is limited to 4.0, because higher ratios cause higher discharge pressures, which affect the valve life of positive displacement compressors. Heat exchangers are used between stages of compression so that the compressed heated gas is cooled to the original suction temperature before being compressed in the next stage. The HP required in a positive displacement compressor is usually estimated from charts provided by the compressor manufacturer. The following equation can be used for large slow-speed compressors with compression ratios greater than 2.5 and for gas specific gravity of 0.65. BHP = 22rNQF where BHP r N Q F

= = = = = = = =

(4.26)

brake horsepower compression ratio per stage number of stages gas flow rate, MMSCFD at suction temperature and 14.4 psia factor that depends on the number of compression stages 1.0 for single-stage compression 1.08 for two-stage compression 1.10 for three-stage compression

In Equation 4.26, the constant 22 is changed to 20 when gas gravity is between 0.8 and 1.0. Also, for compression ratios between 1.5 and 2.0, the constant 22 is replaced with a number between 16 and 18. Example 13 Calculate the BHP required to compress 5 MMSCFD gas at 14.4 psia and 70°F, with an overall compression ratio of 7, considering two-stage compression.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 168 Saturday, April 2, 2005 9:48 AM

168

GAS PIPELINE HYDRAULICS

Solution Considering two identical stages, the compression ratio per stage = 7.0 = 2.65. Using Equation 4.26, we get BHP = 22 × 2.65 × 2 × 5 × 1.08 = 629.64 Centrifugal compressors can be a single-wheel or single-stage compressor or multiwheel or multistage compressor. Single-stage centrifugal compressors have a volume range of 100 to 150,000 ft3/min at actual conditions (ACFM). Multistage centrifugal compressors handle a volume range of 500 to 200,000 ACFM. The operational speeds of centrifugal compressors range from 3000 to 20,000 r/min. The upper limit of speed will be limited by the wheel tip speed and stresses induced in the impeller. Advances in technology have produced compressor wheels operating at speeds in excess of 30,000 r/min. Centrifugal compressors are driven by electric motors, steam turbines, or gas turbines. Sometimes speed increasers are used to increase the speeds necessary to generate the pressure.

4.11 COMPRESSOR PERFORMANCE CURVES

e lin

e

The performance curve of a centrifugal compressor that can be driven at varying speeds typically shows a graphic plot of the inlet flow rate in actual cubic feet per minute (ACFM) against the head or pressure generated at various percentages of the design speed. Figure 4.11 shows a typical centrifugal compressor performance curve or performance map.

Head, ft-lb/lb

surg

100% speed 90% 80% 70% 60%

Flow rate, ACFM Figure 4.11

Typical centrifugal compressor performance curve.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 169 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

169

The limiting curve on the left-hand side is known as the surge line, and the corresponding curve on the right side is known as the stone wall limit. Generally, the performance of a centrifugal compressor follows the “affinity laws.” According to the affinity laws, as the rotational speed of the centrifugal compressor is changed, the inlet flow and head vary as the speed and the square of the speed, respectively, as indicated in the following equations. For compressor speed change, Q2 N 2 = Q1 N1 H2 N 2 = H1 N1

(4.27) 2

(4.28)

where Q1, Q2 = initial and final flow rates H1, H2 = initial and final heads N1, N2 = initial and final compressor speeds In addition, the horsepower for compression varies as the cube of the speed change as follows: HP2 N 2 = HP1 N1

3

(4.29)

An example problem of using the affinity laws to predict the performance of a centrifugal compressor is illustrated next. Example 14 The compressor head and volume flow rate for a centrifugal compressor at 18,000 rpm are as follows: Flow Rate, Q

Head, H

ACFM 360 450 500 600 700 730

ft-lb/lb 10,800 10,200 9700 8200 5700 4900

Using the affinity laws, determine the performance of this compressor at a speed of 20,000 rpm.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 170 Saturday, April 2, 2005 9:48 AM

170

GAS PIPELINE HYDRAULICS

Solution The ratio of speed is 20000 = 1.11 18000 The multiplier for the flow rate is 1.11 and the multiplier for the head is (1.11)2 or 1.232. Using the affinity laws, the performance of the centrifugal compressor at 20,000 rpm is as follows: Flow Rate, Q

Head, H

ACFM 399.6 499.5 555.0 666.0 777.0 810.0

ft-lb/lb 13,306 12,566 11,950 10,102 7,022 6,037

Next, we will explore how the head developed by a centrifugal compressor is calculated from the suction and discharge pressures, the compressibility factor, and the polytropic or adiabatic exponent. The calculation for the actual or inlet flow rate (ACFM) from the standard flow rate will also be illustrated. Finally, knowing the maximum head that can be generated per stage, the number of stages needed will be calculated. Suppose a centrifugal compressor is used to raise the gas pressure from 800 psia to 1440 psia starting at a suction temperature of 70°F and gas flow rate of 80 MMSCFD. The average compressibility factor from the suction to the discharge side is 0.95. The compressibility factor at the inlet is assumed to be 1.0, and the polytropic exponent is 1.3. Gas gravity is 0.6. The head generated by the compressor is calculated as 0.3 1.3 1440 1.3 53.28 H= × 0.95 × (70 + 460) 1 − = 28,146 ft-lb/lb 0.6 0.3 800

The actual flow rate at inlet conditions is calculated using the gas law as Qact =

80 × 14.7 × 1.0 70 + 460 10 6 = 1040.5 ft3/min (ACFM) × × 800 60 + 460 24 × 60

If this particular compressor, according to vendor data, can produce a maximum head per stage of 10,000 ft-lb/lb, the number of stages required to produce the required head is n=

Copyright 2005 by Taylor & Francis Group, LLC

28146 = 3 , approximately. 10000

2785_C004.fm Page 171 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

171

Next, suppose that this compressor has a maximum design speed of 16,000 rpm. The actual operating speed necessary for the three-stage compressor is, according to the affinity laws, 28,146 = 15, 498 3 × 10, 000

N act = 16, 000

Therefore, in order to generate 28,146 ft-lb/lb of head at a gas flow rate of 1040.5 ACFM, this three-stage compressor must run at a speed of 15,498 rpm.

4.12 COMPRESSOR STATION PIPING LOSSES As the gas enters the suction side of the compressor, it flows through a complex piping system within the compressor station. Similarly, the compressed gas leaving the compressor traverses the compressor station discharge piping system that consists of valves and fittings before entering the main pipeline on its way to the next compressor station or delivery terminus. This is illustrated in Figure 4.12. It can be seen from Figure 4.12 that at the compressor station boundary A on the suction side, the gas pressure is P1. This pressure drops to a value Ps at the compressor suction, as the gas flows through the suction piping from A to B. This suction piping, consisting of valves, fittings, filters, and meters, causes a pressure drop of ∆Ps to occur. Therefore, the actual suction pressure at the compressor is Ps = P1 – ∆Ps

(4.30)

where Ps = compressor suction pressure, psia P1 = compressor station suction pressure, psia ∆Ps = pressure loss in compressor station suction piping, psi At the compressor, the gas pressure is raised from Ps to Pd through a compression ratio r as follows: r=

P1

Pd Ps

(4.31)

A

B

Ps

Pd Flow

Figure 4.12

Compressor station suction and discharge piping.

Copyright 2005 by Taylor & Francis Group, LLC

P2

2785_C004.fm Page 172 Saturday, April 2, 2005 9:48 AM

172

GAS PIPELINE HYDRAULICS

where r = compression ratio, dimensionless Pd = compressor discharge pressure, psia The compressed gas then flows through the station discharge piping and loses pressure until it reaches the station discharge valve at the boundary D of the compressor station. If the station discharge pressure is P2, we can write P2 = Pd – ∆Pd

(4.32)

where P2 = compressor station discharge pressure, psia ∆Pd = pressure loss in compressor station discharge piping, psi Generally, the values of ∆Ps and ∆Pd range from 5 to 15 psi. Example 15 A compressor station on a gas transmission pipeline has the following pressures at the station boundaries. The station suction pressure = 850 psia, and the station discharge pressure = 1430 psia. The pressure losses in the suction piping and discharge piping are 5 psi and 10 psi, respectively. Calculate the compression ratio of this compressor station. Solution From Equation 4.30, the compressor suction pressure is Ps = 850 – 5 = 845 psia Similarly, the compressor discharge pressure is Pd = 1430 + 10 = 1440 psia Therefore, the compression ratio is

r=

1440 = 1.70 845

4.13 COMPRESSOR STATION SCHEMATIC A typical compressor station schematic showing the arrangement of the valves, piping, and the compressor itself is shown in Figure 4.13.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 173 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

173

To next compressor station

From previous compressor station A

B MOV

MOV

Pd

Ps

Compressors Figure 4.13

Compressor station schematic.

4.14 SUMMARY We discussed compressing a gas to generate the pressure needed to transport the gas from one point to another along a pipeline. An important parameter known as the compression ratio determines the horsepower required to compress a certain volume of gas and also influences the discharge temperature of the gas exiting the compressor. In a long-distance gas transmission pipeline, the method of locating intermediate compressor stations and minimizing energy lost was discussed. Hydraulically balanced and optimized compressor station locations were also discussed. Calculation of isothermal, adiabatic, and polytropic compression processes was explained and illustrated with sample problems. The HP required for a given compression ratio and calculation of the gas discharge temperature were explained. The different types of compressors, such as positive displacement and centrifugal, were explained, along with their advantages and disadvantages. The need for configuring compressors in series and parallel was explored. The centrifugal compressor performance curve was discussed, and the effect of rotational speed on the flow rate and head using the affinity laws was illustrated with examples. Finally, the impact of the compressor station yard piping pressure drops and how they affect the compression ratio and horsepower were discussed.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 174 Saturday, April 2, 2005 9:48 AM

174

GAS PIPELINE HYDRAULICS

PROBLEMS 1. A natural gas pipeline 120 mi long from Dover to Leeds is constructed of NPS 14 and .250 in. wall thickness pipe, with an MOP of 1400 psig. The gas specific gravity and viscosity are 0.6 and 8 × 10−6 lb/ft-s, respectively. The pipe roughness can be assumed to be 600 µin., and the base pressure and base temperature are 14.7 psia and 60°F, respectively. The gas flow rate is 120 MMSCFD at 70°F, and the delivery pressure required at Leeds is 700 psig. Determine the number and locations of compressor stations required, neglecting elevation difference along the pipeline. Assume Z = 0.90. 2. Calculate the compressor horsepower required for an adiabatic compression of 80 MMSCFD gas with inlet temperature of 70°F and 800 psia pressure. The discharge pressure is 1400 psia. Assume the compressibility factors at suction and discharge conditions to be Z1 = 0.95 and Z2 = 0.88, respectively, and the adiabatic exponent g = 1.3, with the adiabatic efficiency ha = 0.82. If the mechanical efficiency of the compressor driver is 0.94, what BHP is required? Also, calculate the outlet temperature of the gas. 3. Natural gas at 4 Mm3/day and 24°C is compressed isentropically (g = 1.3) from a suction pressure of 6.2 MPa to a discharge pressure of 9.4 MPa in a centrifugal compressor with an isentropic efficiency of 0.82. Calculate the compressor power required, assuming the compressibility factors at suction and discharge conditions to be Z1 = 0.96 and Z2 = 0.87, respectively. If the mechanical efficiency of the compressor driver is 0.94, what is the driver power required? Also, calculate the outlet temperature of the gas. 4. Determine the horsepower required to compress natural gas in a pipeline at a flow rate of 350 MMSCFD and at a compression ratio of 1.6, discharging at 1400 psig pressure. The suction temperature is 80°F. The base temperature and base pressure are 60°F and 14.7 psia, respectively. The gas specific gravity is 0.65, and the compression efficiency is 0.85. What is the discharge temperature of the gas, assuming a polytropic compression exponent of 1.39? The compressibility factor Z = 1.0 at suction conditions and Z = 0.86 at discharge conditions. 5. Determine the horsepower required to compress natural gas in a pipeline at a flow rate of 500 MMSCFD and at a compression ratio of 1.4, discharging at 1200 psia pressure. The suction temperature is 70°F. The base temperature and base pressure are 60°F and 14.7 psia, respectively. The gas specific gravity is 0.6, and assume a compression efficiency of 0.9. What is the discharge temperature of the gas, assuming the polytropic compression coefficient of 1.38? Z = 1.0 at suction conditions and Z = 0.86 at discharge conditions. 6. A gas transmission pipeline is 220 mi long, NPS 24, 0.500 in. wall thickness, and runs from Taylor to Jenks. There is an origin compressor station at Taylor and two intermediate compressor stations at Trent (milepost 70) and Beaver (milepost 130). There are no intermediate flow deliveries or injections, and the inlet flow rate of 500 MMSCFD at Taylor equals the delivery flow rate at Jenks. The delivery pressure required at Jenks is 700 psig, and the MOP of the pipeline is 1440 psig throughout. Neglect the effects of elevation, and assume a constant gas flow temperature of 70°F and constant values of transmission factor F = 20 and

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 175 Saturday, April 2, 2005 9:48 AM

COMPRESSOR STATIONS

175

compressibility factor Z = 0.87 throughout the pipeline. The gas gravity = 0.6, base pressure = 14.7 psia, and base temperature = 60°F. Use a polytropic compression coefficient of 1.4 and a compression efficiency of 0.85. Determine the best locations for the intermediate compressor stations at Trent and Beaver. If the flow rate drops to 350 MMSCFD, will a single intermediate compressor station be sufficient at the reduced flow rate? 7. A compressor station with multiple compressors in series is to provide a gas discharge pressure of 1400 psia. The gas inlet pressure and temperature are 200 psia and 70°F, respectively. How many compressors in series will be required if the discharge temperature is limited to 200°F? The ratio of specific heats g = 1.3. 8. Calculate the BHP required to compress 8 MMSCFD gas at 14.4 psia and 80°F, with an overall compression ratio of 8, considering two-stage compression. 9. The compressor head and volume flow rates for a centrifugal compressor at 15,000 rpm are as follows: Q–ACFM H–ft-lb/lb

720 5400

900 5100

1000 4900

1200 4100

1400 2800

Using the affinity laws, determine the performance of this compressor at a speed of 12,000 rpm. 10. A compressor station on a gas transmission pipeline has the following pressures at the station boundaries: station suction pressure = 840 psig and station discharge pressure = 1410 psig. The pressure losses in the suction piping and discharge piping are 6 psi and 12 psi, respectively. Calculate the compression ratio of this compressor station.

REFERENCES 1. Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003. 2. Westaway, C.R. and Loomis, A.W., Cameron Hydraulic Data, 16th ed., Ingersoll-Rand, Montvale, NJ, 1981. 3. Flow of Fluids through Valves, Fittings and Pipe, Crane Company, New York, 1976. 4. McCain, W.D. Jr., The Properties of Petroleum Fluids, Petroleum Publishing Company, Tulsa, OK, 1973. 5. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 6. Engineering Data Book, 10th ed., Gas Processors Suppliers Association, Tulsa, OK, 1994.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C004.fm Page 176 Saturday, April 2, 2005 9:48 AM

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 177 Friday, April 1, 2005 3:30 PM

CHAPTER

5

Pipe Loops versus Compression

In this chapter we will explore the need for installing pipe loops in order to increase the throughput in a gas pipeline. Looping will be compared to another means of increasing pipeline capacity, such as installing compressor stations. The advantages and disadvantages of looping pipes vs. adding compressor stations will be discussed.

5.1 PURPOSE OF A PIPE LOOP The purpose of a pipe loop that is installed in a segment of a pipeline is to essentially reduce the amount of pressure drop in that section of pipe. By doing so, the overall pressure drop in the pipeline will be reduced. This, in turn, will result in an increased pipeline flow rate at the same inlet pressure. Alternatively, if the flow rate is kept constant, reduction in total pressure required will cause a reduction in pumping horsepower. This is illustrated in Figure 5.1. The pipe loop can be constructed of the same-diameter pipe as the main pipeline, or in some cases it can be of a different size. As we have seen in the analysis of parallel pipes in Chapter 3, the same diameter of pipe loop will result in equal volumes of gas flow in the main pipe as well as the loop. Thus, an NPS 20 pipe looped with an identical NPS 20 pipe segment will reduce the flow to one-half its original value in each pipe. If the loop is larger or smaller in diameter compared to the main pipeline, the volume distribution will not be equal. An NPS 20 pipe looped with an NPS 16 pipe will result in approximately 64% of the flow rate going through the larger-diameter pipe and 36% through the smaller-diameter pipe.

177

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 178 Friday, April 1, 2005 3:30 PM

178

GAS PIPELINE HYDRAULICS

without loop

Pressure

with loop

loop C

A

Figure 5.1

B

Effect of pipe loop.

5.2 PURPOSE OF COMPRESSION We have seen in Chapter 3 and Chapter 4 that installing intermediate compressor stations along a pipeline will increase the flow rate and also reduce the operating pressure in a long gas transmission pipeline. The installation of the intermediate compressor station will result in additional operational and maintenance issues in comparison with pipe loops. Sometimes, additional compression is installed to increase flow rate in preference to looping the pipeline, since looping will involve additional permitting and right-of-way issues and could cost considerably more than adding the new compressor station. Installation of an intermediate compressor to increase flow rate is illustrated in Figure 5.2.

1440 psig

without second compressor station

with

seco

nd co

mpre

ssor

statio

n

800 psig Flow Q = 100 MMSCFD

Dover

NPS 16 pipeline

Kent Distance

Figure 5.2

Adding a compressor station.

Copyright 2005 by Taylor & Francis Group, LLC

Leeds

2785_C005.fm Page 179 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

179

5.3 INCREASING PIPELINE CAPACITY Consider an existing pipeline that is currently limited by the operating pressure that is close to the MAOP of the pipeline. Suppose the capacity of an NPS 16 pipeline is 100 MMSCFD and the discharge pressure at the originating compressor station is 1440 psig, as shown in Figure 5.2. It can be seen that, at the given flow rate and discharge pressure, the delivery pressure is 800 psig. If the pipeline flow rate is increased to 120 MMSCFD without changing the originating pressure of 1440 psig, the increased flow will cause greater pressure drop and, hence, the delivery pressure at the pipeline terminus will drop to some value such as 600 psig. The reduced delivery pressure may or may not be acceptable to the customer receiving the gas. However, we cannot increase the discharge pressure at the beginning of the pipeline to compensate for the drop in delivery pressure because the pressure is already at the MAOP level. How can we increase the flow rate and still provide the same delivery pressure as before? By installing an intermediate compressor station as shown in Figure 5.2, we can pump the increased volume approximately halfway and then boost the pressure at the new compressor station to the same MAOP level for ultimate delivery to the pipeline terminus at 800 psig, as before. This is illustrated in Figure 5.2. Thus, we have been able to achieve the increased pipeline capacity of 120 MMSCFD by installing an additional compressor station at approximately the halfway point along the pipeline. Suppose we want to increase the flow rate further without changing the discharge pressure or the delivery pressure. It is clear that we could install additional intermediate compressor stations as needed to achieve the increased throughput, while maintaining the same delivery pressure. This is illustrated in Figure 5.3, where two additional compressor stations have been installed to increase the pipeline throughput while maintaining the desired delivery pressure at the pipeline terminus.

1440 psig

800 psig

Q Dover

Figure 5.3

NPS 16 pipeline

Multiple compressor stations.

Copyright 2005 by Taylor & Francis Group, LLC

Kent Distance

Leeds

2785_C005.fm Page 180 Friday, April 1, 2005 3:30 PM

180

GAS PIPELINE HYDRAULICS

Pha

se 1

Ph

as

Pressure

e

2

Q

=

MOP = 1440 psig Q=

238

.41

28

ini

MM

SCF

D

tia

8.

41

lQ

M

=

18

800 psig

8.4

M

1M

SC

MS

FD

CF

D

Q 800 psig Windsor m.p. 0

Avon m.p. 37.55

Hart m.p. 57.33

Cardiff m.p. 100

NPS 16 pipeline 100 mi long Figure 5.4

Windsor to Cardiff pipeline.

However, there is a limit to the number of compressor stations that can be installed in a given pipeline system, since the HP required continues to increase with flow rate and, hence, the capital cost and operating costs increase as well. At some point, the cost increases at a very high rate compared to the increase in flow rate. Each pipe size has a particular volume that can be economically transported based upon cost. An additional factor that must be taken into consideration as the flow rate is increased is the resulting higher velocity. As indicated in Chapter 2, the gas velocity must be well below the erosional velocity for the pipe. Example 1 A natural gas pipeline is 100 mi long and is constructed of NPS 16 and 0.250 in. wall thickness and runs from Windsor to Cardiff, as shown in Figure 5.4. 1. Neglecting elevation effects, calculate the maximum throughput capability of this pipeline, based upon an MAOP of 1440 psig and a delivery pressure of 800 psig at Cardiff. The suction pressure at Windsor is 800 psig. 2. Determine the requirement for two expansion scenarios. The phase 1 expansion will increase pipeline throughput by 50 MMSCFD and phase 2 will increase throughput by another 50 MMSCFD. In each case, calculate the number of compressor stations and HP required. The gas flow velocities must be checked to ensure that they are within erosional limits. 3. Also estimate the approximate cost for each of these cases, using an overall installed cost of $2000 per HP for the compressor stations. 4. Compare these expansion cases using pipe loop instead of compression. Thus, for phase 1, instead of building intermediate compressor stations, calculate the amount

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 181 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

181

of pipe loop needed to reduce the pressure drop at the higher flow rate. Similarly, for the phase 2 flow rate, calculate the looping necessary to maintain pressures without adding compressor stations. Estimate the cost of the expansion scenarios using pipe loops instead of compressor stations, based upon an overall installed cost of $500,000 per mile of loop. Assume a transmission factor of 20, gas flow temperature of 80°F, and compressibility factor of 0.85 throughout. Additional data are as follows: gas gravity = 0.6, ratio of specific heats = 1.4, base temperature = 60°F, and base pressure = 14.7 psia. The compressor isentropic efficiency = 0.8, and the mechanical efficiency of the compressor driver is 0.95. Solution 1. First, determine the initial capacity, considering one compressor station at Windsor providing the pressure of 1440 psig needed for delivery pressure of 800 psig at Cardiff. Using General Flow Equation 2.4, we calculate the initial capacity, Q, of the pipeline as follows: 520 1454.72 − 814.72 Q = 38.77 × 20 14.7 0.6 × 540 × 100 × 0.85

0.5

(15.5)2.5 = 188,410,280 SCFD

The HP required is calculated from Equation 4.15: 0.40 1 + 0.85 1 1454.7 1.40 1.40 − 1 = 6357 HP = 0.0857 × 188.41 (540) 0.40 2 0.8 814.7

BHP required =

6357 = 7064 0.95

Checking gas velocities using Equation 2.26, the gas velocity at Windsor is 188.41 × 10 6 14.7 540 u1 = 0.002122 520 1454.7 = 17.46 ft/s 15.52 The velocity at Cardiff is 188.41 × 10 6 14.7 540 u1 = 0.002122 520 814.7 = 31.18 ft/s 15.52 The erosion velocity from Equation 2.31 is

umax = 100

Copyright 2005 by Taylor & Francis Group, LLC

0.85 × 10.73 × 540 = 58.94 ft/s 29 × 0.6 × 814.7

2785_C005.fm Page 182 Friday, April 1, 2005 3:30 PM

182

GAS PIPELINE HYDRAULICS

2. Next, we will calculate the compressor station requirement for the phase 1 flow rate of Q = 188.41 + 50 = 238.41 MMSCFD Assume that an additional compressor station is needed for this flow rate. This will be located at Avon at a distance of L miles from Cardiff, such that a discharge pressure of 1440 psig at Avon will produce a delivery pressure of 800 psig at Cardiff. We will calculate the value of the pipe length, L , using General Flow Equation 2.4 as follows: 520 1454.72 − 814.72 238.41 × 10 6 = 38.77 × 20 14.7 0.6 × 540 × L × 0.85

0.5

(15.5)2.5

Solving for L, we get

L = 62.45 mi Next, calculate the suction pressure at Avon using 1440 psig at Windsor and considering a pipe length of 37.55 (100 − 62.45) mi between Windsor and Avon. Using General Flow Equation 2.4, we get 520 1454.72 − P22 238.41 × 10 = 38.77 × 20 . 14 7 0 . 6 × 540 × 37 . 55 × 0 . 85

0.5

6

(15.5)2.5

Solving for the suction pressure at Avon, we get P2 = 1114.85 psia = 1100.15 psig Therefore, the compression ratio at Avon is r=

1454.7 = 1.30 1114.85

This is a satisfactory compression ratio for a centrifugal compressor. The HP required at Windsor and Avon for phase 1 will be calculated using Equation 4.15. For Windsor, assuming the compressibility factor at suction is 1.0, 0.40 1 + 0.85 1 1454.7 1.40 1.40 − 1 = 8044 HP = 0.0857 × 238.41 (540) 0.40 2 0.8 814.7

Therefore, the BHP required at Windsor for phase 1 =

8044 = 8468. 0.95

Similarly, the HP required at Avon is 0.40 1.40 1 + 0.85 1 1.40 − 1 = 3476 ( . ) ( ) HP = 0.0857 × 238.41 540 1 30 2 0.8 0.40

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 183 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

Therefore, the BHP required at Avon =

183

3476 = 3659. 0.95

The total compressor HP required at both compressor stations for the phase 1 flow rate of 238 MMSCFD is 8468 + 3659 = 12,127 HP Therefore, the incremental HP for phase 1 is ∆HP = 12,127 – 7064 = 5063 HP 3. This represents the additional compressor HP required for phase 1 for the extra 50 MMSCFD flow rate. The cost of this incremental HP, based on $2000 per installed HP, is ∆Cost = 5063 × 2000 = $10.13 million Next, check the gas velocity at the increased flow rate in phase 1 from Equation 2.26. The velocity at Cardiff is 238.41 × 10 6 14.7 540 u1 = 0.002122 520 814.7 = 39.45 ft/s 15.52 This velocity is acceptable, since it is less than the erosion velocity of 58.94 ft/s calculated earlier. The velocity at Windsor at the higher pressure of 1440 psig will be lower and, hence, less than the erosion velocity. Next, consider the phase 2 flow rate of Q = 238.41 + 50 = 288.41 MMSCFD Since phase 2 occurs after phase 1, where the Avon compressor station is already built, we might have to install one compressor station between Windsor and Avon and another between Avon and Cardiff. If we consider this phase independent of phase 1, we could probably install two additional compressor stations between Windsor and Cardiff to handle the phase 2 flow of 288.41 MMSCFD. For now, we will consider a compressor station at Jenks between Windsor and Avon and another one at Hart located between Avon and Cardiff. We will calculate the distance of L miles from Hart to Cardiff, such that a discharge pressure of 1440 psig at Hart will produce a delivery pressure of 800 psig at Cardiff. The value of L is calculated, using General Flow Equation 2.4, as we did before for locating the Avon compressor station, as follows: 520 1454.72 − 814.72 288.41 × 10 6 = 38.77 × 20 14.7 0.6 × 540 × L × 0.85

0.5

(15.5)2.5

Solving for L, we get L = 42.67 mi Therefore, the Hart compressor station will be located at a distance of 19.78 (62.45 − 42.67) mi from Avon. The suction pressure at Hart is calculated next.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 184 Friday, April 1, 2005 3:30 PM

184

GAS PIPELINE HYDRAULICS

Using General Flow Equation 2.4, we get 520 1454.72 − P22 288.41 × 10 6 = 38.77 × 20 14.7 0.6 × 540 × 19.78 × 0.85

0.5

(15.5)2.5

Solving for the suction pressure at Hart, we get P2 = 1201.24 psia = 1186.54 psig Therefore, the compression ratio at Hart is r=

1454.7 = 1.21 1201.24

This is a satisfactory compression ratio. Before determining the location of the Jenks compressor station between Windsor and Avon, calculate the suction pressure at Avon, assuming Jenks doesn’t exist and that the Windsor compressor station pumps directly into Avon, as in phase 1. The suction pressure at Avon, considering 1440 psig at Windsor, is calculated using General Flow Equation 2.4: 520 1454.72 − P2 2 288.41 × 10 6 = 38.77 × 20 14.7 0.6 × 540 × 37.55 × 0.85

0.5

(15.5)2.5

Solving for the suction pressure at Avon, we get P2 = 915.54 psia = 900.84 psig Therefore, the compression ratio at Avon is r=

1454.7 = 1.59 915.54

This is a satisfactory compression ratio. Therefore, for phase 2, we will need only two compressor stations besides Windsor, Avon at milepost 37.55 and Hart at milepost 57.33. Next, calculate the total HP required at Windsor, Avon, and Hart at phase 2 flow rates. The HP required at Windsor is, using Equation 4.15, 0.40 1 + 0.85 1 1454.7 1.40 1.40 − HP = 0.0857 × 288.41 540 1 ( ) 2 0.8 814.7 = 9731 0.40

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 185 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

185

Therefore, the BHP required at Windsor for phase 2 =

9731 = 10,243. 0.95

Similarly, the HP required at Avon is 0.40 1 + 0.85 1 1.40 HP = 0.0857 × 288.41 (1.59) 1.40 − 1 = 7652 (540) 2 0.8 0.40

Therefore, the BHP required at Avon =

7652 = 8055. 0.95

The HP required at Hart is 0.40 1 + 0.85 1 1.40 HP = 0.0857 × 288.41 (540) (1.21) 1.40 − 1 = 3023 0.40 2 0.8

Therefore, the BHP required at Hart =

3023 = 3182. 0.95

The total compressor HP required at all three compressor stations for phase 2 is 10,243 + 8055 + 3182 = 21,480 HP The incremental HP for phase 2 compared to phase 1 is ∆HP = 21,480 – 12,127 = 9353 HP This represents the additional compression HP required for phase 2 compared to phase 1, for the additional 50 MMSCFD flow rate. The cost of this incremental HP, based on $2000 per installed HP, is ∆Cost = 9353 × 2000 = $18.71 million Next, check the velocity at increased flow rates in phase 2 from Equation 2.26. The velocity at Cardiff is 288.41 × 10 6 14.7 540 u1 = 0.002122 520 814.7 = 47.72 ft/s 15.52 This velocity is acceptable, since it is less than the erosion velocity. The velocity at higher pressures will be well within the limits. 4. In the preceding analysis, we accomplished the increase in flow rates for phase 1 and phase 2 by adding intermediate compressor stations. The capital cost for phase 1 expansion was $10.13 million and for the phase 2 expansion was an additional $18.71 million. Next, we will explore the two expansions by installing pipe loops without additional intermediate compressor stations. For phase 1, assume that L miles of the pipe near Cardiff will be looped. The reason we picked this section is because in Chapter 3, we found that looping close to the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 186 Friday, April 1, 2005 3:30 PM

186

GAS PIPELINE HYDRAULICS

downstream end is more beneficial than looping near the upstream end, as long as the flowing temperature was constant. Following the methodology of Chapter 3, we will determine the equivalent diameter of the pipe loop as follows: Assuming the loop to be of the same diameter as the main piping and L1 = L2, using Equation 3.18, we get Const1 = 1.0 Therefore, the equivalent diameter, using Equation 3.17, is 1/ 5

1 + 1 2 De = D1 1

= 1.32 D1 = 1.32 × 15.5 = 20.46 in.

Considering L miles of pipe of inside diameter 20.46 in., calculate the upstream pressure at the beginning of the loop as shown in Figure 5.5. The downstream pressure at Cardiff is 800 psig, and the upstream pressure P is unknown. Using General Flow Equation 2.4, 520 P 2 − 814.72 238.41 × 10 6 = 38.77 × 20 14.7 0.6 × 540 × L × 0.85

0.5

(20.46)2.5

(5.1)

There are two unknowns, P and L, in Equation 5.1. We need another equation to solve for both variables. For this, the pipe segment from Windsor to the start of the loop will be examined. Considering 1440 psig at Windsor, calculate the downstream pressure P at the beginning of the loop for a pipe length of (100 – L). Using General Flow Equation 2.4, 520 1454.72 − P 2 238.41 × 10 6 = 38.77 × 20 14.7 0.6 × 540 × (100 − L ) × 0.85

0.5

(15.5)2.5

(5.2)

MOP = 1440 psig

Phas

P = 962 psig

= 23

8 MM

Pressure

e 1Q

SCF

D 800 psig

Loop - 50.03 mi

238 MMSCFD 800 psig Windsor m.p. 0

NPS 16 pipeline 100 mi long Figure 5.5

Pipe loop for phase 1.

Copyright 2005 by Taylor & Francis Group, LLC

Cardiff m.p. 100

2785_C005.fm Page 187 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

187

Eliminating P from Equation 5.1 and Equation 5.2, we solve for L as follows: L = 50.03 mi Substituting this value of L in Equation 5.1 and solving for P, P = 976.76 psia = 962.06 psig Therefore, for phase 1, without an intermediate compressor station, flow increase can be achieved by looping 50.03 mi of pipe upstream of Cardiff. The installed cost of this pipe loop is 50.03 × $500,000 = $25.02 million In addition to this cost of pipe loop, we must also include the increased horsepower requirement at Windsor for the phase 1 flow rate. Since the discharge pressure at Windsor is still 1440 psig as before, the HP is the same as that calculated earlier. The incremental HP is (8468 – 7064) = 1404 HP. At $2000 per installed HP, the extra cost for incremental HP is 2000 × 1404 = $2.81 million Thus, for phase 1, the cost of looping pipe upstream of Cardiff and increased HP cost at Windsor compressor station is $25.02 + $2.81 = $27.83 million This compares with $10.13 million calculated earlier for phase 1 using a compressor station at Avon. Even though at first sight the looping appears to be a more expensive option, we must also consider the increased operating cost when adding a compressor station. The annual operating cost for the compressor station can be estimated considering the fuel consumption, operating and maintenance costs, and other costs. In Chapter 10, we will discuss more details of capital cost, operating cost, and cost of service. For now, we will only look at capital costs. For phase 2, at a flow rate of 288.41 MMSCFD, similarly calculate the amount of pipe loop needed, without adding any intermediate compression. The length of loop L required is calculated as follows: Using General Flow Equation 2.4, 520 P 2 − 814.72 288.41 × 10 = 38.77 × 20 . . 14 7 0 6 × 540 × L × 0 . 85 6

0.5

(20.46)2.5

(5.3)

Considering 1440 psig at Windsor, calculate the downstream pressure P at the beginning of the loop for a pipe length of (100 – L). Using General Flow Equation 2.4, 520 1454.72 − P 2 288.41 × 10 6 = 38.77 × 20 14.7 0.6 × 540 × (100 − L ) × 0.85

0.5

(15.5)2.5

(5.4)

Eliminating P from Equation 5.3 and Equation 5.4, we solve for the loop length L as L = 76.26 mi

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 188 Friday, April 1, 2005 3:30 PM

188

GAS PIPELINE HYDRAULICS

MOP = 1440 psig

Pressure

P = 1130 psig Pha

se 2

Q=

288

MM

SCF

D

800 psig

Loop - 76.26 mi

288 MMSCFD 800 psig Windsor m.p. 0

Cardiff m.p. 100 NPS 16 pipeline 100 mi long

Figure 5.6

Pipe loop for phase 2.

This is shown in Figure 5.6. Substituting this value of L in Equation 5.4 and solving for P, P = 1144.54 psia = 1129.84 psig Therefore, the installed cost of this pipe loop is 76.26 × $500,000 = $38.13 million In addition to the cost of pipe loop, we must include the increased horsepower requirement at Windsor for phase 2. Since the discharge pressure at Windsor is still 1440 psig as before, the HP is the same as that calculated earlier. The incremental HP is (10,243 – 8468) = 1775 HP more than that required for phase 1. At $2000 per installed HP, the incremental cost is 1775 × $2000 = $3.55 million compared to phase 1. Thus, for phase 2 the total incremental cost of additional looping over phase 1 and increased HP at the Windsor compressor station is ($38.13 – $25.02) + $3.55 = $16.66 million The costs of the initial case and the two expansion scenarios for the compressor station option and the pipe loop option are summarized in Table 5.1 and Table 5.2. Table 5.1 Windsor to Cardiff Pipeline Expansion—Compressor Station Option Phase

Flow, MMSCFD

Initial 1 2

188.41 238.41 288.41

Copyright 2005 by Taylor & Francis Group, LLC

Compressor BHP Required Windsor Avon Hart 7,064 8,468 10,243

— 3,659 8,055

— — 3,182

Compression Cost, $ million

Incremental Cost, $ million

14.13 24.25 42.96

— 10.13 18.71

2785_C005.fm Page 189 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

189

Table 5.2 Windsor to Cardiff Pipeline Expansion—Pipe Loop Option Pipe Compressor Pipe Compression Loop Total Incremental Flow, BHP Loop, Cost, Cost, Cost, Cost, Phase MMSCFD Windsor mi $ million $ million $ million $ million Initial 1 2

188.41 238.41 288.41

7,064 8,468 10,243

— 50.03 76.26

14.13 16.94 20.49

— 25.02 38.13

14.13 41.96 58.62

— 27.83 16.66

5.4 REDUCING POWER REQUIREMENTS In an existing pipeline for a given flow rate, we can calculate the HP required based upon the number of compressor stations, their suction and discharge pressures, and flow rate. Suppose we are interested in reducing the HP required and, hence, the annual operating cost of the pipeline. If the flow rate is not reduced, the only way power consumption can be reduced is to reduce the overall pressure drop between compressor stations. If the pipeline is 100 miles long and at a flow rate of 100 MMSCFD, an origin compressor station and an intermediate compressor station are required, each station operating at 900 psia suction and 1400 psia discharge pressures. The HP required will depend upon the compression ratio of (1400/900) or 1.56. Since the flow rate is constant, HP can be reduced by increasing the suction pressure or decreasing the discharge pressure, both of which reduce the compression ratio. Since the objective is to operate a gas pipeline at the highest possible pressure for efficiency, we will not reduce the discharge pressure. That leaves us the option of only increasing the suction pressure. Suction pressure can be increased by reducing the pressure drop in the pipeline segment upstream of the compressor station. Since the flow rate and pipe diameter are fixed, the pressure drop in a pipe segment can be decreased by installing a pipe loop. Therefore, looping a segment of pipeline, thereby reducing the pressure drop, will result in a decrease in HP and annual operating cost. We will illustrate this using an example.

Example 2 A natural gas (specific gravity = 0.60) pipeline is 130 mi long and is constructed of NPS 20, 0.500 in. wall thickness pipeline (MAOP = 1440 psig) that runs from Anaheim to Ventura. At a flow rate of 300 MMSCFD, an intermediate compressor at Brentwood (milepost 70) is needed. Calculate the total HP required for both compressor stations. In order to reduce the power consumption by 30% at the present flow rate, it is proposed to loop the pipeline. Calculate the extent of looping required. For simplicity, use the General Flow equation with a transmission factor F = 20 and compressibility factor of 0.90. The gas flow temperature is 60°F and base pressure is 14.7 psia. The base temperature is 60°F. The delivery pressure required at Ventura is 800 psig. The discharge pressure at Anaheim is 1440 psig, and the suction pressure is 900 psig. Use 80% isentropic efficiency and 95% mechanical efficiency for compressors. The gas specific heat ratio g = 1.4.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 190 Friday, April 1, 2005 3:30 PM

190

GAS PIPELINE HYDRAULICS

Solution Using General Flow Equation 2.4, calculate the discharge pressure required at the Brentwood compressor station for 800 psig delivery pressure at Ventura. 520 P12 − 814.72 300 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × 60 × 0.9

0.5

(19.0)2.5

Solving for the discharge pressure at Brentwood, P1 = 1216 psia Next, calculate the suction pressure at Brentwood, applying the General Flow equation to the pipeline segment 70 mi long between Anaheim and Brentwood. 520 1454.72 − P22 300 × 10 = 38.77 × 20 14.7 0.6 × 520 × 70 × 0.9

0.5

6

(19.0)2.5

Solving for the suction pressure at Brentwood, P2 = 1079.25 psia The compression ratio at Anaheim is r=

1454.7 = 1.59 914.7

The HP required at Anaheim is calculated using Equation 4.15: 0.40 1 + 0.9 1 1454.7 1.40 1.40 520 ( ) − HP = 0.0857 × 300 1 2 0.8 914.7 = 7876 0.40

where the compressibility factor for suction conditions is assumed to be 1.0. Considering a mechanical efficiency of 95%, the BHP required at Anaheim is BHP =

7876 = 8291 0.95

Similarly, calculate the compression ratio and BHP for the Brentwood compressor station. The compression ratio at Brentwood is r=

1216 = 1.127 1079.25

The HP required at Brentwood is 0.40 1.40 1.9 1 HP = 0.0857 × 300 (520) (1.127) 1.40 − 1 = 1931 0.40 2 0.8

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 191 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

191

and the BHP is BHP =

1931 = 2033 0.95

By looping the pipe segment between Anaheim and Brentwood using NPS 20 pipe, the flow rate through each pipe will be one-half the inlet flow of 300 MMSCFD at Anaheim. The suction pressure at Brentwood is calculated using the General Flow equation as 520 1454.72 − P22 150 × 10 = 38.77 × 20 0.6 × 520 × 70 × 0.9 . 14 7

0.5

(19.0)2.5

6

Solving for P2, we get P2 = 1371 psia Since this pressure is more than the discharge pressure of 1216 psia calculated earlier for Brentwood, we conclude that the Brentwood station will not be needed if we loop the entire 70 mi pipe segment from Anaheim to Brentwood. This would reduce the total BHP required to 8291 from (8291 + 2033) calculated earlier for the Anaheim and Brentwood compressor stations. The reduction in BHP is ∆BHP =

2033 = 0.197, or 19.7% 8291 + 2033

Since the objective is to reduce the power consumption by 30%, we must do more than just loop the pipe segment between Anaheim and Brentwood. We will recalculate the discharge pressure at Anaheim without the Brentwood compressor station, such that the delivery pressure at Ventura is 800 psig. The reduced discharge pressure at Anaheim due to the 70 mi pipe loop will reduce the compression ratio and, hence, the HP at Anaheim. Using the General Flow equation for the pipe segment between Anaheim and Brentwood, the downstream pressure at Brentwood must equal the 1216 psia calculated earlier to ensure 800 psig delivery at Ventura. Therefore, considering half the total flow rate through each NPS 20 pipe section of the loop, 520 P12 − 12162 150 × 10 = 38.77 × 20 14.7 0.6 × 520 × 70 × 0.9 6

Solving for the discharge pressure at Anaheim, we get P1 = 1310 psia

Copyright 2005 by Taylor & Francis Group, LLC

0.5

(19.0)2.5

2785_C005.fm Page 192 Friday, April 1, 2005 3:30 PM

192

GAS PIPELINE HYDRAULICS

This reduced discharge pressure at Anaheim causes the compression ratio to be r=

1310 = 1.432 914.7

The revised HP at Anaheim is calculated as 0.40 1.40 1.9 1 HP = 0.0857 × 300 (1.432) 1.40 − 1 = 6003 (520) 0.40 2 0.8

BHP =

6003 = 6319 0.95

Therefore, the total reduction in HP is (8291 + 2033) – 6319 = 4005 Or the percentage reduction in HP is 4005 = 0.39 or 39% 8291 + 2033 This is well above the 30% reduction in power required. If we reduce the loop pipe length slightly from 70 mi, we will realize the required 30% reduction. This is left as an exercise for the reader.

5.5 LOOPING IN DISTRIBUTION PIPING Another example of pipe loops is as follows. Consider a distribution piping system as shown in Figure 5.7. Gas enters the pipeline at A at a flow rate of 60 MMSCFD, and after making gas deliveries at B of 20 MMSCFD and at C of 30 MMSCFD, the remaining 10 MMSCFD of gas proceeds to D, where an additional 10 MMSCFD enters the pipeline, which is delivered to the terminus at E. The last segment of pipe has a flow rate of (60 − 20 − 30 + 10) or 20 MMSCFD. Suppose it is desired to bring in an extra 10 MMSCFD gas at D so that the delivery at E is increased to 30 MMSCFD. If the delivery pressure at E is to remain the same at 600 psig, it is clear that the pressure at D will need to be increased to handle the extra flow rate in pipe segment DE. This, in turn, will raise all pressures upstream of D. Thus, the pressures 60 MMSCFD

40 MMSCFD B

A

20 MMSCFD Figure 5.7

Distribution piping.

Copyright 2005 by Taylor & Francis Group, LLC

10 MMSCFD C

30 MMSCFD

20 MMSCFD D

10 MMSCFD

600 psig E

2785_C005.fm Page 193 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

193

at A, B, and C will all increase, resulting in an increased HP requirement at A. However, by looping the section DE, we can maintain all pressures the same as before. Assume that for the initial case, where the injection at D is 10 MMSCFD, the pressure at D is 900 psig. The delivery pressure at E is to be maintained constant at 600 psig. If the entire length of pipe DE is looped with an identical pipe size, the equivalent diameter Deq is such that at 30 MMSCFD, the pressure drop in the diameter Deq is the same as the pressure drop in the original pipe diameter D at 20 MMSCFD. From General Flow Equation 2.4, the flow rate is directly proportional to the square root of (P12 – P22) and also to the pipe diameter raised to the power of 2.5, keeping everything else the same. P1 and P2 are the upstream and downstream pressures in a pipe segment. Since we want the upstream and downstream pressures for the pipe segment DE to be the same at both flow rates, at 20 MMSCFD,

(

)

(

)

20 = C P12 − P22

0.5

D 2.5

(A)

Deq 2.5

(B)

and at 30 MMSCFD, 30 = C P12 − P22

0.5

where C is a constant for the pipe segment. By dividing one equation by the other, we get Deq D

2.5

=

30 = 1.5 20

The equivalent diameter is 1

Deq = D(1.5) 2.5 If the initial pipe size of DE was 12.00 in. inside diameter, we need an equivalent diameter of 1

Deq = 12 × (1.5) 2.5 = 14.11 in. Next, we need to determine the loop diameter required that will produce the equivalent diameter just calculated. Since the pressure drop in each pipe loop is the same, if Q1 and Q2 represent the flow rates in the main pipe and loop respectively, Q1 + Q2 = 30

Copyright 2005 by Taylor & Francis Group, LLC

(5.5)

2785_C005.fm Page 194 Friday, April 1, 2005 3:30 PM

194

GAS PIPELINE HYDRAULICS

and Q1 Q2 30 = = 2.5 2.5 Deq (12) ( D)2.5

(5.6)

where D is the loop diameter to be calculated and the main pipe is 12.00 in. diameter. Solving for Q1, we get 12 Q1 = 30 14.11

2.5

= 20.01 MMSCFD

and the flow rate through the loop is Q2 = 30 − 20.01 = 9.99 MMSCFD Therefore, from Equation 5.6, 9.99 30 = 2.5 D (14.11)2.5 Solving for D, we get 1

9.99 2.5 D= × 14.11 = 9.09 in. 30 Therefore, by looping the entire length DE of the existing 12 in. diameter pipe with a pipe having an inside diameter of 9.09 in., we will maintain the same pressure at all points as before. A slightly different case of looping is one in which the inlet flow at A needs to be increased so that the increased volume can be delivered at B, while keeping all pressures the same as before. Suppose the delivery volume at B needs to be increased to 30 MMSCFD, without changing other deliveries or receipt. The inlet volume at A will increase from 60 MMSCFD to 70 MMSCFD, and the delivery volume at B will increase from 20 MMSCFD to 30 MMSCFD. We will loop the section AB such that the pressure at A and B remain the same as before, so that the volumes and pressures at all points downstream of B remain the same. This will be illustrated by calculating the pressures and the size of the pipe loop required in the next example. Example 3 In a gas distribution pipeline, 60 MMSCFD enters the pipeline at A, as shown in Figure 5.8. If the delivery at B is increased from 20 MMSCFD to 30 MMSCFD by increasing the inlet flow at A, keeping all downstream flow rates the same, calculate the looping necessary for section AB to ensure pressures are not changed throughout the pipeline. Pipe AB is NPS 14, 0.250 in. wall thickness; BC is NPS 12, 0.250 in.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 195 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

70 MMSCFD

40 MMSCFD B

A

20 MMSCFD

10 MMSCFD

30 MMSCFD

600 psig

D

C

30 MMSCFD Figure 5.8

195

E

10 MMSCFD

Looping a distribution piping.

wall thickness; CD is NPS 10, 0.250 in. wall thickness; and DE is NPS 12, 0.250 in. wall thickness. The delivery pressure at E is fixed at 600 psig. The pipe lengths are as follow: AB = 12 mi BC = 18 mi CD = 20 mi DE = 8 mi The gas gravity is 0.60, and the flow temperature is 60°F. The compressibility factor and transmission factor can be assumed to be 0.85 and 20, respectively, throughout the pipeline. The base pressure and base temperature are 14.7 psia and 60°F, respectively. Solution First, the pressures at A and B must be calculated for the initial flow rates. Starting at E, for a delivery pressure of 600 psig at E, the pressures at D, C, and B will be calculated sequentially. Applying the General Flow equation for the 8 mi section DE of inside diameter 12.25 in. and at a flow rate of 20 MMSCFD, 520 PD2 − 614.72 20 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × 8 × 0.85

0.5

(12.25)2.5

Solving for PD, we get PD = 618.02 psia = 603 psig Next, calculate the pressure at C, considering 10 MMSCFD flow through the 20 mi section of pipe CD, of inside diameter 10.25 in.: 520 PC2 − 618.022 10 × 10 = 38.77 × 20 14.7 0.6 × 520 × 20 × 0.85 6

Solving for PC, we get PC = 623.04 psia = 608.34 psig

Copyright 2005 by Taylor & Francis Group, LLC

0.5

(10.25)2.5

2785_C005.fm Page 196 Friday, April 1, 2005 3:30 PM

196

GAS PIPELINE HYDRAULICS

Similarly, the pressure at B is calculated considering 40 MMSCFD flow through the 18 mi section of pipe BC, with an inside diameter 12.25 in.: 520 PB2 − 623.04 2 40 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × 18 × 0.85

0.5

(12.25)2.5

Solving for PB, we get PB = 651.90 psia = 637.20 psig Next, calculate the pressure at A considering 60 MMSCFD flow through the 12 mi section of pipe AB, with an inside diameter 13.5 in.: 520 PA2 − 651.90 2 60 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × 12 × 0.85

0.5

(13.5)2.5

Solving for PA, we get PA = 677.45 psia = 662.75 psig Therefore, the pipe section AB, when flowing 60 MMSCFD of gas, has the following pressures: PA = 677.45 psia PB = 651.90 psia When the delivery rate at B is increased from 20 MMSCFD to 30 MMSCFD, the flow rate in pipe segment AB increases from 60 MMSCFD to 70 MMSCFD. Since the pressures at A and B are to remain the same as before, the pipe segment AB must be looped to reduce the pressure drop at the higher flow rate. We will assume the entire 12 mi length will be looped. Next, we calculate the equivalent diameter required for segment AB, using the General Flow equation, so the pressures at A and B are the same as before. 520 677.452 − 651.90 2 70 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × 12 × 0.85

0.5

( D)2.5

Solving for the diameter, D = 14.36 in. The equivalent diameter of the looped line AB must be 14.36 in. to keep pressures the same as calculated. From Equation 3.17 and Equation 3.18, the diameter of the loop can be calculated, knowing the equivalent diameter just calculated. From Equation 3.17, 1

1 + Const1 2 5 14.36 = 13.50 Const1

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 197 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

197

Solving for Const1 Const1 = 5.99 From Equation 3.18, and since L1 = L2 = 12 mi, 13.5 5.99 = D2

5

Solving for the pipe loop diameter D2, D2 = 6.6 in. Therefore, the pipe section AB must be looped with a pipe of inside diameter 6.6 in. for the entire length of 12 mi. We could also increase the loop diameter and reduce the pipe length that is looped to get the same effect. For example, increasing the loop diameter to 10 in. will reduce the length of looping needed. Suppose we decide on an NPS 10, 0.250 in. wall thickness pipe for the loop length of L mi. upstream of B. The equivalent diameter will be calculated using Equation 3.17 and Equation 3.18. From Equation 3.18, 13.5 Const1 = 10.25

5

= 1.9908

and from Equation 3.17, the equivalent diameter is 1

1 + 1.9908 2 5 = 15.89 in. De = 13.50 1.9908 The pressure at the start of the loop will be calculated from General Flow Equation 2.4: 520 P 2 − 651.90 2 70 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × L × 0.85

0.5

(15.89)2.5

Simplifying, P2 – 651.90 2 = 1705L

(5.7)

Next, consider the unlooped portion of pipe AB from A to the starting point of the loop. Using General Flow Equation 2.4, we get 677.452 − P 2 520 70 × 10 6 = 38.77 × 20 14.7 0.6 × 520 × (12 − L ) × 0.85

Copyright 2005 by Taylor & Francis Group, LLC

0.5

(13.5)2.5

2785_C005.fm Page 198 Friday, April 1, 2005 3:30 PM

198

GAS PIPELINE HYDRAULICS

Simplifying, 677.452 – P2 = 3851.92 (12 – L)

(5.8)

Eliminating P from Equation 5.7 and Equation 5.8 and solving for L, we get L = 5.71 mi Therefore, by looping the existing NPS 14 pipe from A to B with an identical NPS 14 pipe, 5.71 mi long (measured upstream from B), the pressures will be the same as before the increased delivery volume at B.

5.6 SUMMARY We discussed two ways to increase the throughput of a gas pipeline: using intermediate compressor stations and installing pipe loops. With intermediate compressor stations, the flow rate can be increased to fully utilize pipe MAOP. However, adding compressor stations causes increased capital cost as well as annual operating and maintenance costs. On the other hand, by installing a pipe loop, the effective diameter of the pipe is increased, resulting in a lower pressure drop. Therefore, additional flow rate can be realized without installing an intermediate compressor station. Looping an existing pipeline causes increase in capital but very little increase in operating and maintenance costs compared to installing intermediate compressor stations. We also discussed how the HP required can be reduced by installing a pipe loop. On distribution piping, an example of increasing delivery rate to certain locations using pipe loops, without changing pipe pressures in the rest of the pipeline, was also illustrated.

PROBLEMS 1. A natural gas pipeline from Compton to Merced is 100 mi long and is constructed of NPS 14, 0.250 in. wall thickness. The pipeline elevation profile is essentially flat. The MAOP of the pipeline is 1280 psig. The gas delivery pressure at Merced is 600 psig. What is the maximum pipeline throughput with an origin compressor station at Compton? The gas gravity is 0.6 and gas flowing temperature is 80°F. Use the Colebrook equation for pressure drop with a friction factor of 0.01. The compressibility factor can be assumed to be constant at 0.88. If the flow rate increases by 50 MMSCFD, calculate the increased HP required at Compton and the HP required at an intermediate compressor station at Vale. Instead of the intermediate compressor station at Vale, a portion of the pipe is looped. What length of NPS 14 loop will be needed? The base pressure and base temperature are 14.7 psia and 60°F, respectively. The compressor isentropic efficiency = 0.8, and the mechanical efficiency of the compressor driver is 0.95. 2. A natural gas (specific gravity = 0.60) pipeline is 120 mi long and is constructed of NPS 20, 0.500 in. wall thickness (MAOP = 1000 psig), and runs from Akers to Coburn. At a flow rate of 250 MMSCFD, an intermediate compressor at

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 199 Friday, April 1, 2005 3:30 PM

PIPE LOOPS VERSUS COMPRESSION

199

Bradley (milepost 65) is required. Calculate the total HP required. In order to reduce the power consumption by 20% at the present flow rate, it is proposed to loop the pipeline. Calculate the length of looping required. Use the Panhandle A equation with 95% efficiency. The compressibility factor can be assumed constant at 0.90. The gas flow temperature is 70°F, and the base pressure and base temperature are 14.7 psia and 60°F, respectively. The delivery pressure required at Coburn is 750 psig. The discharge pressure at Akers is 1000 psig, and the suction pressure is 850 psig. Use 80% isentropic efficiency and 95% mechanical efficiency for the compressors. 3. In a gas distribution pipeline, similar to that shown in Figure 5.8, gas enters the pipeline at A at a flow rate of 50 MMSCFD. At B and C, deliveries of 10 MMSCFD and 20 MMSCFD are made. At D, an additional volume of gas at 15 MMSCFD enters the pipeline. Calculate the pressures at the various pipe nodes A, B, C, and D, considering a delivery pressure of 500 psig at E. If the incoming volume at D is increased to 25 MMSCFD and all pressures are to remain the same, how much of the pipe DE should be looped? The pipe lengths are as follows: AB: NPS 12, 0.250 length = 18 mi BC: NPS 10, 0.250 length = 24 mi CD: NPS 8, 0.250 length = 16 mi DE: NPS 12, 0.250 length = 20 mi

REFERENCES 1. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 2. Engineering Data Book, 10th ed., Gas Processors Suppliers Association, Tulsa, OK, 1994. 3. Pipeline Design for Hydrocarbon Gases and Liquids, American Society of Civil Engineers, New York, 1975.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 200 Friday, April 1, 2005 3:30 PM

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 201 Friday, April 1, 2005 3:29 PM

CHAPTER

6

Pipe Analysis In this chapter we will discuss the mechanical strength needed for a pipeline transporting gas. We will analyze the impact of pipe diameter, wall thickness, material of construction, and specific safety requirements dictated by design codes and state and federal regulations. Also covered will be testing requirements and classification of pipelines based upon their proximity to human dwellings and industrial establishments and population density. The importance of mainline block valves and calculation of blowdown time to isolate sections of a gas pipeline will also be discussed.

6.1 PIPE WALL THICKNESS In Chapter 3 we calculated the pressure needed to transport a given volume of gas through a pipeline. The internal pressure in a pipe causes the pipe wall to be stressed, and if allowed to reach the yield strength of the pipe material, it could cause permanent deformation of the pipe and ultimate failure. Obviously, the pipe should have sufficient strength to handle the internal pressure safely. In addition to the internal pressure due to gas flowing through the pipe, the pipe might also be subjected to external pressure. External pressure can result from the weight of the soil above the pipe in a buried pipeline and also by the loads transmitted from vehicular traffic in areas where the pipeline is located below roads, highways, and railroads. The deeper the pipe is buried, the higher will be the soil load on the pipe. However, the pressure transmitted to the pipe due to vehicles above ground will diminish with the depth of the pipe below the ground surface. Thus, the external pressure due to vehicular loads on a buried pipeline that is 6 ft below ground will be less than that on a pipeline that is at a depth of 4 ft. In most cases involving buried pipelines transporting gas and other compressible fluids, the effect of the internal pressure is more than that of external loads. Therefore, the necessary minimum wall thickness will be dictated by the internal pressure in a gas pipeline. The minimum wall thickness required to withstand the internal pressure in a gas pipeline will depend upon the pressure, pipe diameter, and pipe material. The larger 201

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 202 Friday, April 1, 2005 3:29 PM

202

GAS PIPELINE HYDRAULICS

the pressure or diameter, the larger would be the wall thickness required. Higherstrength steel pipes will require less wall thickness to withstand the given pressure compared to low-strength materials. The commonly used formula to determine the wall thickness for internal pressure is known as Barlow’s equation. This equation has been modified to take into account design factors and type of pipe joints (seamless, welded, etc.) and is incorporated into design codes such as DOT Code of Federal Regulations Part 192 and ASME B31.8 Standards. See Chapter 9 for a full list of design codes and standards used in the design, construction, and operation of gas pipelines.

6.2 BARLOW’S EQUATION When a circular pipe is subject to internal pressure, the pipe material at any point will have two stress components at right angles to each other. The larger of the two stresses is known as the hoop stress and acts along the circumferential direction. Hence, it is also called the circumferential stress. The other stress is the longitudinal stress, also known as the axial stress, which acts in a direction parallel to the pipe axis. Figure 6.1 shows a cross section of a pipe subject to internal pressure. An element of the pipe wall material is shown with the two stresses Sh and Sa in perpendicular directions. Both stresses will increase as the internal pressure is increased. As will be shown shortly, the hoop stress Sh is the larger of the two stresses and, hence, will govern the minimum wall thickness required for a given internal pressure. In its basic form, Barlow’s equation relates the hoop stress in the pipe wall to the internal pressure, pipe diameter, and wall thickness as follows: Sh =

PD 2t

(6.1)

L

Axial stress - Sa

Axial stress - Sa

Sh Pressure - P

Diameter - D

Figure 6.1

Stresses in pipe subject to internal pressure.

Copyright 2005 by Taylor & Francis Group, LLC

s - Sh

tres

ps Hoo

Sh

2785_C006.fm Page 203 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

where Sh = P = D = t =

203

hoop or circumferential stress in pipe material, psi internal pressure, psi pipe outside diameter, in. pipe wall thickness, in.

Similar to Equation 6.1, the axial (or longitudinal) stress, Sa, is given by the following equation: Sa =

PD 4t

(6.2)

Note that in these equations the pipe diameter used is the outside diameter, not the inside diameter as we used in Chapter 2 and Chapter 3. For example, consider an NPS 20 pipe, 0.500 in. wall thickness, that is subject to an internal gas pressure of 1200 psig. The pipe wall material will be stressed in the circumferential direction by the hoop stress given by Equation 6.1 as follows: Sh =

1200 × 20 = 24,000 psig 2 × 0.500

and in accordance with Equation 6.2, the axial stress in the pipe wall is Sa =

1200 × 20 = 12,000 psig 4 × 0.500

Barlow’s equation is valid only for thin-walled cylindrical pipes. Most pipelines transporting gases and liquids generally fall in this category. There are instances in which pipes carrying gases and petroleum liquids, subject to high external loads, such as deep submarine pipelines, may be classified as thick-walled pipes. The governing equations for such thick-walled pipes are different and more complex. We will introduce these formulas for information only. 6.3 THICK-WALLED PIPES Consider a thick-walled pipe with an outside diameter DO and inside diameter of Di, subject to an internal pressure of P. The greatest stress in the pipe wall will be found to occur in the circumferential direction near the inner surface of the pipe. This stress can be calculated from the following equation: Smax =

(

P Do2 + Di2

(D

2 o

−D

2 i

)

)

(6.3)

The pipe wall thickness is t=

Copyright 2005 by Taylor & Francis Group, LLC

Do − Di 2

(6.4)

2785_C006.fm Page 204 Friday, April 1, 2005 3:29 PM

204

GAS PIPELINE HYDRAULICS

Rewriting Equation 6.3 in terms of outside diameter and wall thickness, we get D 2 + ( Do − 2t )2 Smax = P o2 2 Do − ( Do − 2t ) Simplifying further,

Smax

PDo = 2t

( ) ( ) ( )

t 1 − Do + 2 1 + Dt o

2

t Do

(6.5)

In the limiting case, a thin-walled pipe is one in which the wall thickness is very small compared to the diameter Do. In this case (t /D) is small compared to 1 and, therefore, can be neglected in Equation 6.5. Therefore, the approximation for thinwalled pipes from Equation 6.5 becomes Smax =

PDo 2t

which is the same as Barlow’s Equation 6.1 for hoop stress. Example 1 A gas pipeline is subject to an internal pressure of 1400 psig. It is constructed of steel pipe with 24 in. outside diameter and 0.75 in. wall thickness. Calculate the maximum hoop stress in the pipeline, considering both the thin-walled approach and the thick-walled equation. What is the error in assuming that the pipe is thin walled? Solution Pipe inside diameter = 24 – 2 × 0.75 = 22.5 in. From Equation 6.1 for thin-walled pipe, Barlow’s equation gives the maximum hoop stress as Sh =

1400 × 24 = 22,400 psig 2 × 0.75

Considering the thick-walled pipe formula Equation 6.3, Smax =

1400(24 2 + 22.52 ) = 21,723 psig (24 2 − 22.52 )

Therefore, by assuming thin-walled pipe, the hoop stress is overestimated by approximately 22, 400 − 21, 723 = 0.0312 or 3.12% 21, 723

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 205 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

205

6.4 DERIVATION OF BARLOW’S EQUATION Since Barlow’s equation is the basic equation for pipes under internal pressure, it is appropriate to understand how the formula is derived, which is the subject of this section. Consider a circular pipe of length L, outside diameter D, and wall thickness t as shown in Figure 6.1. We consider the cross section of one-half portion of this pipe. The pipe is subject to an internal pressure of P psig. Within the pipe material, the hoop stress Sh and the axial stress Sa act at right angles to each other as shown. Considering the one-half section of the pipe, for balancing the forces in the direction of the hoop stress Sh, we can say that Sh, acting on the two rectangular areas L × t , balances the internal pressure on the projected area D × L. Therefore, P × D × L = Sh × L × t × 2

(6.6)

Solving for Sh , we get the derivation of Equation 6.1 as Sh =

PD 2t

Now we will look at the balancing of longitudinal forces. The internal pressure P acting on the cross-sectional area of pipe π4 D 2 produces the bursting force. This is balanced by the axial resisting force Sa acting on the area π Dt . Therefore,

π 2 D = Sa × π Dt 4

(6.7)

Solving for Sa, we get the derivation of Equation 6.2 as Sa =

PD 4t

It can be seen from the preceding equations that the hoop stress is twice the axial stress and, therefore, is the governing stress. Consider a pipe with 20 in. outside diameter and 0.500 in. wall thickness subject to an internal pressure of 1000 psig. From Barlow’s Equation 6.1 and Equation 6.2, we calculate the hoop stress and axial stress as follows: Sh =

1000 × 20 = 20,000 psig 2 × 0.500

Sa =

1000 × 20 = 10,000 psig 4 × 0.500

Therefore, we are able to determine the stress levels in the pipe material for a given internal pressure, pipe diameter, and wall thickness. If the above-calculated

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 206 Friday, April 1, 2005 3:29 PM

206

GAS PIPELINE HYDRAULICS

values are within the stress limits of the pipe material, we can conclude that the NPS 20 pipe with 0.500 in. wall thickness is adequate for the internal pressure of 1000 psig. The yield stress of the pipe material represents the stress at which the pipe material yields and undergoes permanent deformation. Therefore, we must ensure that the stress calculations above do not come dangerously close to the yield stress. Frequently, we have to solve the reverse problem of determining the wall thickness of a pipeline for a given pressure. For example, suppose the pipe is constructed of steel with a yield strength of 52,000 psi and we are required to determine what wall thickness is needed for NPS 20 pipe to withstand 1400 psig internal pressure. If we are allowed to stress the pipe material to no more than 60% of the yield stress, we can easily calculate the minimum wall thickness required using Equation 6.1, as follows: 0.6 × 52, 000 =

1400 × 20 2t

Here, we have equated the hoop stress per Barlow’s equation to 60% yield strength of the pipe material. Solving for pipe wall thickness, we get t = 0.4487 in. Suppose we used the nearest standard wall thickness of 0.500 in. The actual hoop stress can then be calculated from Barlow’s equation as Sh =

1400 × 20 = 28,000 psi 2 × 0.5

28,000 Therefore, the pipe will be stressed to 52 ,000 = 0.54 or 54% of yield stress, which is less than the 60% we started with. Incidentally, the actual axial or longitudinal stress in the preceding example will be one-half the hoop stress or 14,000 psi. Therefore, in this basic example, we used Barlow’s equation to calculate the pipe wall thickness required for a NPS 20 pipe to withstand an internal pressure of 1400 psig without stressing the pipe material beyond 60% of its yield strength. In the foregoing, we arbitrarily picked 60% of the yield stress of pipe material to calculate the pipe wall thickness. We did not use 100% of the yield stress because, in this case, the pipe material would yield at the given pressure, which obviously cannot be allowed. In design, we generally use a design factor that is a number less than 1.00 that represents the fraction of the yield stress of the pipe material that the pipe can be stressed to. Gas pipelines are designed with various design factors ranging from 0.4 to 0.72. This means that the pipe hoop stress is allowed to be between 40 and 72% of the yield strength of pipe material. The actual percentage will depend on various factors and will be discussed shortly. The yield stress used in the calculation of pipe wall thickness is called the specified minimum yield strength (SMYS) of pipe material. Thus, in the preceding example, we calculated the pipe wall thickness based on a design factor of 0.6 or allowed the pipe stress to go up to 60% of the SMYS.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 207 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

207

Table 6.1 Pipe Material and Yield Strength Pipe Material API 5LX Grade

Specified Minimum Yield Strength (SMYS), psi

X42 X46 X52 X56 X60 X65 X70 X80 X90

42,000 46,000 52,000 56,000 60,000 65,000 70,000 80,000 90,000

6.5 PIPE MATERIAL AND GRADE Steel pipes used in gas pipeline systems generally conform to API 5L and 5LX specifications. These are manufactured in grades ranging from X42 to X90 with SMYS, as shown in Table 6.1. Sometimes API 5L grade B pipe with 35,000 psi SMYS is also used in certain installations.

6.6 INTERNAL DESIGN PRESSURE EQUATION We indicated earlier in this chapter that Barlow’s equation, in a modified form, is used in designing gas pipelines. The following form of Barlow’s equation is used in design codes for petroleum transportation systems to calculate the allowable internal pressure in a pipeline based upon given diameter, wall thickness, and pipe material. P=

2tSEFT D

(6.8)

where P = internal pipe design pressure, psig D = pipe outside diameter, in. t = pipe wall thickness, in. S = specified minimum yield strength (SMYS) of pipe material, psig E = seam joint factor, 1.0 for seamless and submerged arc welded (SAW) pipes. F = design factor, usually 0.72 for cross-country gas pipelines, but can be as low as 0.4, depending on class location and type of construction T = temperature deration factor = 1.00 for temperatures below 250°F It must be noted that in the foregoing, we used the outside diameter of the pipe and not the inside diameter as used in Chapter 2 and Chapter 3 for pressure drop calculations.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 208 Friday, April 1, 2005 3:29 PM

208

GAS PIPELINE HYDRAULICS

Table 6.2 Pipe Seam Joint Factors Specification ASTM A53

ASTM ASTM ASTM ASTM ASTM ASTM ASTM ASTM

A106 A134 A135 A139 A211 A333 A333 A381

ASTM A671 ASTM A672 ASTM A691 API 5L

API 5LX

API 5LS

Seam Joint Factor (E)

Pipe Class Seamless Electric Resistance Welded Furnace Lap Welded Furnace Butt Welded Seamless Electric Fusion Arc Welded Electric Resistance Welded Electric Fusion Welded Spiral Welded Pipe Seamless Welded Double Submerged Arc Welded Electric-Fusion-Welded Electric-Fusion-Welded Electric-Fusion-Welded Seamless Electric Resistance Welded Electric Flash Welded Submerged Arc Welded Furnace Lap Welded Furnace Butt Welded Seamless Electric Resistance Welded Electric Flash Welded Submerged Arc Welded Electric Resistance Welded Submerged Arc Welded

1 1 0.8 0.6 1 0.8 1 0.8 0.8 1 1 1 1 1 1 1 1 1 1 0.8 0.6 1 1 1 1 1 1

The seam joint factor E used in Equation (6.8) varies with the type of pipe material and welding employed. Seam joint factors are given in Table 6.2 for the most commonly used pipe and joint types. The internal design pressure calculated from Equation (6.8) is known as the maximum allowable operating pressure (MAOP) of the pipeline. This term has been shortened to maximum operating pressure (MOP) in recent years. Throughout this book we will use MOP and MAOP interchangeably. The design factor F has values ranging from 0.4 to 0.72, as mentioned earlier. Table 6.3 lists the values of the design factor based upon class locations. The class locations, in turn, depend on the population density in the vicinity of the pipeline. Table 6.3 Design Factors for Steel Pipe

Copyright 2005 by Taylor & Francis Group, LLC

Class Location

Design Factor, F

1 2 3 4

0.72 0.60 0.50 0.40

2785_C006.fm Page 209 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

209

220 yards pipe centerline 220 yards

1 mi Figure 6.2

Class location unit.

6.7 CLASS LOCATION The following definitions of class 1 through class 4 are taken from DOT 49 CFR, Part 192 (see Reference section for details). The class location unit (CLU) is defined as an area that extends 220 yards on either side of the center line of a 1-mi section of pipe, as indicated in Figure 6.2. Class 1 Offshore gas pipelines are Class 1 locations. For onshore pipelines, any class location unit that has 10 or fewer buildings intended for human occupancy is termed Class 1. Class 2 This is any class location unit that has more than 10 but fewer than 46 buildings intended for human occupancy. Class 3 This is any class location unit that has 46 or more buildings intended for human occupancy or an area where the pipeline is within 100 yards of a building or a playground, recreation area, outdoor theatre, or other place of public assembly that is occupied by 20 or more people at least 5 days a week for 10 weeks in any 12-month period. The days and weeks need not be consecutive. Class 4 This is any class location unit where buildings with four or more stories above ground exist.

The temperature deration factor T is equal to 1.00 up to gas temperature 250°F, as indicated in Table 6.4. Table 6.4 Temperature Deration Factors Temperature °F

°C

Deration Factor T

250 or less 300 350 400 450

121 or less 149 177 204 232

1.000 0.967 0.033 0.900 0.867

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 210 Friday, April 1, 2005 3:29 PM

210

GAS PIPELINE HYDRAULICS

Example 2 A gas pipeline is constructed of API 5L X65 steel, NPS 16, 0.250 in. wall thickness. Calculate the MAOP of this pipeline for class 1 through class 4 locations. Use a temperature deration factor of 1.00. Solution Using Equation 6.8, the MAOP is given by P=

2 × 0.250 × 65, 000 × 1.0 × 0.72 × 1.0 = 1462.5 psig for class 1 16

Similarly, MAOP = 1462.5 ×

0.6 = 1218.8 psig for class 2 0.72

MAOP = 1462.5 ×

0.5 = 1015.62 psig for class 3 0.72

MAOP = 1462.5 ×

0.4 = 812.5 psig for class 4 0.72

6.8 MAINLINE VALVES Mainline valves are installed in gas pipelines so that portions of the pipeline can be isolated for hydrostatic testing and maintenance. Valves are also necessary to separate sections of pipe and minimize gas loss that can occur due to pipe rupture from construction damage. Design codes specify the spacing of these valves based upon class location, which in turn depends on the population density around the pipeline. The following lists the maximum spacing between mainline valves in gas transmission piping. These are taken from ASME B31.8 code. Class Location 1 2 3 4

Valve Spacing 20 15 10 5

miles miles miles miles

It can be seen from the preceding that the valve spacing is shorter as the pipeline traverses high-population areas. This is necessary as a safety feature to protect the inhabitants in the vicinity of the pipeline by restricting the amount of gas that might escape due to rupture of the pipeline. These mainline valves must be full-opening, through-conduit type valves such that scraper pigs and inspection tools can pass through these valves without any obstruction. Therefore, ball valves and gate valves are used of the welded construction rather than flanged type. Buried valves have

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 211 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

211

extended stems with elevated valve operators located above ground, with lubrication and bleed lines brought above ground for easy access and maintenance.

6.9 HYDROSTATIC TEST PRESSURE When a pipeline is designed to operate at a certain MOP, it must be tested to ensure that it is structurally sound and can withstand safely the internal pressure before being put into service. Generally, gas pipelines are hydrotested with water by filling the test section of the pipe with water and pumping the pressure up to a value higher than the MAOP and holding it at this test pressure for a period of 4 to 8 hours. The magnitude of the test pressure is specified by design code, and it is usually 125% of the operating pressure. Thus, a pipeline designed to operate continuously at 1000 psig will be hydrotested to a minimum pressure of 1250 psig. Consider a pipeline NPS 24, with 0.375 in. wall thickness, constructed of API 5L X65 pipe. Using a temperature deration factor of 1.00, we calculate the MOP of this pipeline from Equation 6.8 for class 1 location as follows: P=

2 × 0.375 × 65, 000 × 1.0 × 0.72 × 1.0 = 1462.5 psig 24

Since the pipe fittings and valves will be ANSI 600, we will establish an MOP of 1440 psig for this pipeline. Therefore, the hydrotest pressure will be 1.25 × 1440 = 1800 psig If the pipeline is designed to be below ground, the test pressure is held constant for a period of 8 hours, and it is thoroughly checked for leaks. Above-ground pipelines are tested for a period of 4 hours. If the design factor used in the MOP calculation is 0.72 (class 1), the hoop stress is allowed to reach 72% of the SMYS of pipe material. Testing this pipe at 125% of MOP will result in the hoop stress reaching a value of 1.25 × 0.72 = 0.90 or 90% of SMYS. Thus, by hydrotesting the pipe at 1.25 times the operating pressure, we are stressing the pipe material to 90% of the yield strength. Generally, the hydrotest pressure is given such that the hoop stress has a range of values, such as 90 to 95% of SMYS. Therefore, in the preceding example, the minimum and maximum hydrotest pressures will be as follows: Minimum hydrotest pressure = 1.25 × 1440 = 1800 psig Maximum hydrotest pressure = 1800 × (95/90) = 1900 psig It can be seen from Equation 6.1 that the 1800 psig internal pressure will cause a hoop stress of Sh =

Copyright 2005 by Taylor & Francis Group, LLC

1800 × 24 = 57,600 psi 2 × 0.375

2785_C006.fm Page 212 Friday, April 1, 2005 3:29 PM

212

GAS PIPELINE HYDRAULICS

Dividing this hoop stress by the SMYS, we get the lower limit of the hydrotest pressure as 57, 600 = 0.89 = 89% of SMYS 65, 000 Similarly, by proportion, the maximum hydrotest pressure of 1900 psig will cause a hoop stress of Sh =

1900 × 57, 600 = 60,800 = 94% of SMYS 1800

Therefore, in this example, the hydrotest envelope of 1800 to 1900 psig is equivalent to stressing the pipe in the range of 89 to 94% of SMYS. In the preceding analysis we have not taken into consideration the pipeline elevation profile in calculating the hydrotest pressures. Generally, a long pipeline is divided into test sections and the hydrotest pressures are established for each section, taking into account the elevations along the pipeline profile. The reason for subdividing the pipeline into sections for hydrotesting will be evident from the following example. Consider, for example, a pipeline 50 mi long with an elevation profile as shown in Figure 6.3. The elevation of the starting point, Norwalk, is 300 ft, whereas the pipeline terminus, Lakewood, is at an elevation of 1200 ft. If the entire 50 mi length of the pipeline were filled with water for hydrotesting, the static pressure difference between the two ends due to elevation will be as follows: Pressure difference = (1200 – 300) × 0.433 = 389.7 psig The factor 0.433 is the conversion factor from feet of water to pressure in psig. It can be seen that the pipe section at the low elevation point at Norwalk will be at a higher pressure than the pipeline at the high elevation end at Lakewood by almost 390 psig. Therefore, if we pump the water in the line to the required hydrotest

Hydrotest pressure: 1800 psig

1410 psig

1200 ft 390 psig Pipeline elevation profile 300 ft Norwalk Figure 6.3

50 mi

Pipeline with elevation profile—impact on hydrotest.

Copyright 2005 by Taylor & Francis Group, LLC

Lakewood

2785_C006.fm Page 213 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

213

pressure of 1800 psig at Norwalk, the corresponding water pressure at Lakewood will be 1800 – 390 = 1410 psig Conversely, if we pump the water in the line to the required hydrotest pressure of 1800 psig at Lakewood, the corresponding water pressure at Norwalk will be 1800 + 390 = 2190 psig This is shown in Figure 6.3. The pressure of 2190 psig at Norwalk will result in a hoop stress of Sh =

2190 × 24 = 70,080 psi 2 × 0.375

This is equivalent to 70, 080 = 1.08 = 108% of SMYS 65, 000 Obviously, we have exceeded the yield strength of the pipe material, and this is not acceptable. On the other hand, with 1800 psig test pressure at Norwalk, the corresponding test pressure at Lakewood is calculated to be 1410 psig. Even though the pipe section at the low end at Norwalk has the requisite test pressure (125% MOP), the pipe section at the higher elevation at Lakewood will see only 1410 × 125 = 98% MOP 1800 This will not be an acceptable hydrotest, because we have not been able to test the entire pipeline at the correct hydrotest pressure, which must be at least 125% of the MAOP. The solution to this dilemma is to break the length of 50 mi into several sections such that each section can be tested separately at the required test pressure. These test sections will have smaller elevation differences between the ends of the test sections. Therefore, each section will be hydrotested to pressures close to the required minimum pressure. Figure 6.4 shows such a pipeline subdivided into sections suitable for hydrotesting. Using the hydrotest envelope of 90 to 95% of SMYS, we will be able to adjust the test pressures for each section such that even with some elevation difference between the ends of each test section, the hydrotest pressures may be close to the required pressures. This will not be possible if we have one single test section with significant elevation difference between the two ends, as illustrated in Figure 6.4. Table 6.5 through Table 6.13 list the internal design pressure and hydrostatic test pressure for various pipe diameters and pipe materials ranging from X42 to X90.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 214 Friday, April 1, 2005 3:29 PM

214

GAS PIPELINE HYDRAULICS

Hydrotest pressure profile

Pipeline elevation profile San Juan Figure 6.4

250 km

Cadiz

Hydrotesting by subdividing pipeline.

Example 3 A gas pipeline, NPS 20, 0.500 in. wall thickness, is constructed of API 5L X52 pipe. (a) Calculate the design pressures for class 1 through class 4 locations. (b) What is the range of hydrotest pressures for each of these class locations? Assume joint factor = 1.00 and temperature deration factor = 1.00. Solution Using Equation 6.8, the internal design pressure is P=

2 × 0.500 × 52, 000 × 1.00 × 1.0 × F = 2600 F 20

where F = design factor = 0.72 for class 1 Therefore, the design pressures for class 1 through class 4 are as follows: Class Class Class Class

1 2 3 4

= = = =

2600 2600 2600 2600

× × × ×

0.72 0.60 0.50 0.40

= = = =

1872 1560 1300 1040

psig psig psig psig

The range of hydrotest pressures is such that the hoop stress will be between 90 and 95% of SMYS.

Copyright 2005 by Taylor & Francis Group, LLC

Wall Thickness in.

4.5

0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500

6.625

8.625

10.75

12.75

14.00

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09

42000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 3185 4529 5873 7137 2282 2556 3944 5131 1753 1942 2258 2847 1407 1727 2054 2813 1186 1565 1779 1926 2372 1080 1348 1620 1888 2160

2654 3774 4894 5947 1902 2130 3286 4275 1461 1619 1882 2372 1172 1439 1711 2344 988 1304 1482 1605 1976 900 1123 1350 1573 1800

2212 3145 4079 4956 1585 1775 2739 3563 1217 1349 1568 1977 977 1199 1426 1953 824 1087 1235 1337 1647 750 936 1125 1311 1500

1770 2516 3263 3965 1268 1420 2191 2850 974 1079 1254 1582 781 960 1141 1563 659 870 988 1070 1318 600 749 900 1049 1200

3982 5662 7342 8921 2853 3195 4930 6413 2191 2428 2822 3559 1758 2159 2567 3516 1482 1957 2224 2407 2965 1350 1685 2025 2360 2700

4203 5976 7749 9416 3011 3373 5204 6769 2313 2563 2979 3756 1856 2279 2709 3712 1565 2065 2347 2541 3129 1425 1778 2138 2491 2850

4424 6291 8157 9912 3170 3550 5477 7126 2435 2698 3136 3954 1953 2399 2852 3907 1647 2174 2471 2675 3294 1500 1872 2250 2622 3000

Copyright 2005 by Taylor & Francis Group, LLC

215

(continued )

2785_C006.fm Page 215 Friday, April 1, 2005 3:29 PM

Pipe Material API 5L X42 Diameter in.

PIPE ANALYSIS

Table 6.5 Pipeline Internal Design Pressures and Test Pressures

Pipe Material API 5L X42

SMYS

Wall Thickness in.

Weight lb/ft

16.00

0.250 0.312 0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750

42.05 52.27 62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25

18.00

20.00

22.00

24.00

26.00

Copyright 2005 by Taylor & Francis Group, LLC

42000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 945 1179 1418 1652 1890 840 1048 1260 1468 1680 943 1134 1321 1512 1699 1031 1375 1718 2062 945 1101 1260 1416 1575 1890 872 1163 1454 1745

788 983 1181 1377 1575 700 874 1050 1224 1400 786 945 1101 1260 1416 859 1145 1432 1718 788 918 1050 1180 1313 1575 727 969 1212 1454

656 819 984 1147 1313 583 728 875 1020 1167 655 788 918 1050 1180 716 955 1193 1432 656 765 875 984 1094 1313 606 808 1010 1212

525 655 788 918 1050 467 582 700 816 933 524 630 734 840 944 573 764 955 1145 525 612 700 787 875 1050 485 646 808 969

1181 1474 1772 2065 2363 1050 1310 1575 1835 2100 1179 1418 1652 1890 2124 1289 1718 2148 2577 1181 1377 1575 1770 1969 2363 1090 1454 1817 2181

1247 1556 1870 2180 2494 1108 1383 1663 1937 2217 1245 1496 1744 1995 2242 1360 1814 2267 2720 1247 1453 1663 1869 2078 2494 1151 1535 1918 2302

1313 1638 1969 2294 2625 1167 1456 1750 2039 2333 1310 1575 1835 2100 2360 1432 1909 2386 2864 1313 1530 1750 1967 2188 2625 1212 1615 2019 2423

GAS PIPELINE HYDRAULICS

Diameter in.

2785_C006.fm Page 216 Friday, April 1, 2005 3:29 PM

216

Table 6.5 Pipeline Internal Design Pressures and Test Pressures (Continued)

32.00

34.00

36.00

42.00

110.64 146.85 182.73 218.27 118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

810 1080 1350 1620 756 1008 1260 1512 709 945 1181 1418 667 889 1112 1334 630 840 1050 1260 540 720 900 1080 1440

675 900 1125 1350 630 840 1050 1260 591 788 984 1181 556 741 926 1112 525 700 875 1050 450 600 750 900 1200

563 750 938 1125 525 700 875 1050 492 656 820 984 463 618 772 926 438 583 729 875 375 500 625 750 1000

450 600 750 900 420 560 700 840 394 525 656 788 371 494 618 741 350 467 583 700 300 400 500 600 800

1013 1350 1688 2025 945 1260 1575 1890 886 1181 1477 1772 834 1112 1390 1668 788 1050 1313 1575 675 900 1125 1350 1800

1069 1425 1781 2138 998 1330 1663 1995 935 1247 1559 1870 880 1174 1467 1760 831 1108 1385 1663 713 950 1188 1425 1900

1125 1500 1875 2250 1050 1400 1750 2100 984 1313 1641 1969 926 1235 1544 1853 875 1167 1458 1750 750 1000 1250 1500 2000

217

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 217 Friday, April 1, 2005 3:29 PM

30.00

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

PIPE ANALYSIS

28.00

Pipe Material API 5L X46 Diameter in. 4.5

6.625

8.625

10.75

12.75

16.00

Copyright 2005 by Taylor & Francis Group, LLC

0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500 0.250 0.312

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09 42.05 52.27

46000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 3489 4961 6433 7816 2500 2800 4319 5619 1920 2127 2473 3118 1540 1892 2249 3081 1299 1714 1948 2109 2598 1183 1476 1774 2068 2366 1035 1292

2907 4134 5361 6514 2083 2333 3599 4683 1600 1773 2061 2598 1284 1576 1874 2567 1082 1429 1624 1758 2165 986 1230 1479 1723 1971 863 1076

2423 3445 4467 5428 1736 1944 3000 3902 1333 1477 1717 2165 1070 1314 1562 2140 902 1191 1353 1465 1804 821 1025 1232 1436 1643 719 897

1938 2756 3574 4342 1389 1555 2400 3122 1067 1182 1374 1732 856 1051 1249 1712 722 952 1082 1172 1443 657 820 986 1149 1314 575 718

4361 6201 8041 9770 3125 3499 5399 7024 2400 2659 3091 3898 1926 2365 2811 3851 1624 2143 2435 2637 3247 1479 1845 2218 2585 2957 1294 1615

4603 6545 8488 10313 3298 3694 5699 7414 2533 2807 3263 4114 2033 2496 2968 4065 1714 2262 2571 2783 3427 1561 1948 2341 2728 3121 1366 1704

4845 6890 8934 10856 3472 3888 5999 7804 2667 2955 3435 4331 2140 2627 3124 4279 1804 2381 2706 2930 3608 1643 2050 2464 2872 3286 1438 1794

GAS PIPELINE HYDRAULICS

14.00

Wall Thickness in.

2785_C006.fm Page 218 Friday, April 1, 2005 3:29 PM

218

Table 6.6 Pipeline Internal Design Pressures and Test Pressures

24.00

26.00

28.00

30.00

Copyright 2005 by Taylor & Francis Group, LLC

1294 1508 1725 767 957 1150 1340 1533 861 1035 1206 1380 1551 941 1255 1568 1882 863 1005 1150 1293 1438 1725 796 1062 1327 1592 739 986 1232 1479 690

1078 1256 1438 639 797 958 1117 1278 718 863 1005 1150 1293 784 1045 1307 1568 719 838 958 1077 1198 1438 663 885 1106 1327 616 821 1027 1232 575

863 1005 1150 511 638 767 893 1022 574 690 804 920 1034 627 836 1045 1255 575 670 767 862 958 1150 531 708 885 1062 493 657 821 986 460

1941 2261 2588 1150 1435 1725 2010 2300 1292 1553 1809 2070 2327 1411 1882 2352 2823 1294 1508 1725 1939 2156 2588 1194 1592 1990 2388 1109 1479 1848 2218 1035

2048 2387 2731 1214 1515 1821 2122 2428 1363 1639 1910 2185 2456 1490 1986 2483 2980 1366 1591 1821 2047 2276 2731 1261 1681 2101 2521 1171 1561 1951 2341 1093

2156 2513 2875 1278 1595 1917 2234 2556 1435 1725 2010 2300 2585 1568 2091 2614 3136 1438 1675 1917 2154 2396 2875 1327 1769 2212 2654 1232 1643 2054 2464 1150 (continued )

2785_C006.fm Page 219 Friday, April 1, 2005 3:29 PM

22.00

1553 1809 2070 920 1148 1380 1608 1840 1033 1242 1447 1656 1861 1129 1505 1882 2258 1035 1206 1380 1551 1725 2070 955 1274 1592 1911 887 1183 1479 1774 828

219

20.00

62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25 110.64 146.85 182.73 218.27 118.65

PIPE ANALYSIS

18.00

0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375

Pipe Material API 5L X46 Diameter in.

32.00

34.00

36.00

42.00

Weight lb/ft

0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

46000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 1104 1380 1656 776 1035 1294 1553 731 974 1218 1461 690 920 1150 1380 591 789 986 1183 1577

920 1150 1380 647 863 1078 1294 609 812 1015 1218 575 767 958 1150 493 657 821 986 1314

767 958 1150 539 719 898 1078 507 676 846 1015 479 639 799 958 411 548 685 821 1095

613 767 920 431 575 719 863 406 541 676 812 383 511 639 767 329 438 548 657 876

1380 1725 2070 970 1294 1617 1941 913 1218 1522 1826 863 1150 1438 1725 739 986 1232 1479 1971

1457 1821 2185 1024 1366 1707 2048 964 1285 1607 1928 910 1214 1517 1821 780 1040 1301 1561 2081

1533 1917 2300 1078 1438 1797 2156 1015 1353 1691 2029 958 1278 1597 1917 821 1095 1369 1643 2190

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

SMYS

Wall Thickness in.

2785_C006.fm Page 220 Friday, April 1, 2005 3:29 PM

220

Table 6.6 Pipeline Internal Design Pressures and Test Pressures (Continued)

4.5

6.625

8.625

10.75

12.75

14.00

Wall Thickness in. 0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09

52000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 3944 5608 7272 8836 2826 3165 4883 6352 2170 2405 2796 3525 1741 2138 2542 3483 1468 1938 2202 2384 2936 1337 1669 2006 2337 2674

3286 4673 6060 7363 2355 2637 4069 5293 1809 2004 2330 2937 1451 1782 2119 2902 1224 1615 1835 1987 2447 1114 1391 1671 1948 2229

2739 3894 5050 6136 1962 2198 3391 4411 1507 1670 1941 2448 1209 1485 1766 2419 1020 1346 1529 1656 2039 929 1159 1393 1623 1857

2191 3115 4040 4909 1570 1758 2713 3529 1206 1336 1553 1958 967 1188 1412 1935 816 1077 1224 1325 1631 743 927 1114 1299 1486

4930 7010 9090 11045 3532 3956 6103 7940 2713 3006 3494 4406 2177 2673 3178 4353 1835 2423 2753 2981 3671 1671 2086 2507 2922 3343

5203 7399 9595 11658 3728 4176 6443 8381 2864 3173 3689 4651 2298 2822 3355 4595 1937 2557 2906 3146 3875 1764 2202 2646 3084 3529

5477 7788 10100 12272 3925 4395 6782 8822 3014 3340 3883 4896 2419 2970 3531 4837 2039 2692 3059 3312 4078 1857 2318 2786 3246 3714

Copyright 2005 by Taylor & Francis Group, LLC

221

(continued )

2785_C006.fm Page 221 Friday, April 1, 2005 3:29 PM

Pipe Material API 5L X52 Diameter in.

PIPE ANALYSIS

Table 6.7 Pipeline Internal Design Pressures and Test Pressures

Pipe Material API 5L X52 Diameter in. 16.00

18.00

20.00

22.00

26.00

Copyright 2005 by Taylor & Francis Group, LLC

Weight lb/ft

0.250 0.312 0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750

42.05 52.27 62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25

52000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 1170 1460 1755 2045 2340 1040 1298 1560 1818 2080 1168 1404 1636 1872 2104 1276 1702 2127 2553 1170 1363 1560 1753 1950 2340 1080 1440 1800 2160

975 1217 1463 1704 1950 867 1082 1300 1515 1733 973 1170 1363 1560 1753 1064 1418 1773 2127 975 1136 1300 1461 1625 1950 900 1200 1500 1800

813 1014 1219 1420 1625 722 901 1083 1262 1444 811 975 1136 1300 1461 886 1182 1477 1773 813 947 1083 1218 1354 1625 750 1000 1250 1500

650 811 975 1136 1300 578 721 867 1010 1156 649 780 909 1040 1169 709 945 1182 1418 650 757 867 974 1083 1300 600 800 1000 1200

1463 1825 2194 2556 2925 1300 1622 1950 2272 2600 1460 1755 2045 2340 2630 1595 2127 2659 3191 1463 1704 1950 2192 2438 2925 1350 1800 2250 2700

1544 1927 2316 2698 3088 1372 1713 2058 2399 2744 1541 1853 2159 2470 2776 1684 2245 2807 3368 1544 1799 2058 2314 2573 3088 1425 1900 2375 2850

1625 2028 2438 2841 3250 1444 1803 2167 2525 2889 1622 1950 2272 2600 2922 1773 2364 2955 3545 1625 1894 2167 2435 2708 3250 1500 2000 2500 3000

GAS PIPELINE HYDRAULICS

24.00

SMYS

Wall Thickness in.

2785_C006.fm Page 222 Friday, April 1, 2005 3:29 PM

222

Table 6.7 Pipeline Internal Design Pressures and Test Pressures (Continued)

32.00

34.00

36.00

42.00

110.64 146.85 182.73 218.27 118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

1003 1337 1671 2006 936 1248 1560 1872 878 1170 1463 1755 826 1101 1376 1652 780 1040 1300 1560 669 891 1114 1337 1783

836 1114 1393 1671 780 1040 1300 1560 731 975 1219 1463 688 918 1147 1376 650 867 1083 1300 557 743 929 1114 1486

696 929 1161 1393 650 867 1083 1300 609 813 1016 1219 574 765 956 1147 542 722 903 1083 464 619 774 929 1238

557 743 929 1114 520 693 867 1040 488 650 813 975 459 612 765 918 433 578 722 867 371 495 619 743 990

1254 1671 2089 2507 1170 1560 1950 2340 1097 1463 1828 2194 1032 1376 1721 2065 975 1300 1625 1950 836 1114 1393 1671 2229

1323 1764 2205 2646 1235 1647 2058 2470 1158 1544 1930 2316 1090 1453 1816 2179 1029 1372 1715 2058 882 1176 1470 1764 2352

1393 1857 2321 2786 1300 1733 2167 2600 1219 1625 2031 2438 1147 1529 1912 2294 1083 1444 1806 2167 929 1238 1548 1857 2476

223

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 223 Friday, April 1, 2005 3:29 PM

30.00

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

PIPE ANALYSIS

28.00

Pipe Material API 5L X56 Diameter in. 4.5

6.625

8.625

10.75

12.75

16.00

Copyright 2005 by Taylor & Francis Group, LLC

0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500 0.250 0.312

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09 42.05 52.27

56000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 4247 6039 7831 9516 3043 3408 5258 6841 2337 2590 3011 3796 1875 2303 2738 3751 1581 2087 2372 2568 3162 1440 1797 2160 2517 2880 1260 1572

3539 5033 6526 7930 2536 2840 4382 5701 1948 2158 2509 3163 1563 1919 2282 3126 1318 1739 1976 2140 2635 1200 1498 1800 2098 2400 1050 1310

2949 4194 5438 6608 2113 2367 3652 4750 1623 1798 2091 2636 1302 1599 1901 2605 1098 1449 1647 1783 2196 1000 1248 1500 1748 2000 875 1092

2359 3355 4351 5286 1691 1893 2921 3800 1299 1439 1673 2109 1042 1279 1521 2084 878 1160 1318 1427 1757 800 998 1200 1398 1600 700 874

5309 7549 9789 11894 3804 4260 6573 8551 2922 3237 3763 4745 2344 2879 3423 4688 1976 2609 2965 3210 3953 1800 2246 2700 3146 3600 1575 1966

5604 7968 10333 12555 4015 4497 6938 9026 3084 3417 3972 5009 2474 3039 3613 4949 2086 2754 3129 3388 4173 1900 2371 2850 3321 3800 1663 2075

5899 8388 10876 13216 4226 4734 7303 9501 3246 3597 4181 5272 2605 3199 3803 5209 2196 2899 3294 3566 4392 2000 2496 3000 3496 4000 1750 2184

GAS PIPELINE HYDRAULICS

14.00

Wall Thickness in.

2785_C006.fm Page 224 Friday, April 1, 2005 3:29 PM

224

Table 6.8 Pipeline Internal Design Pressures and Test Pressures

22.00

24.00

26.00

28.00

1890 2202 2520 1120 1398 1680 1958 2240 1258 1512 1762 2016 2266 1375 1833 2291 2749 1260 1468 1680 1888 2100 2520 1163 1551 1938 2326 1080 1440 1800 2160

1575 1835 2100 933 1165 1400 1631 1867 1048 1260 1468 1680 1888 1145 1527 1909 2291 1050 1224 1400 1574 1750 2100 969 1292 1615 1938 900 1200 1500 1800

1313 1530 1750 778 971 1167 1360 1556 874 1050 1224 1400 1574 955 1273 1591 1909 875 1020 1167 1311 1458 1750 808 1077 1346 1615 750 1000 1250 1500

1050 1224 1400 622 777 933 1088 1244 699 840 979 1120 1259 764 1018 1273 1527 700 816 933 1049 1167 1400 646 862 1077 1292 600 800 1000 1200

2363 2753 3150 1400 1747 2100 2447 2800 1572 1890 2202 2520 2832 1718 2291 2864 3436 1575 1835 2100 2360 2625 3150 1454 1938 2423 2908 1350 1800 2250 2700

2494 2906 3325 1478 1844 2217 2583 2956 1660 1995 2325 2660 2990 1814 2418 3023 3627 1663 1937 2217 2492 2771 3325 1535 2046 2558 3069 1425 1900 2375 2850

2625 3059 3500 1556 1941 2333 2719 3111 1747 2100 2447 2800 3147 1909 2545 3182 3818 1750 2039 2333 2623 2917 3500 1615 2154 2692 3231 1500 2000 2500 3000

Copyright 2005 by Taylor & Francis Group, LLC

225

(continued )

2785_C006.fm Page 225 Friday, April 1, 2005 3:29 PM

20.00

62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25 110.64 146.85 182.73 218.27

PIPE ANALYSIS

18.00

0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750

Pipe Material API 5L X56 Diameter in. 30.00

32.00

34.00

36.00

42.00

Weight lb/ft

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

56000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 1008 1344 1680 2016 945 1260 1575 1890 889 1186 1482 1779 840 1120 1400 1680 720 960 1200 1440 1920

840 1120 1400 1680 788 1050 1313 1575 741 988 1235 1482 700 933 1167 1400 600 800 1000 1200 1600

700 933 1167 1400 656 875 1094 1313 618 824 1029 1235 583 778 972 1167 500 667 833 1000 1333

560 747 933 1120 525 700 875 1050 494 659 824 988 467 622 778 933 400 533 667 800 1067

1260 1680 2100 2520 1181 1575 1969 2363 1112 1482 1853 2224 1050 1400 1750 2100 900 1200 1500 1800 2400

1330 1773 2217 2660 1247 1663 2078 2494 1174 1565 1956 2347 1108 1478 1847 2217 950 1267 1583 1900 2533

1400 1867 2333 2800 1313 1750 2188 2625 1235 1647 2059 2471 1167 1556 1944 2333 1000 1333 1667 2000 2667

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

SMYS

Wall Thickness in.

2785_C006.fm Page 226 Friday, April 1, 2005 3:29 PM

226

Table 6.8 Pipeline Internal Design Pressures and Test Pressures (Continued)

4.5

6.625

8.625

10.75

12.75

14.00

Wall Thickness in. 0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09

60000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 4550 6470 8390 10195 3260 3652 5634 7329 2504 2775 3226 4067 2009 2467 2934 4019 1694 2236 2541 2751 3388 1543 1925 2314 2697 3086

3792 5392 6992 8496 2717 3043 4695 6108 2087 2312 2688 3389 1674 2056 2445 3349 1412 1864 2118 2293 2824 1286 1605 1929 2247 2571

3160 4493 5827 7080 2264 2536 3912 5090 1739 1927 2240 2824 1395 1713 2037 2791 1176 1553 1765 1911 2353 1071 1337 1607 1873 2143

2528 3595 4661 5664 1811 2029 3130 4072 1391 1542 1792 2259 1116 1371 1630 2233 941 1242 1412 1528 1882 857 1070 1286 1498 1714

5688 8088 10488 12744 4075 4565 7042 9162 3130 3469 4032 5084 2512 3084 3667 5023 2118 2795 3176 3439 4235 1929 2407 2893 3371 3857

6004 8537 11071 13452 4302 4818 7434 9671 3304 3661 4256 5366 2651 3256 3871 5302 2235 2951 3353 3630 4471 2036 2541 3054 3558 4071

6320 8987 11653 14160 4528 5072 7825 10180 3478 3854 4480 5649 2791 3427 4074 5581 2353 3106 3529 3821 4706 2143 2674 3214 3746 4286

Copyright 2005 by Taylor & Francis Group, LLC

227

(continued )

2785_C006.fm Page 227 Friday, April 1, 2005 3:29 PM

Pipe Material API 5L X60 Diameter in.

PIPE ANALYSIS

Table 6.9 Pipeline Internal Design Pressures and Test Pressures

Pipe Material API 5L X60 Diameter in. 16.00

18.00

20.00

22.00

26.00

Copyright 2005 by Taylor & Francis Group, LLC

Weight lb/ft

0.250 0.312 0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750

42.05 52.27 62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25

60000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 1350 1685 2025 2360 2700 1200 1498 1800 2098 2400 1348 1620 1888 2160 2428 1473 1964 2455 2945 1350 1573 1800 2023 2250 2700 1246 1662 2077 2492

1125 1404 1688 1967 2250 1000 1248 1500 1748 2000 1123 1350 1573 1800 2023 1227 1636 2045 2455 1125 1311 1500 1686 1875 2250 1038 1385 1731 2077

938 1170 1406 1639 1875 833 1040 1250 1457 1667 936 1125 1311 1500 1686 1023 1364 1705 2045 938 1093 1250 1405 1563 1875 865 1154 1442 1731

750 936 1125 1311 1500 667 832 1000 1165 1333 749 900 1049 1200 1349 818 1091 1364 1636 750 874 1000 1124 1250 1500 692 923 1154 1385

1688 2106 2531 2950 3375 1500 1872 2250 2622 3000 1685 2025 2360 2700 3035 1841 2455 3068 3682 1688 1967 2250 2529 2813 3375 1558 2077 2596 3115

1781 2223 2672 3114 3563 1583 1976 2375 2768 3167 1778 2138 2491 2850 3203 1943 2591 3239 3886 1781 2076 2375 2670 2969 3563 1644 2192 2740 3288

1875 2340 2813 3278 3750 1667 2080 2500 2913 3333 1872 2250 2622 3000 3372 2045 2727 3409 4091 1875 2185 2500 2810 3125 3750 1731 2308 2885 3462

GAS PIPELINE HYDRAULICS

24.00

SMYS

Wall Thickness in.

2785_C006.fm Page 228 Friday, April 1, 2005 3:29 PM

228

Table 6.9 Pipeline Internal Design Pressures and Test Pressures (Continued)

32.00

34.00

36.00

42.00

110.64 146.85 182.73 218.27 118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

1157 1543 1929 2314 1080 1440 1800 2160 1013 1350 1688 2025 953 1271 1588 1906 900 1200 1500 1800 771 1029 1286 1543 2057

964 1286 1607 1929 900 1200 1500 1800 844 1125 1406 1688 794 1059 1324 1588 750 1000 1250 1500 643 857 1071 1286 1714

804 1071 1339 1607 750 1000 1250 1500 703 938 1172 1406 662 882 1103 1324 625 833 1042 1250 536 714 893 1071 1429

643 857 1071 1286 600 800 1000 1200 563 750 938 1125 529 706 882 1059 500 667 833 1000 429 571 714 857 1143

1446 1929 2411 2893 1350 1800 2250 2700 1266 1688 2109 2531 1191 1588 1985 2382 1125 1500 1875 2250 964 1286 1607 1929 2571

1527 2036 2545 3054 1425 1900 2375 2850 1336 1781 2227 2672 1257 1676 2096 2515 1188 1583 1979 2375 1018 1357 1696 2036 2714

1607 2143 2679 3214 1500 2000 2500 3000 1406 1875 2344 2813 1324 1765 2206 2647 1250 1667 2083 2500 1071 1429 1786 2143 2857

229

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 229 Friday, April 1, 2005 3:29 PM

30.00

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

PIPE ANALYSIS

28.00

4.5

6.625

8.625

10.75

12.75

16.00

Copyright 2005 by Taylor & Francis Group, LLC

0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500 0.250 0.312

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09 42.05 52.27

65000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 4930 7010 9090 11045 3532 3956 6103 7940 2713 3006 3494 4406 2177 2673 3178 4353 1835 2423 2753 2981 3671 1671 2086 2507 2922 3343 1463 1825

4108 5841 7575 9204 2943 3297 5086 6617 2261 2505 2912 3672 1814 2228 2648 3628 1529 2019 2294 2484 3059 1393 1738 2089 2435 2786 1219 1521

3423 4868 6312 7670 2453 2747 4238 5514 1884 2088 2427 3060 1512 1856 2207 3023 1275 1682 1912 2070 2549 1161 1449 1741 2029 2321 1016 1268

2739 3894 5050 6136 1962 2198 3391 4411 1507 1670 1941 2448 1209 1485 1766 2419 1020 1346 1529 1656 2039 929 1159 1393 1623 1857 813 1014

6162 8762 11362 13806 4415 4945 7629 9925 3391 3758 4368 5507 2721 3341 3973 5442 2294 3028 3441 3726 4588 2089 2607 3134 3652 4179 1828 2282

6504 9249 11993 14573 4660 5220 8053 10477 3580 3966 4611 5813 2872 3527 4193 5744 2422 3196 3632 3933 4843 2205 2752 3308 3855 4411 1930 2408

6847 9736 12624 15340 4906 5494 8477 11028 3768 4175 4853 6119 3023 3713 4414 6047 2549 3365 3824 4140 5098 2321 2897 3482 4058 4643 2031 2535

GAS PIPELINE HYDRAULICS

14.00

Wall Thickness in.

2785_C006.fm Page 230 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures

Pipe Material API 5L X65 Diameter in.

230

Table 6.10

22.00

24.00

26.00

28.00

2194 2556 2925 1300 1622 1950 2272 2600 1460 1755 2045 2340 2630 1595 2127 2659 3191 1463 1704 1950 2192 2438 2925 1350 1800 2250 2700 1254 1671 2089 2507

1828 2130 2438 1083 1352 1625 1894 2167 1217 1463 1704 1950 2192 1330 1773 2216 2659 1219 1420 1625 1827 2031 2438 1125 1500 1875 2250 1045 1393 1741 2089

1523 1775 2031 903 1127 1354 1578 1806 1014 1219 1420 1625 1827 1108 1477 1847 2216 1016 1184 1354 1522 1693 2031 938 1250 1563 1875 871 1161 1451 1741

1219 1420 1625 722 901 1083 1262 1444 811 975 1136 1300 1461 886 1182 1477 1773 813 947 1083 1218 1354 1625 750 1000 1250 1500 696 929 1161 1393

2742 3196 3656 1625 2028 2438 2841 3250 1825 2194 2556 2925 3288 1994 2659 3324 3989 1828 2130 2438 2740 3047 3656 1688 2250 2813 3375 1567 2089 2612 3134

2895 3373 3859 1715 2141 2573 2998 3431 1927 2316 2698 3088 3470 2105 2807 3509 4210 1930 2249 2573 2892 3216 3859 1781 2375 2969 3563 1654 2205 2757 3308

3047 3551 4063 1806 2253 2708 3156 3611 2028 2438 2841 3250 3653 2216 2955 3693 4432 2031 2367 2708 3044 3385 4063 1875 2500 3125 3750 1741 2321 2902 3482

Copyright 2005 by Taylor & Francis Group, LLC

231

(continued )

2785_C006.fm Page 231 Friday, April 1, 2005 3:29 PM

20.00

62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25 110.64 146.85 182.73 218.27

PIPE ANALYSIS

18.00

0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750

30.00

32.00

34.00

36.00

42.00

Weight lb/ft

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

65000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 1170 1560 1950 2340 1097 1463 1828 2194 1032 1376 1721 2065 975 1300 1625 1950 836 1114 1393 1671 2229

975 1300 1625 1950 914 1219 1523 1828 860 1147 1434 1721 813 1083 1354 1625 696 929 1161 1393 1857

813 1083 1354 1625 762 1016 1270 1523 717 956 1195 1434 677 903 1128 1354 580 774 967 1161 1548

650 867 1083 1300 609 813 1016 1219 574 765 956 1147 542 722 903 1083 464 619 774 929 1238

1463 1950 2438 2925 1371 1828 2285 2742 1290 1721 2151 2581 1219 1625 2031 2438 1045 1393 1741 2089 2786

1544 2058 2573 3088 1447 1930 2412 2895 1362 1816 2270 2724 1286 1715 2144 2573 1103 1470 1838 2205 2940

1625 2167 2708 3250 1523 2031 2539 3047 1434 1912 2390 2868 1354 1806 2257 2708 1161 1548 1935 2321 3095

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

SMYS

Wall Thickness in.

2785_C006.fm Page 232 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures (Continued)

Pipe Material API 5L X65 Diameter in.

232

Table 6.10

Pipe Material API 5L X70 Diameter in. 4.5

6.625

8.625

10.75

12.75

14.00

Wall Thickness in. 0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09

70000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 5309 7549 9789 11894 3804 4260 6573 8551 2922 3237 3763 4745 2344 2879 3423 4688 1976 2609 2965 3210 3953 1800 2246 2700 3146 3600

4424 6291 8157 9912 3170 3550 5477 7126 2435 2698 3136 3954 1953 2399 2852 3907 1647 2174 2471 2675 3294 1500 1872 2250 2622 3000

3687 5242 6798 8260 2642 2958 4565 5938 2029 2248 2613 3295 1628 1999 2377 3256 1373 1812 2059 2229 2745 1250 1560 1875 2185 2500

2949 4194 5438 6608 2113 2367 3652 4750 1623 1798 2091 2636 1302 1599 1901 2605 1098 1449 1647 1783 2196 1000 1248 1500 1748 2000

6636 9436 12236 14868 4755 5325 8216 10689 3652 4047 4704 5931 2930 3598 4278 5860 2471 3261 3706 4012 4941 2250 2808 3375 3933 4500

7005 9960 12916 15694 5019 5621 8673 11282 3855 4271 4965 6261 3093 3798 4516 6186 2608 3442 3912 4235 5216 2375 2964 3563 4152 4750

7373 10484 13596 16520 5283 5917 9129 11876 4058 4496 5227 6590 3256 3998 4753 6512 2745 3624 4118 4458 5490 2500 3120 3750 4370 5000

Copyright 2005 by Taylor & Francis Group, LLC

233

(continued )

2785_C006.fm Page 233 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures

PIPE ANALYSIS

Table 6.11

16.00

18.00

20.00

22.00

26.00

Copyright 2005 by Taylor & Francis Group, LLC

Weight lb/ft

0.250 0.312 0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750

42.05 52.27 62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25

70000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 1575 1966 2363 2753 3150 1400 1747 2100 2447 2800 1572 1890 2202 2520 2832 1718 2291 2864 3436 1575 1835 2100 2360 2625 3150 1454 1938 2423 2908

1313 1638 1969 2294 2625 1167 1456 1750 2039 2333 1310 1575 1835 2100 2360 1432 1909 2386 2864 1313 1530 1750 1967 2188 2625 1212 1615 2019 2423

1094 1365 1641 1912 2188 972 1213 1458 1699 1944 1092 1313 1530 1750 1967 1193 1591 1989 2386 1094 1275 1458 1639 1823 2188 1010 1346 1683 2019

875 1092 1313 1530 1750 778 971 1167 1360 1556 874 1050 1224 1400 1574 955 1273 1591 1909 875 1020 1167 1311 1458 1750 808 1077 1346 1615

1969 2457 2953 3441 3938 1750 2184 2625 3059 3500 1966 2363 2753 3150 3541 2148 2864 3580 4295 1969 2294 2625 2951 3281 3938 1817 2423 3029 3635

2078 2594 3117 3633 4156 1847 2305 2771 3229 3694 2075 2494 2906 3325 3737 2267 3023 3778 4534 2078 2422 2771 3114 3464 4156 1918 2558 3197 3837

2188 2730 3281 3824 4375 1944 2427 2917 3399 3889 2184 2625 3059 3500 3934 2386 3182 3977 4773 2188 2549 2917 3278 3646 4375 2019 2692 3365 4038

GAS PIPELINE HYDRAULICS

24.00

SMYS

Wall Thickness in.

2785_C006.fm Page 234 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures (Continued)

Pipe Material API 5L X70 Diameter in.

234

Table 6.11

32.00

34.00

36.00

42.00

110.64 146.85 182.73 218.27 118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

1350 1800 2250 2700 1260 1680 2100 2520 1181 1575 1969 2363 1112 1482 1853 2224 1050 1400 1750 2100 900 1200 1500 1800 2400

1125 1500 1875 2250 1050 1400 1750 2100 984 1313 1641 1969 926 1235 1544 1853 875 1167 1458 1750 750 1000 1250 1500 2000

938 1250 1563 1875 875 1167 1458 1750 820 1094 1367 1641 772 1029 1287 1544 729 972 1215 1458 625 833 1042 1250 1667

750 1000 1250 1500 700 933 1167 1400 656 875 1094 1313 618 824 1029 1235 583 778 972 1167 500 667 833 1000 1333

1688 2250 2813 3375 1575 2100 2625 3150 1477 1969 2461 2953 1390 1853 2316 2779 1313 1750 2188 2625 1125 1500 1875 2250 3000

1781 2375 2969 3563 1663 2217 2771 3325 1559 2078 2598 3117 1467 1956 2445 2934 1385 1847 2309 2771 1188 1583 1979 2375 3167

1875 2500 3125 3750 1750 2333 2917 3500 1641 2188 2734 3281 1544 2059 2574 3088 1458 1944 2431 2917 1250 1667 2083 2500 3333

235

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 235 Friday, April 1, 2005 3:29 PM

30.00

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

PIPE ANALYSIS

28.00

4.5

6.625

8.625

10.75

12.75

16.00

Copyright 2005 by Taylor & Francis Group, LLC

0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500 0.250 0.312

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09 42.05 52.27

80000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 6067 8627 11187 13594 4347 4869 7512 9772 3339 3700 4301 5423 2679 3290 3911 5358 2259 2982 3388 3668 4518 2057 2567 3086 3596 4114 1800 2246

5056 7189 9323 11328 3623 4057 6260 8144 2783 3083 3584 4519 2233 2742 3260 4465 1882 2485 2824 3057 3765 1714 2139 2571 2997 3429 1500 1872

4213 5991 7769 9440 3019 3381 5217 6786 2319 2569 2987 3766 1860 2285 2716 3721 1569 2071 2353 2547 3137 1429 1783 2143 2497 2857 1250 1560

3371 4793 6215 7552 2415 2705 4173 5429 1855 2055 2389 3013 1488 1828 2173 2977 1255 1656 1882 2038 2510 1143 1426 1714 1998 2286 1000 1248

7584 10784 13984 16992 5434 6086 9390 12216 4174 4625 5376 6778 3349 4112 4889 6698 2824 3727 4235 4585 5647 2571 3209 3857 4495 5143 2250 2808

8005 11383 14761 17936 5736 6424 9912 12894 4406 4882 5675 7155 3535 4341 5161 7070 2980 3934 4471 4840 5961 2714 3387 4071 4745 5429 2375 2964

8427 11982 15538 18880 6038 6762 10433 13573 4638 5139 5973 7532 3721 4569 5433 7442 3137 4141 4706 5095 6275 2857 3566 4286 4994 5714 2500 3120

GAS PIPELINE HYDRAULICS

14.00

Wall Thickness in.

2785_C006.fm Page 236 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures

Pipe Material API 5L X80 Diameter in.

236

Table 6.12

22.00

24.00

26.00

28.00

2700 3146 3600 1600 1997 2400 2797 3200 1797 2160 2517 2880 3237 1964 2618 3273 3927 1800 2098 2400 2698 3000 3600 1662 2215 2769 3323 1543 2057 2571 3086

2250 2622 3000 1333 1664 2000 2331 2667 1498 1800 2098 2400 2698 1636 2182 2727 3273 1500 1748 2000 2248 2500 3000 1385 1846 2308 2769 1286 1714 2143 2571

1875 2185 2500 1111 1387 1667 1942 2222 1248 1500 1748 2000 2248 1364 1818 2273 2727 1250 1457 1667 1873 2083 2500 1154 1538 1923 2308 1071 1429 1786 2143

1500 1748 2000 889 1109 1333 1554 1778 998 1200 1398 1600 1798 1091 1455 1818 2182 1000 1165 1333 1499 1667 2000 923 1231 1538 1846 857 1143 1429 1714

3375 3933 4500 2000 2496 3000 3496 4000 2246 2700 3146 3600 4046 2455 3273 4091 4909 2250 2622 3000 3372 3750 4500 2077 2769 3462 4154 1929 2571 3214 3857

3563 4152 4750 2111 2635 3167 3690 4222 2371 2850 3321 3800 4271 2591 3455 4318 5182 2375 2768 3167 3559 3958 4750 2192 2923 3654 4385 2036 2714 3393 4071

3750 4370 5000 2222 2773 3333 3884 4444 2496 3000 3496 4000 4496 2727 3636 4545 5455 2500 2913 3333 3747 4167 5000 2308 3077 3846 4615 2143 2857 3571 4286

Copyright 2005 by Taylor & Francis Group, LLC

237

(continued )

2785_C006.fm Page 237 Friday, April 1, 2005 3:29 PM

20.00

62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25 110.64 146.85 182.73 218.27

PIPE ANALYSIS

18.00

0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750

30.00

32.00

34.00

36.00

42.00

Weight lb/ft

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

80000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 1440 1920 2400 2880 1350 1800 2250 2700 1271 1694 2118 2541 1200 1600 2000 2400 1029 1371 1714 2057 2743

1200 1600 2000 2400 1125 1500 1875 2250 1059 1412 1765 2118 1000 1333 1667 2000 857 1143 1429 1714 2286

1000 1333 1667 2000 938 1250 1563 1875 882 1176 1471 1765 833 1111 1389 1667 714 952 1190 1429 1905

800 1067 1333 1600 750 1000 1250 1500 706 941 1176 1412 667 889 1111 1333 571 762 952 1143 1524

1800 2400 3000 3600 1688 2250 2813 3375 1588 2118 2647 3176 1500 2000 2500 3000 1286 1714 2143 2571 3429

1900 2533 3167 3800 1781 2375 2969 3563 1676 2235 2794 3353 1583 2111 2639 3167 1357 1810 2262 2714 3619

2000 2667 3333 4000 1875 2500 3125 3750 1765 2353 2941 3529 1667 2222 2778 3333 1429 1905 2381 2857 3810

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

SMYS

Wall Thickness in.

2785_C006.fm Page 238 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures (Continued)

Pipe Material API 5L X80 Diameter in.

238

Table 6.12

Pipe Material API 5L X90 Diameter in. 4.5

6.625

8.625

10.75

12.75

14.00

Wall Thickness in. 0.237 0.337 0.437 0.531 0.250 0.280 0.432 0.562 0.250 0.277 0.322 0.406 0.250 0.307 0.365 0.500 0.250 0.330 0.375 0.406 0.500 0.250 0.312 0.375 0.437 0.500

SMYS Weight lb/ft 10.79 14.98 18.96 22.51 17.02 18.97 28.57 36.39 22.36 24.70 28.55 35.64 28.04 34.24 40.48 54.74 33.38 43.77 49.56 53.52 65.42 36.71 45.61 54.57 63.30 72.09

90000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 6826 9706 12586 15293 4891 5477 8451 10994 3757 4162 4838 6101 3014 3701 4400 6028 2541 3354 3812 4127 5082 2314 2888 3471 4045 4629

5688 8088 10488 12744 4075 4565 7042 9162 3130 3469 4032 5084 2512 3084 3667 5023 2118 2795 3176 3439 4235 1929 2407 2893 3371 3857

4740 6740 8740 10620 3396 3804 5869 7635 2609 2890 3360 4237 2093 2570 3056 4186 1765 2329 2647 2866 3529 1607 2006 2411 2809 3214

3792 5392 6992 8496 2717 3043 4695 6108 2087 2312 2688 3389 1674 2056 2445 3349 1412 1864 2118 2293 2824 1286 1605 1929 2247 2571

8532 12132 15732 19116 6113 6847 10564 13742 4696 5203 6048 7626 3767 4626 5500 7535 3176 4193 4765 5159 6353 2893 3610 4339 5057 5786

9006 12806 16606 20178 6453 7227 11150 14506 4957 5492 6384 8049 3977 4883 5806 7953 3353 4426 5029 5445 6706 3054 3811 4580 5338 6107

9480 13480 17480 21240 6792 7608 11737 15269 5217 5781 6720 8473 4186 5140 6112 8372 3529 4659 5294 5732 7059 3214 4011 4821 5619 6429

Copyright 2005 by Taylor & Francis Group, LLC

239

(continued )

2785_C006.fm Page 239 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures

PIPE ANALYSIS

Table 6.13

16.00

18.00

20.00

22.00

26.00

Copyright 2005 by Taylor & Francis Group, LLC

Weight lb/ft

0.250 0.312 0.375 0.437 0.500 0.250 0.312 0.375 0.437 0.500 0.312 0.375 0.437 0.500 0.562 0.375 0.500 0.625 0.750 0.375 0.437 0.500 0.562 0.625 0.750 0.375 0.500 0.625 0.750

42.05 52.27 62.58 72.64 82.77 47.39 58.94 70.59 81.97 93.45 65.60 78.60 91.30 104.13 116.67 86.61 114.81 142.68 170.21 94.62 109.97 125.49 140.68 156.03 186.23 102.63 136.17 169.38 202.25

90000 psig Internal Design Pressure, psig Hydrostatic Test Pressure, psig Class 1 Class 2 Class 3 Class 4 90% SMYS 95% SMYS 100% SMYS 2025 2527 3038 3540 4050 1800 2246 2700 3146 3600 2022 2430 2832 3240 3642 2209 2945 3682 4418 2025 2360 2700 3035 3375 4050 1869 2492 3115 3738

1688 2106 2531 2950 3375 1500 1872 2250 2622 3000 1685 2025 2360 2700 3035 1841 2455 3068 3682 1688 1967 2250 2529 2813 3375 1558 2077 2596 3115

1406 1755 2109 2458 2813 1250 1560 1875 2185 2500 1404 1688 1967 2250 2529 1534 2045 2557 3068 1406 1639 1875 2108 2344 2813 1298 1731 2163 2596

1125 1404 1688 1967 2250 1000 1248 1500 1748 2000 1123 1350 1573 1800 2023 1227 1636 2045 2455 1125 1311 1500 1686 1875 2250 1038 1385 1731 2077

2531 3159 3797 4425 5063 2250 2808 3375 3933 4500 2527 3038 3540 4050 4552 2761 3682 4602 5523 2531 2950 3375 3794 4219 5063 2337 3115 3894 4673

2672 3335 4008 4670 5344 2375 2964 3563 4152 4750 2668 3206 3736 4275 4805 2915 3886 4858 5830 2672 3114 3563 4004 4453 5344 2466 3288 4111 4933

2813 3510 4219 4916 5625 2500 3120 3750 4370 5000 2808 3375 3933 4500 5058 3068 4091 5114 6136 2813 3278 3750 4215 4688 5625 2596 3462 4327 5192

GAS PIPELINE HYDRAULICS

24.00

SMYS

Wall Thickness in.

2785_C006.fm Page 240 Friday, April 1, 2005 3:29 PM

Pipeline Internal Design Pressures and Test Pressures (Continued)

Pipe Material API 5L X90 Diameter in.

240

Table 6.13

32.00

34.00

36.00

42.00

110.64 146.85 182.73 218.27 118.65 157.53 196.08 234.29 126.66 168.21 209.43 250.31 134.67 178.89 222.78 266.33 142.68 189.57 236.13 282.35 166.71 221.61 276.18 330.41 437.88

1736 2314 2893 3471 1620 2160 2700 3240 1519 2025 2531 3038 1429 1906 2382 2859 1350 1800 2250 2700 1157 1543 1929 2314 3086

1446 1929 2411 2893 1350 1800 2250 2700 1266 1688 2109 2531 1191 1588 1985 2382 1125 1500 1875 2250 964 1286 1607 1929 2571

1205 1607 2009 2411 1125 1500 1875 2250 1055 1406 1758 2109 993 1324 1654 1985 938 1250 1563 1875 804 1071 1339 1607 2143

964 1286 1607 1929 900 1200 1500 1800 844 1125 1406 1688 794 1059 1324 1588 750 1000 1250 1500 643 857 1071 1286 1714

2170 2893 3616 4339 2025 2700 3375 4050 1898 2531 3164 3797 1787 2382 2978 3574 1688 2250 2813 3375 1446 1929 2411 2893 3857

2290 3054 3817 4580 2138 2850 3563 4275 2004 2672 3340 4008 1886 2515 3143 3772 1781 2375 2969 3563 1527 2036 2545 3054 4071

2411 3214 4018 4821 2250 3000 3750 4500 2109 2813 3516 4219 1985 2647 3309 3971 1875 2500 3125 3750 1607 2143 2679 3214 4286

241

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 241 Friday, April 1, 2005 3:29 PM

30.00

0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 0.375 0.500 0.625 0.750 1.000

PIPE ANALYSIS

28.00

2785_C006.fm Page 242 Friday, April 1, 2005 3:29 PM

242

GAS PIPELINE HYDRAULICS

For class 1, the range of hydrotest pressures is (1.25 × 1872) psig to (1.3194 × 1872) psig = 2340 psig to 2470 psig where 1.3194 is equal to the factor 1.25 × 95/90, representing the upper limit of the hydrotest envelope. For class 2, the range of hydrotest pressures is (1.25 × 1560) psig to (1.3194 × 1560) psig = 1950 psig to 2058 psig For class 3, the range of hydrotest pressures is (1.25 × 1300) psig to (1.3194 × 1300) psig = 1625 psig to 1715 psig For class 4, the range of hydrotest pressures is (1.25 × 1040) psig to (1.3194 × 1040) psig = 1300 psig to 1372 psig

6.10 BLOWDOWN CALCULATIONS Blowdown valves and piping systems are installed around the mainline valve in a gas transmission piping system in order to evacuate gas from sections of pipeline in the event of an emergency or for maintenance purposes. The objective of the blowdown assembly is to remove gas from the pipeline once the pipe section is isolated by closing the mainline block valves in a reasonable period of time. The pipe size required to blow down a section of pipe will depend on the gas gravity, pipe diameter, length of pipe section, pressure in the pipeline, and blowdown time. AGA recommends the following equation to estimate the blowdown time: 1

T= where T = P1 = G = D = L = d = Fc =

1

0.0588 P1 3G 2 D 2 LFc d2

blowdown time, min initial pressure, psia gas gravity (air = 1.00) pipe inside diameter, in. length of pipe section, mi inside diameter of blowdown pipe, in. choke factor (as follows)

Choke factor list Ideal nozzle = 1.0 Through gate = 1.6 Regular gate = 1.8 Regular lube plug = 2.0 Venturi lube plug = 3.2

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(6.9)

2785_C006.fm Page 243 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

243

In SI units, 1

1

0.0192 P1 3G 2 D 2 LFc T= d2 where P1 = D = L = d =

(SI units)

(6.10)

initial pressure, kPa pipe inside diameter, mm length of pipe section, km pipe inside diameter of blowdown, mm

Other symbols are as defined before. Example 4 Calculate the blowdown time required for an NPS 6, 0.250 in. wall thickness, blowdown assembly on an NPS 24 pipe, 0.500 in. wall thickness, considering a 5 mi pipe section starting at a pressure of 1000 psia. The gas gravity is 0.6 and choke factor = 1.8. Solution Pipe inside diameter = 24 – 2 × 0.500 = 23 in. Blowdown pipe inside diameter = 6.625 – 2 × 0.250 = 6.125 in. Using Equation 6.9, we get 1

1

0.0588 × (1000) 3 (0.6) 2 (23)2 × 5 × 1.8 T= = 58 min, approximately 6.1252

6.11 DETERMINING PIPE TONNAGE Frequently in pipeline design, we are interested in knowing the amount of pipe used so that we can determine the total cost of pipe. A convenient formula for calculating the weight per unit length of pipe used by pipe vendors is given in Equation 6.11. In USCS units, pipe weight in lb/ft is calculated for a given diameter and wall thickness as follows: w = 10.68 × t × (D – t) where w = pipe weight, lb/ft D = pipe outside diameter, in. t = pipe wall thickness, in.

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(6.11)

2785_C006.fm Page 244 Friday, April 1, 2005 3:29 PM

244

GAS PIPELINE HYDRAULICS

The constant 10.68 in Equation 6.11 includes the density of steel and, therefore, the equation is only applicable to steel pipe. For other pipe material, we can ratio the densities to obtain the pipe weight for nonsteel pipe. In SI units, the pipe weight in kg/m is found from w = 0.0246 × t × (D – t)

(SI units)

(6.12)

where w = pipe weight, kg/m D = pipe outside diameter, mm t = pipe wall thickness, mm Example 5 Calculate the total amount of pipe in a 10 mi pipeline, NPS 20, 0.500 in wall thickness. If pipe costs $700 per ton, determine the total pipeline cost. Solution Using Equation 6.11, the weight per foot of pipe is

w = 10.68 × 0.500 × (20 – 0.500) = 101.46 lb/ft Therefore, the total pipe tonnage in 10 miles of pipe is Tonnage = 101.46 × 5280 × 10/2000 = 2679 tons Total pipeline cost = 2679 × 700 = $1,875,300 Example 6 A 60 km pipeline consists of 20 km of DN 500, 12 mm wall thickness pipe connected to a 40 km length of DN 400, 10 mm wall thickness pipe. What are the total metric tons of pipe? Solution Using Equation 6.12, the weight per meter of DN 500 pipe is

w = 0.0246 × 12 × (500 – 12) = 144.06 kg/m and the weight per meter of DN 400 pipe is

w = 0.0246 × 10 × (400 – 10) = 95.94 kg/m Therefore, the total pipe weight for 20 km of DN 500 pipe and 40 km of DN 400 pipe is Weight = (20 × 144.06) + (40 × 95.94) = 6719 tons Total metric tons = 6719

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 245 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

245

Example 7 Calculate the MOP for NPS 16 pipeline, 0.250 in wall thickness, constructed of API 5LX-52 steel. What minimum wall thickness is required for an internal working pressure of 1440 psi? Use class 2 construction with design factor F = 0.60 and for an operating temperature below 250°F. Solution Using Equation 6.8, the internal design pressure is P=

2 × 0.250 × 52, 000 × 0.60 × 1.0 × 1.0 = 975 psig 16

For an internal working pressure of 1440 psi, the wall thickness required is 1440 =

2 × t × 52, 000 × 0.6 × 1.0 16

Solving for t, we get Wall thickness t = 0.369 in. The nearest standard pipe wall thickness is 0.375 in.

Example 8 A natural gas pipeline, 600 km long, is constructed of DN 800 pipe and has a required operating pressure of 9 MPa. Compare the cost of using X-60 and X-70 steel pipe. The material costs of the two grades of pipe are as follows: Pipe Grade

Material Cost—$/tonne

X-60 X-70

800 900

Use a class 1 design factor and temperature deration factor of 1.00. Solution We will first determine the wall thickness of pipe required to withstand the operating pressure of 9 MPa. Using Equation 6.8, the pipe wall thickness required for X-60 pipe (60,000 psi = 414 MPa) is t=

9 × 800 = 12.08 mm. Use 13 mm wall thickness. 2 × 414 × 1.0 × 0.72 × 1.0

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 246 Friday, April 1, 2005 3:29 PM

246

GAS PIPELINE HYDRAULICS

Similarly, the pipe wall thickness required for X-70 pipe (70,000 psi = 483 MPa) is t=

9 × 800 = 10.35 mm. Use 11 mm wall thickness. 2 × 483 × 1.0 × 0.72 × 1.0

The pipe weight in kg/m will be calculated using Equation 6.12. For X-60 pipe, Weight per meter = 0.0246 × 13 × (800 –13) = 251.68 kg/m Therefore, the total cost of 600 km pipeline at $800 per ton of X-60 pipe is Total cost = 600 × 251.68 × 800 = $120.81 million Similarly, the pipe weight in kg/m for X-70 pipe is Weight per meter = 0.0246 × 11 × (800 – 11) = 213.50 kg/m Therefore, the total cost of 600 km pipeline at $900 per ton of X-70 pipe is Total cost = 600 × 213.50 × 900 = $115.29 million Therefore, the X-70 pipe will cost less than the X-60 pipe. The difference in cost is $120.81 – $115.29 = $5.52 million.

6.12 SUMMARY In this chapter we discussed how to calculate the pipe wall thickness required to withstand an internal pressure in a gas pipeline using Barlow’s equation. The influence of the population density in the vicinity of the pipeline on the required pipe wall thickness by reducing the allowable hoop stress in the high population areas was explained by way of class locations. We explored the range of pressures required to hydrotest pipeline sections to ensure safe operation of the pipeline. The effect of pipeline elevations on determining a testing plan by sectioning the pipeline was covered. The need for isolating portions of the pipeline by properly spaced mainline valves and the method of calculating the time required for evacuating gas from the pipeline sections were also discussed. Finally, a simple method of calculating the pipe tonnage was explained. PROBLEMS 1. A gas pipeline is constructed of API 5L X70 steel, NPS 24, 0.375 in. wall thickness. Calculate the MAOP of this pipeline for a class 1 design factor and a temperature deration factor of 1.00. 2. A gas pipeline, DN 500, 12 mm wall thickness, is constructed of API 5L X65 pipe. (a) Calculate the design pressures for class 1 and class 2 locations. (b) What is the range of hydrotest pressures for each of these class locations? Assume the joint factor = 1.00 and temperature deration factor = 1.00.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 247 Friday, April 1, 2005 3:29 PM

PIPE ANALYSIS

247

3. Calculate the total tonnage of pipe material for a 24 mi pipeline, NPS 16, 0.375 in. wall thickness. If pipe costs $700 per ton, determine the total pipeline cost. 4. A 50 km pipeline consists of 15 km of DN 400, 10 mm wall thickness pipe connected to a 35 km length of DN 300, 8 mm wall thickness pipe. (a) Calculate the total metric tons of pipe. (b) If this pipeline were replaced with a single 50 km long pipeline, DN 400, API 5LX-65 material, what minimum wall thickness would be required for a class 1 design at an MOP of 9 MPa? 5. Calculate the minimum wall thickness for NPS 16 pipeline constructed of API 5LX-60 steel to withstand an internal pressure of 1440 psi. Use a class 1 design and temperature deration factor of 1.00. 6. A natural gas pipeline, 240 mi long, is constructed of NPS 24 pipe and has a required operating pressure of 1200 psig. Compare the cost of using X-70 or X-80 steel pipe. The material cost of X-70 pipe is $850/ton, and for X-80 pipe it is $1000/ton. Use a class 1 design factor and temperature deration factor of 1.00. 7. A natural gas pipeline, NPS 24, traverses a hilly terrain with elevations ranging from 300 ft at Norwalk to 4500 ft at the Fulton summit (milepost 50), followed by an elevation of 500 ft at the pipeline terminus at Danby. The pipeline is 100 mi long and is constructed of API 5L X-65 pipe as follows:

Section Norwalk to mp 30 mp 30 to Fulton Fulton to mp 70 mp 70 to Danby

Wall Thickness (in.)

Class

0.500 0.500 0.375 0.375

2 1 2 3

Determine a hydrostatic test plan for this pipeline, considering a test pressure envelope of 90 to 95% yield. What is the minimum number of test sections required?

REFERENCES 1. Nayyar, M.L., Piping Handbook, 7th ed., McGraw-Hill, New York, 2000. 2. Baumeister, T., Ed., Standard Handbook for Mechanical Engineers, 7th ed., McGrawHill, New York, 1967. 3. Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003. 4. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 5. Department of Transportation—DOT Code of Federal Regulation 49CFR Part 192, Oct. 2000. 6. ASME B31.8: Gas Transmission and Distribution Piping Systems, 2003–2005.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C006.fm Page 248 Friday, April 1, 2005 3:29 PM

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 249 Friday, April 1, 2005 3:28 PM

CHAPTER

7

Thermal Hydraulics In this chapter we will further discuss thermal hydraulics, which was briefly reviewed in Chapter 3. The importance of taking into account the gas temperature variation along a gas transmission pipeline and its impact on pressure drop and flow rate will be explained. Isothermal hydraulics, which formed the majority of calculations in Chapter 2 and Chapter 3, will be compared with thermal hydraulics. Since manual calculation of gas pipeline hydraulics, considering thermal effects, is quite laborious and time consuming, we will use examples of pipeline simulation cases using a popular gas pipeline hydraulics software application. 7.1 ISOTHERMAL VERSUS THERMAL HYDRAULICS In the previous chapters the hydraulic analysis of gas flow through pipelines was mainly done based upon isothermal or constant temperature flow. This assumption is fairly good in long-distance pipelines where the gas temperature reaches a constant value equal to or close to the surrounding soil (or ambient) temperature at large distances from the compressor stations. However, upon compressing the gas, depending on the compression ratio, the outlet temperature of the gas from the compressor station can be considerably higher than that of the ambient air or surrounding soil. In Chapter 4, Example 8, we found that when gas is compressed adiabatically from a 60°F suction temperature and a compression ratio of 2.0, the discharge temperature is 278.3°F. Since pipe coating limitations restrict temperatures to about 140 to 150°F, cooling of the compressed gas is necessary at the downstream side of the compressor station. In this example, assuming gas cooling results in a discharge temperature of 140°F as gas enters the pipeline, we find that the temperature difference between the gas at 140°F and the surrounding soil at 70°F will cause heat transfer to take place between the pipeline gas and the surrounding soil. It is found that the gas temperature drops off rapidly for the first few miles and eventually reaches a temperature close to the soil temperature. Additionally, in a long transmission pipeline, the soil temperature can vary along the pipeline as well, causing different heat transfer rates at locations along the pipeline. This is illustrated in Figure 7.1. 249

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 250 Friday, April 1, 2005 3:28 PM

250

GAS PIPELINE HYDRAULICS

Tinlet = 140°F

Tem

pera

ture

grad

ient Toutlet = 70°F

Compressor station Figure 7.1

Terminus

Temperature variation in a gas pipeline.

In some instances, the expansion of gas as it flows along a pipeline can result in gas temperature reaching a slightly lower temperature than the surrounding soil. This is called the Joule-Thompson cooling effect. Thus, if the soil temperature is fairly constant at 70°F, due to the Joule-Thompson effect, the final temperature of the gas at the terminus of the pipeline can drop to 60 or 65°F. This is illustrated in Figure 7.2. This cooling of gas below the surrounding soil temperature depends on the pressure differential and the Joule-Thompson coefficient. Ignoring this cooling effect will result in a more conservative (lower flow rate for a given pressure drop) flow rate calculation, since cooler temperature means less pressure drop in a gas pipeline and, hence, higher flow rate. Thermal hydraulics is the study of gas pipeline pressures, temperatures, and flow rates, taking into account the thermal properties of the soil, pipe, and pipe insulation, if any. Due to such variation in gas temperature, calculation of pressure drop must be made by considering short lengths of pipe that make up the total pipeline. For example, if the pipeline is 50 mi long, we subdivide it into short segments of 1 or 2 mi lengths and apply the General Flow equation for each pipe segment, considering

Gas temperature

Gas inlet temperature = 140°F

Ground temperature = 70°F

Distance Figure 7.2

Joule-Thompson cooling effect in a gas pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

Gas outlet temperature = 60–65°F

2785_C007.fm Page 251 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

251

an average gas temperature and an average ambient soil temperature. Starting with the upstream pressure of segment 1, the downstream pressure will be calculated assuming an average temperature for segment 1. Next, using the calculated downstream pressure as the upstream pressure for segment 2, we calculate the downstream pressure for segment 2. The process is continued until all segments of the pipeline are covered. It must be noted that the variation of temperature from segment to segment must be taken into account to calculate the compressibility factor used in the General Flow equation. The following equation is the General Flow equation that we used frequently in Chapter 2 and Chapter 3: T P 2 − e s P22 Q = 38.77 F b 1 Pb GT f Le Z

0.5 5

D 2.5

(7.1)

After examining Equation 7.1, we see that if the gas flowing temperature Tf is constant (isothermal flow) throughout the length of the pipeline Le, the compressibility factor Z of the gas will depend on the average pressure of the pipe segment. If the gas temperature also varies along the pipeline, based on the preceding discussions, the compressibility factor Z will change in a different manner, since it is a function of both the gas temperature and gas pressure. Therefore, it is seen that calculation of P1 or P2 for a given flow rate Q from Equation 7.1 will yield different results if isothermal conditions do not exist. The calculation of gas temperature at any point along the pipeline, taking into account the heat transfer between the gas and surrounding soil, is quite complicated. It does not lend itself easily to manual calculations. We will discuss the method of calculation for thermal hydraulics in this section. To accurately take into account the temperature variations, a suitable gas pipeline hydraulics simulation program must be used since, as indicated earlier, manual calculation is quite laborious and time consuming. Several commercial simulation programs are available to model steady-state gas pipeline hydraulics. These programs calculate the gas temperatures and pressures by taking into consideration variations of soil temperature, pipe burial depth, and thermal conductivities of pipe, insulation, and soil. One such software program is GASMOD, marketed by SYSTEK Technologies, Inc. (www.systek.us). Appendix D includes a sample simulation of a gas pipeline using the GASMOD software. In this chapter we will use GASMOD to illustrate thermal hydraulics analysis.

7.2 TEMPERATURE VARIATION AND GAS PIPELINE MODELING Consider a buried pipeline transporting gas from point A to point B. We will analyze a short segment of length ∆ L of this pipe, as shown in Figure 7.3, and apply the principles of heat transfer to determine how the gas temperature varies along the pipeline. The upstream end of the pipe segment of length ∆ L is at temperature T1 and the downstream end at temperature T2. The average gas temperature in this segment is represented by T. The outside soil temperature at this location is Ts . Assume steadystate conditions and the mass flow rate of gas to be m. The gas flow from the upstream

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 252 Friday, April 1, 2005 3:28 PM

252

GAS PIPELINE HYDRAULICS

Soil temperature Ts

A

T2

T

T1 Q

B

∆L 1

Figure 7.3

2

Analysis of temperature variation.

end to the downstream end of the segment causes a temperature drop of ∆T. The heat loss from the gas can be represented by ∆ H = − mCp∆T where ∆H m Cp ∆T

= = = =

(7.2)

heat transfer rate, Btu/h mass flow rate of gas, lb/h average specific heat of gas, Btu/lb/ °F temperature difference = T1 – T2, °F

The negative sign in Equation 7.2 indicates loss of heat from upstream temperature T1 to downstream temperature T2. Next, we consider the heat transfer from the gas to the surrounding soil in terms of the overall heat transfer coefficient U and the difference in temperature between the gas and surrounding soil represented by (T – Ts ). Therefore, we can write the following equation for heat transfer: ∆ H = U∆ A (T – Ts) where U ∆A T Ts D

(7.3)

= overall heat transfer coefficient, Btu/h/ft2 / °F = surface area of pipe for heat transfer = p D∆ L = average gas temperature in pipe segment, °F = average soil temperature surrounding pipe segment, °F = pipe inside diameter, ft

Equating the two values of the heat transfer rate ∆ H from Equation 7.2 and Equation 7.3, we get −mCp∆T = U∆A (T – Ts) Simplifying, we get πUD ∆T = − ∆L T − Ts mCp

Copyright 2005 by Taylor & Francis Group, LLC

(7.4)

2785_C007.fm Page 253 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

253

Rewriting Equation 7.4 in differential form and integrating, we get

∫

2

1

dT = T − Ts

πUD

∫ − mCp dL 1

(7.5)

2

Integrating and simplifying, we get T2 − Ts = e −θ T1 − Ts

(7.6)

where e = base of natural logarithms (e = 2.718…) and

θ=

πUD∆ L mCp

(7.7)

Simplifying Equation 7.6 further, we get the downstream temperature of the pipe segment of length ∆ L as T2 = Ts + (T1 − Ts )e −θ

(7.8)

It can be seen from Equation 7.8 that as the pipe length increases, the term e −θ approaches zero and the temperature T2 becomes equal to soil temperature Ts. Therefore, in a long gas pipeline, the gas temperature ultimately equals the surrounding soil temperature. This is illustrated in Figure 7.1. In the preceding analysis, we made several simplifying assumptions. We assumed that the soil temperature and the overall heat transfer coefficient remained constant and ignored the Joule-Thompson effect as gas expands through a pipeline. In a long pipeline, the soil temperature may actually vary along the pipeline and, therefore, must be taken into account in these calculations. One approach would be to subdivide the pipeline into segments that have constant soil temperatures and perform calculations for each segment separately. The Joule-Thompson effect causes the gas to cool slightly due to expansion. Therefore, in a long pipeline, the gas temperature at the delivery point can fall below that of the ground or soil temperature, as indicated in Figure 7.2.

7.3 REVIEW OF SIMULATION MODEL REPORTS To illustrate thermal effects in a gas pipeline, we will analyze a gas transmission pipeline, first using the method outlined in Chapter 3. Next, we will analyze the same pipeline, taking into account the thermal conductivity and soil temperatures. The latter method requires some form of computer simulation models. To do this, we have chosen the commercially available software known as GASMOD. We will compare the results of the isothermal hydraulics of Chapter 3 with the thermal hydraulics using GASMOD. Examples will be used to illustrate the comparison.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 254 Friday, April 1, 2005 3:28 PM

254

GAS PIPELINE HYDRAULICS

Example 1 A natural gas pipeline system is being built from Rockport to Concord, a distance of 240 miles. The pipeline is constructed of NPS 30, 0.500 in. wall thickness, API 5L-X60 pipe. The MOP is 1400 psig. Gas enters the Rockport compressor station at 70°F and 800 psig pressure. The soil temperature can be assumed to be 60°F throughout. The gas flow rate is 420 MMSCFD, and the gas specific gravity and viscosity are 0.6 and 0.000008 lb/ft-s, respectively. The contract delivery pressure required at Concord is 500 psig. Assume an isothermal flow at 70°F and a gas specific heat ratio of 1.29. Use a compressor adiabatic efficiency of 80% and mechanical efficiency of 98%. Use the General Flow equation with a Colebrook friction factor, assuming a pipe internal roughness of 700 µ in. Calculate the pressure profile and the compressor horsepower required at Rockport. Compare these results with thermal hydraulic analysis using GASMOD. Assume a base pressure of 14.7 psia and base temperature of 60°F. The pipeline elevation profile is essentially flat. Solution Inside diameter of pipe D = 30 – 2 × 0.500 = 29 in. First, we calculate the Reynolds number from Equation 2.34: 14.7 0.6 × 420 × 10 6 = 14,671,438 R = 0.0004778 60 + 460 0.000008 × 29 Next, using Colebrook Equation 2.39, we calculate the friction factor as 0.0007 2.51 = −2Log10 + f 3.7 × 29 14, 671, 438 f

1

Solving by trial and error, we get f = 0.0097 Therefore, the transmission factor is, using Equation 2.42, F=

2 0.0097

= 20.33

To calculate the compressibility factor Z, the average pressure is required. Since the inlet pressure is unknown, we will calculate an approximate value of Z using a value of 110% of the delivery pressure for the average pressure. The average pressure is Pavg = 1.1 × (500 + 14.7) = 566.17 psia = 551.47 psig

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 255 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

255

Using CNGA Equation 1.34, we calculate the value of compressibility factor as Z=

1 (566.17−14.7)×344,400(10)1.785× 0.6 1 + 5303.825

= 0.9217

Since there is no elevation difference between the beginning of the pipeline and the end of the pipeline, the elevation component in Equation 2.7 can be neglected and es = 1. The outlet pressure is P2 = 500 + 14.7 = 514.7 psia. From General Flow Equation 2.4, substituting given values, we get 520 P12 − 514.72 420 × 10 6 = 38.77 × 20.33 14.7 0.6 × 530 × 240 × 0.9217

0.5

(29.0)2.5

Solving for the upstream pressure, we get P1 = 1021.34 psia = 1006.64 psig Using this value of P1, we calculate the new average pressure using Equation 2.14: Pavg =

1021.34 × 514.7 2 1021.34 + 514.7 − = 795.87 psia 1021.34 + 514.7 3

This compares with the value of 566.17 psia we assumed initially for calculating Z. Obviously, we were way off. Recalculating Z using the recently calculated value of Pavg, we get Z=

1 (795.87−14.7)×344,400(10)1.785× 0.6 1 + 5303.825

= 0.8926

We will now recalculate the inlet pressure using this value of Z. From General Flow Equation 2.4, we get 520 P12 − 514.72 420 × 10 6 = 38.77 × 20.33 14.7 0.6 × 530 × 240 × 0.8926

0.5

(29.0)2.5

Solving for the upstream pressure, we get P1 = 1009.24 psia = 994.54 psig This compares with 1021.34 psia calculated earlier. This is almost 1% different. We could repeat the process and get a better approximation.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 256 Friday, April 1, 2005 3:28 PM

256

GAS PIPELINE HYDRAULICS

Using this recently calculated value of P1, we calculate the new average pressure using Equation 2.14 as Pavg =

1009.24 × 514.7 2 1009.24 + 514.7 − = 788.72 psia 1009.24 + 514.7 3

This compares with the previous approximation of 795.87 psia. The error is less than 1%. Recalculating Z using this value of Pavg , we get Z=

1 (788.72−14.7)×344,400(10)1.785× 0.6 1 + 5303.825

= 0.8935

We will now recalculate the inlet pressure using this value of Z. From General Flow Equation 2.4, we get 520 P12 − 514.72 420 × 10 = 38.77 × 20.33 14.7 0.6 × 530 × 240 × 0.8935 6

0.5

(29.0)2.5

Solving for the upstream pressure, we get P1 = 1009.62 psia = 994.92 psig This compares with 1009.24 psia calculated earlier. The difference is less than 0.04%; therefore, we can stop iterating any further. The HP required at the Rockport compressor station will be calculated using Equation 4.15 as follows: 0.29 1 + 0.8935 1 1009.62 1.29 1.29 HP = 0.0857 × 420 − 1 = 4962 (70 + 460) 0.8 814.7 2 0.29

Using Equation 4.17, we calculate the driver horsepower required, based on a mechanical efficiency of 0.98. BHP required =

4962 = 5063 0.98

The final results are: Inlet pressure at Rockport = 994.92 psig Delivery pressure at Concord = 500 psig at a flow rate of 420 MMSCFD BHP required at Rockport compressor station = 5063 HP

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 257 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

257

It must be noted that the preceding calculations ignored any elevation changes along the pipeline. If we had considered the pipe elevations at Rockport and Concord, the result would have been different. This pressure of 994.92 psig required at the Rockport compressor station was calculated assuming a constant gas flowing temperature of 70°F and considering the pipeline as one single segment 240 mi long. As explained in earlier chapters, the calculation accuracy is improved if we subdivide the pipeline into short segments. By doing so, we calculate the upstream pressure of each segment starting with the last segment near Concord. If the pipeline is divided into 100 equal pipe segments of 2.4 mi each, the pressure P100 at the upstream end of the last segment is calculated using the General Flow equation, considering a 500 psig downstream pressure. Next, using this calculated pressure, P100, we calculate the upstream pressure P99 of the 99th segment. The process is repeated until all segments are covered and the value of the pressure P1 at Rockport is calculated. This is illustrated in Figure 7.4. By subdividing the pipeline in this fashion, we are improving the accuracy of calculations. Of course, manual calculation in this manner is going to be quite laborious and time consuming, and we should use some form of a computer program to perform this task. Next, we will compare the isothermal calculation results with thermal hydraulics using the GASMOD program. Given next is the output report from the GASMOD program.

Pressure

P1

P99 P100 500 psig Subdivided into 100 segments

Rockport

Concord Distance = 240 mi

Figure 7.4

Subdividing pipe into segments.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 258 Friday, April 1, 2005 3:28 PM

258

GAS PIPELINE HYDRAULICS

******** GASMOD — GAS PIPELINE HYDRAULIC SIMULATION ******* ************ 32-bit Version 5.00.100 ************ DATE:

16-September-2004 TIME: 21:30:37

PROJECT DESCRIPTION: Pipeline from Rockport to Concord Case Number:

1139

Pipeline data file:

C:\GASMOD32\RockportPipeline.TOT

Pressure drop formula: Pipeline efficiency: Compressibility factor method: Inlet gas gravity(air=1.0): Inlet gas viscosity:

Colebrook-White 1.00 CNGA 0.60000 0.0000080(lb/ft-sec)

CALCULATION OPTIONS: Branch pipe calculations: Loop pipe calculations: Compressor fuel calculated: Joule-Thompson effect included: Customized output: Holding delivery pressure at terminus

NO NO NO NO NO

******** Calculations Based on Specified Thermal Conductivities of Pipe, Soil, and Insulation ******** Origin suction temperature: Base temperature: Base pressure: Origin suction pressure: Delivery pressure: Minimum pressure: Gas specific heat ratio: Maximum gas velocity:

70.00(degF) 60.00(degF) 14.700(psig) 800.00(psig) 500.35(psig) 100.00(psig) 1.29 50.00(ft/sec)

Inlet flow rate: Outlet flow rate:

420.0000(MMSCFD) 420.0000(MMSCFD)

**************** PIPELINE PROFILE DATA *********** Distance (mi)

Elevation (ft)

Diameter (in.)

Thickness (in.)

Roughness (in.)

0.00 10.00 20.00 30.00

250.00 250.00 250.00 250.00

30.000 30.000 30.000 30.000

0.500 0.500 0.500 0.500

0.000700 0.000700 0.000700 0.000700

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 259 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

40.00 50.00 60.00 70.00 100.00 120.00 140.00 150.00 170.00 190.00 200.00 220.00 240.00

259

250.00 250.00 250.00 250.00 250.00 250.00 250.00 250.00 250.00 250.00 250.00 250.00 250.00

30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000

0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500

0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700

******** THERMAL CONDUCTIVITY AND INSULATION DATA *********

Distance (mi)

Cover (in)

0.000 10.000 20.000 30.000 40.000 50.000 60.000 70.000 100.000 120.000 140.000 150.000 170.000 190.000 200.000 220.000 240.000

36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000

****************

Thermal Conductivity (Btu/hr/ft/degF) Pipe Soil Insulation 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000

0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800

0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020

Insul.Thk (in)

Soil Temp (degF)

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00

COMPRESSOR STATION DATA **************

FLOW RATES, PRESSURES, AND TEMPERATURES:

Name

Flow Suct. Disch. Rate Press. Press. (MMSCFD) (psig) (psig)

Rockport 420.00

Copyright 2005 by Taylor & Francis Group, LLC

800.00 996.17

Compr. Suct. Disch. Suct. Disch. Ratio Loss. Loss. Temp. Temp MaxPipe (psig) (psig) (degF) (degF) (degF) Temp 1.2408 0.00

0.00

70.00

102.92 140.00

2785_C007.fm Page 260 Friday, April 1, 2005 3:28 PM

260

GAS PIPELINE HYDRAULICS

********* COMPRESSOR EFFICIENCY, HP, AND FUEL USED ********

Name

Compr Mech. Overall Horse Distance Effy. Effy. Effy. Power (mi) (%) (%) (%)

Rockport 0.00

80.00 98.00 78.40

Fuel Factor (MCF/day/ HP)

4,786.94 0.2000

Total Compressor Station Horsepower:

Fuel Used (MMSCFD)

------

4,786.94

**************** LOCATIONS AND FLOW RATES **************** Location Distance Flow in/out Gravity Viscosity Pressure GasTemp. (mi) (MMSCFD) (lb/ft-sec) (psig) (degF) Rockport 0.00 Concord 240.00

420.0000 -420.0000

0.6000 0.6000

0.00000800 0.00000800

996.17 500.35

102.92 60.00

***** REYNOLD'S NUMBER AND HEAT TRANSFER COEFFICIENT *****

Distance (mi) 0.000 10.000 20.000 30.000 40.000 50.000 60.000 70.000 100.000 120.000 140.000 150.000 170.000 190.000 200.000 220.000 240.000

FrictFactor Transmission Reynold'sNum. (Darcy) Factor 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438.

0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099

HeatTransCoeff Compressibility (Btu/hr/ Factor ft2/degF) (CNGA)

20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13

0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361

0.8865 0.8791 0.8747 0.8725 0.8718 0.8722 0.8733 0.8767 0.8821 0.8870 0.8909 0.8952 0.9012 0.9061 0.9114 0.9171 0.9171

******* PIPELINE TEMPERATURE AND PRESSURE PROFILE ****** Distance Diameter Flow Velocity Press. GasTemp. SoilTemp. MAOP Location (mi) (in) (MMSCFD) (ft/sec) (psig) (degF) (degF) (psig) 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 100.00

30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000

Copyright 2005 by Taylor & Francis Group, LLC

420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 420.0000

15.44 15.44 15.44 15.44 15.44 15.44 15.44 15.44 15.44

996.17 979.59 963.25 946.96 930.60 914.07 897.30 880.23 826.86

102.92 87.37 77.22 70.74 66.66 64.11 62.54 61.56 60.36

60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00

1400.00 Rockport 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00

2785_C007.fm Page 261 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

120.00 140.00 150.00 170.00 190.00 200.00 220.00 240.00

30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000

420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 420.0000

261

15.44 15.44 15.44 15.44 15.44 15.44 15.44 15.44

789.12 749.30 728.51 684.85 637.95 613.06 559.72 500.35

60.14 60.05 60.03 60.01 60.00 60.00 60.00 60.00

60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00

1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00

Concord

************* LINE PACK VOLUMES AND PRESSURES ************* Distance (mi)

Pressure (psig)

Line Pack (million std.cu.ft)

0.00 996.17 0.0000 10.00 979.59 17.8138 20.00 963.25 17.9714 30.00 946.96 17.9566 40.00 930.60 17.8193 50.00 914.07 17.5990 60.00 897.30 17.3235 70.00 880.23 17.0114 100.00 826.86 48.9787 120.00 789.12 30.7496 140.00 749.30 29.1424 150.00 728.51 13.9438 170.00 684.85 26.5765 190.00 637.95 24.7442 200.00 613.06 11.6489 220.00 559.72 21.7584 240.00 500.35 19.5576 Total line pack in main pipeline = 350.5950(million std.cu.ft) ************** End of GASMOD Output Report ************* It can be seen from the GASMOD thermal hydraulic analysis report that the inlet pressure at Rockport is 996.17 psig, whereas the manual calculation considering isothermal flow yielded an inlet pressure of 994.92 psig at the Rockport compressor station. Thus, taking into account the temperature variation of the gas along the pipeline, the pressure required at Rockport is approximately 4 psig higher. This does not seem to be very significant. However, in many cases the temperature variation along the pipeline will cause pressures calculated to be significantly different. To recap, the manual calculations were based on an isothermal gas flow temperature of 70°F, whereas the thermal hydraulics shows variation of the gas temperature ranging from 102.92°F at the Rockport compressor discharge to 60°F at Concord. The gas temperature reaches the soil temperature of 60°F at approximately milepost 190, after which it remains constant at 60°F. Figure 7.5 shows the temperature variation in this case.

Next, we will illustrate the calculation of the pressure and temperature profile considering the pipeline elevation difference between Rockport and Concord and considering a branch pipeline bringing in an additional 200 MMSCFD.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 262 Friday, April 1, 2005 3:28 PM

262

GAS PIPELINE HYDRAULICS

Tinlet = 102.92°F T10 = 87.37°F T20 = 77.22°F T30 = 70.74°F Tem

pera

Rockport Figure 7.5

ture

T140 = 60.05°F

grad

ient

Toutlet = 60°F

240 mi

Concord

Gas temperature variation—Rockport to Concord pipeline.

Example 2 Consider a natural gas pipeline system from Rockport (elevation 250 ft) to Concord (elevation 800 ft), a distance of 240 mi. The pipeline is constructed of NPS 30, 0.500 in. wall thickness, API 5L-X60 pipe. The MOP is 1400 psig. The pipeline elevation profile is listed below: Milepost

Elevation

Location

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 100.00 120.00 140.00 150.00 170.00 190.00 200.00 220.00 240.00

250.00 300.00 200.00 320.00 400.00 375.00 410.00 430.00 450.00 500.00 400.00 600.00 700.00 710.00 720.00 750.00 800.00

Rockport

Vale

Concord

Gas enters the Rockport compressor station at 70°F and 800 psig pressure. The soil temperature can be assumed to be 60°F throughout. The gas flow rate is 420 MMSCFD, and the gas specific gravity and viscosity are 0.6 and 0.000008 lb/ft-s, respectively. At Vale (milepost 100, elevation 450 ft), a branch pipeline 80 mi long, NPS 24, 0.375 in. wall thickness, brings in an additional 200 MMSCFD gas from a gathering facility at Drake. The elevation at Drake is 300 ft and that at Vale is 450 ft.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 263 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

263

The pipeline elevation profile for the branch pipe from Drake to Vale is as follows:

Milepost

Elevation

Location

0.00 10.00 20.00 40.00 50.00 70.00 80.00

300.00 100.00 125.00 200.00 250.00 300.00 450.00

Drake

Vale

The inlet temperature at the beginning of the branch is 70°F in Figure 7.6. The contract delivery pressure required at Concord is 500 psig. Assume an isothermal flow at 70°F and gas specific heat ratio of 1.29. Use a compressor adiabatic efficiency of 80% and mechanical efficiency of 98%. Use the General Flow equation with a Colebrook friction factor, assuming a pipe internal roughness of 700 µ in. Calculate the pressure profile and the compressor horsepower required at Rockport. Compare these results with thermal hydraulic analysis using GASMOD. Assume a base pressure of 14.7 psia and base temperature of 60°F. Solution Inside diameter of pipe D = 30 – 2 × 0.500 = 29 in. The calculation of the pressure at milepost 100 will be done first. This is because we know the delivery pressure at Concord and the 140 mi segment from Vale (at milepost 100 to the pipeline terminus at Concord) flows 620 MMSCFD. In comparison, the first 100 mi from Rockport to Vale, carries only 420 MMSCFD, and both upstream and downstream pressures are unknown. After finding the pressure at Vale, we can

20

0

M

80

M

m

SC

iN

FD

PS

24

Drake elevation 300 ft

420 MMSCFD 240 mi NPS 30 Rockport m.p. 0 elevation 250 ft Figure 7.6

Vale m.p. 100 elevation 450 ft

Rockport to Concord pipeline with branch from Drake.

Copyright 2005 by Taylor & Francis Group, LLC

Concord m.p. 240 elevation 800 ft

2785_C007.fm Page 264 Friday, April 1, 2005 3:28 PM

264

GAS PIPELINE HYDRAULICS

calculate the upstream pressure at Rockport, considering the 100 mi segment at the lower flow rate. First, we calculate the Reynolds number from Equation 2.34 for a 620 MMSCFD flow rate: 14.7 0.6 × 620 × 10 6 = 21,657,837 R = 0.0004778 60 + 460 0.000008 × 29 Next, using Colebrook Equation 2.39, we calculate the friction factor as 0.0007 2.51 = −2Log10 + . 3 7 29 × 21, 657, 837 f f

1

Solving by trial and error, we get f = 0.0096 Therefore, the transmission factor is, using Equation 2.42, F=

2 0.0096

= 20.41

To calculate the compressibility factor Z, the average pressure is required. Since the inlet pressure at Vale is unknown, we will calculate an approximate value of Z using a value of 110% of the delivery pressure for the average pressure. The average pressure is Pavg = 1.1 × (500 + 14.7) = 566.17 psia = 551.47 psig Using CNGA Equation 1.34, we calculate the value of the compressibility factor as Z=

1 = 0.9217 (566.17−14.7)×344,400(10)1.785× 0.6 + 1 5303.825

Since there is an elevation difference of 350 (800 − 450) ft between Vale and Concord, we must apply the elevation correction according to Equation 2.7. Using Equation 2.10, the elevation adjustment parameter is 800 − 450 s = 0.0375 × 0.6 = 0.0161 530 × 0.9217 The equivalent length Le from Equation 2.9 is Le =

Copyright 2005 by Taylor & Francis Group, LLC

140(1.0163 − 1) = 141.74 mi 0.0161

2785_C007.fm Page 265 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

265

Next, using Equation 2.7, we calculate the pressure at Vale, milepost 100, as follows: 520 P12 − 1.0163 × 514.72 620 × 10 = 38.77 × 20.41 14.7 0.6 × 530 × 141.74 × 0.9217

0.5

292.5

6

Solving for the upstream pressure at Vale, we get P1 = 1122.49 psia = 1107.79 psig Using this calculated value of P1, we calculate the new average pressure using Equation 2.14 as Pavg =

2 1122.49 × 514.7 = 856.2 psia = 841.5 psig 1122.49 + 514.7 − 3 1122.49 + 514.7

This compares with the previous approximation of 551.47 psig. Obviously, the assumed value was way off. Recalculating the compressibility factor, using the recently calculated average pressure, Z=

1 (856.2−14.7)×344,400(10)1.785× 0.6 1 + 5303.8825

= 0.8852

Recalculating the pressure at Vale, we get 520 P12 − 1.0163 × 514.72 620 × 10 6 = 38.77 × 20.41 14.7 0.6 × 530 × 141.74 × 0.8852

0.5

292.5

Solving for the upstream pressure at Vale, we get P1 = 1104.88 psia = 1090.18 psig Compared to the last calculated value, the difference is: 1090.18 – 1107.79 = −17.61 psig or 1.6%. One more iteration would get us closer to the correct value. We recalculate the new average pressure using Equation 2.14 as Pavg =

1104.88 × 514.7 2 1104.88 + 514.7 − = 845.63 psia = 830.93 psig 1104.88 + 514.7 3

We recalculate the compressibility factor, using the recently calculated average pressure, as Z=

1 (845.63−14.7)×344,400(10)1.785× 0.6 1 + 5303.825

= 0.8865

Recalculating the pressure at Vale, we get 520 P12 − 1.0163 × 514.72 620 × 10 = 38.77 × 20.41 14.7 0.6 × 530 × 141.74 × 0.8865 6

Copyright 2005 by Taylor & Francis Group, LLC

0.5

292.5

2785_C007.fm Page 266 Friday, April 1, 2005 3:28 PM

266

GAS PIPELINE HYDRAULICS

Solving for the upstream pressure at Vale, we get P1 = 1106.15 psia = 1091.45 psig Compared to the last calculated value, the difference is: 1090.18 – 1091.45 = −1.27 psig or 0.12%. This is close enough, and no further iteration is needed. Therefore, the pressure at Vale = 1106.15 psia = 1091.45 psig. Next, using this as the downstream pressure for the pipe segment from Rockport to Vale, we calculate the upstream pressure at Rockport as follows, for a flow rate of 420 MMSCFD. From previous calculations at 420 MMSCFD (Example 1), the Reynolds number is R = 14,671,438 The friction factor was calculated as f = 0.0097 and the transmission factor was F=

2 0.0097

= 20.33

To calculate the compressibility factor Z, the average pressure is required. Since the inlet pressure at Rockport is unknown, we will calculate an approximate value of Z using a value of 110% of the downstream pressure at Vale for the average pressure. The average pressure is Pavg = 1.1 × 1106.15 = 1216.77 psia = 1202.07 psig Using this, we calculate the compressibility factor as Z=

1 (1216.77−14.7)×344,400(10)1.785× 0.6 1 + 5303.825

= 0.8437

Since there is an elevation difference of 200 (450 − 250) ft between Rockport and Vale, the elevation correction according to Equation 2.7 must be applied. Using Equation 2.10, the elevation adjustment parameter is 450 − 250 = 0.0101 s = 0.0375 × 0.6 530 × 0.8437 The equivalent length Le from Equation 2.9 is Le =

Copyright 2005 by Taylor & Francis Group, LLC

100(1.0101 − 1) = 100 mi 0.0101

2785_C007.fm Page 267 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

267

Calculating the pressure at Rockport, we get 520 P12 − 1.0101 × 1106.152 420 × 10 6 = 38.77 × 20.33 14.7 0.6 × 530 × 100 × 0.8437

0.5

292.5

Solving for the upstream pressure at Rockport, we get P1 = 1238.04 psia = 1223.34 psig Next, we will compare the isothermal calculation results with thermal hydraulics using the GASMOD program. Given below is the output report from the GASMOD program. ******** GASMOD — GAS PIPELINE HYDRAULIC SIMULATION ******* ************ 32-bit Version 5.00.100 ************ DATE: PROJECT DESCRIPTION: Pipeline from Rockport to Concord Case number: Pipeline data file:

17-September-2004 TIME: 07:16:19

1143 C:\GASMOD32\RockportPipeline.TOT

Pressure drop formula: Pipeline efficiency: Compressibility factor method: Inlet gas gravity (air=1.0): Inlet gas viscosity:

Colebrook-White 1.00 CNGA 0.60000 0.0000080(lb/ft-sec)

CALCULATION OPTIONS: Branch pipe calculations: Loop pipe calculations: Compressor fuel calculated: Joule-Thompson effect included: Customized output: Holding delivery pressure at terminus

YES NO NO NO NO

**** Calculations Based on Specified Thermal Conductivities of Pipe, Soil, and Insulation **** Origin suction temperature: Base temperature: Base pressure: Origin suction pressure: Delivery pressure: Minimum pressure: Gas specific heat ratio: Maximum gas velocity:

70.00(degF) 60.00(degF) 14.700(psig) 800.00(psig) 499.66(psig) 100.00(psig) 1.29 50.00(ft/sec)

Inlet flow rate: Outlet flow rate:

420.0000(MMSCFD) 620.0000(MMSCFD)

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 268 Friday, April 1, 2005 3:28 PM

268

GAS PIPELINE HYDRAULICS

**************** PIPELINE PROFILE DATA *********** Distance (mi)

Elevation (ft)

Diameter (in)

Thickness (in)

Roughness (in)

250.00 300.00 200.00 320.00 400.00 375.00 410.00 430.00 450.00 500.00 400.00 600.00 700.00 710.00 720.00 750.00 800.00

30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000

0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500 0.500

0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 100.00 120.00 140.00 150.00 170.00 190.00 200.00 220.00 240.00

********* THERMAL CONDUCTIVITY AND INSULATION DATA ********

Distance (mi)

Cover (in)

0.000 10.000 20.000 30.000 40.000 50.000 60.000 70.000 100.000 120.000 140.000 150.000 170.000 190.000 200.000 220.000 240.000

36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000

Copyright 2005 by Taylor & Francis Group, LLC

Thermal Conductivity (Btu/hr/ft/degF) Insul.Thk Pipe Soil Insulation (in) 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000

0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800

0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

Soil Temp (degF) 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00

2785_C007.fm Page 269 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

269

**************** COMPRESSOR STATION DATA ************** FLOW RATES, PRESSURES, AND TEMPERATURES: Flow Suct. Disch. Suct. Disch. Suct. Disch. MaxPipe Rate Press. Press. Compr. Loss. Loss. Temp. Temp Temp Name (MMSCFD) (psig) (psig) Ratio (psig) (psig) (degF) (degF) (degF) Rockport 420.00

800.00 1223.97 1.5204 0.00

0.00

70.00

132.24 140.00

****** COMPRESSOR EFFICIENCY, HP, AND FUEL USED *********

Name

Compr Mech. Overall Distance Effy. Effy. Effy. Horse (mi) (%) (%) (%) Power

Rockport 0.00

80.00 98.00 78.40

Fuel Fuel Factor Used (MCF/day/HP) (MMSCFD)

9,513.40 0.2000

------

Total Compressor Station Horsepower: 9,513.40

**************** LOCATIONS AND FLOW RATES **************** Location Distance Flow in/out Gravity Viscosity Pressure GasTemp. (mi) (MMSCFD) (lb/ft-sec) (psig) (degF) Rockport 0.00 100.00 Concord 240.00

420.0000 200.0000 -620.0000

0.6000 0.6000 0.6000

0.00000800 0.00000800 0.00000800

1223.97 1085.71 499.66

132.24 60.65 60.01

****** REYNOLD'S NUMBER AND HEAT TRANSFER COEFFICIENT *****

Distance (mi) Reynold'sNum. 0.000 10.000 20.000 30.000 40.000 50.000 60.000 70.000 100.000 120.000 140.000 150.000 170.000 190.000 200.000 220.000 240.000

14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 14,671,438. 21,657,838. 21,657,838. 21,657,838. 21,657,838. 21,657,838. 21,657,838. 21,657,838. 21,657,838. 21,657,838.

Copyright 2005 by Taylor & Francis Group, LLC

FrictFactor Transmission Factor (Darcy) 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0099 0.0098 0.0098 0.0098 0.0098 0.0098 0.0098 0.0098 0.0098 0.0098

20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.13 20.25 20.25 20.25 20.25 20.25 20.25 20.25 20.25 20.25

HeatTransCoeff Compressibility (Btu/hr/ Factor ft2/degF) (CNGA) 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3361 0.3365 0.3365 0.3365 0.3365 0.3365 0.3365 0.3365 0.3365 0.3365

0.8820 0.8678 0.8578 0.8516 0.8479 0.8460 0.8454 0.8465 0.8516 0.8589 0.8650 0.8721 0.8821 0.8903 0.8998 0.9148 0.9148

2785_C007.fm Page 270 Friday, April 1, 2005 3:28 PM

270

GAS PIPELINE HYDRAULICS

******** PIPELINE TEMPERATURE AND PRESSURE PROFILE ******** Distance (mi)

Diameter Flow Velocity Press. GasTemp. SoilTemp. MAOP Location (in) (MMSCFD) (ft/sec) (psig) (degF) (degF) (psig)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 100.00 120.00 140.00 150.00 170.00 190.00 200.00 220.00 240.00

30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000 30.000

420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 420.0000 620.0000 620.0000 620.0000 620.0000 620.0000 620.0000 620.0000 620.0000 620.0000

12.60 12.60 12.60 12.60 12.60 12.60 12.60 12.60 18.60 18.60 18.60 18.60 18.60 18.60 18.60 18.60 18.60

1223.97 1208.59 1197.97 1181.38 1166.06 1153.87 1139.88 1126.25 1085.71 1023.96 961.88 922.84 847.15 765.54 720.84 620.90 499.66

132.24 107.14 90.16 79.02 71.88 67.38 64.56 62.81 60.65 60.34 60.17 60.13 60.06 60.03 60.02 60.01 60.01

60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00 60.00

1400.00 Rockport 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 Concord

************ LINE PACK VOLUMES AND PRESSURES ************

Distance (mi)

Pressure (psig)

Line Pack (million std.cu.ft)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 100.00 120.00 140.00 150.00 170.00 190.00 200.00 220.00 240.00

1223.97 1208.59 1197.97 1181.38 1166.06 1153.87 1139.88 1126.25 1085.71 1023.96 961.88 922.84 847.15 765.54 720.84 620.90 499.66

0.0000 21.2972 22.0280 22.4450 22.5706 22.5714 22.4716 22.2810 65.4581 41.3973 38.6813 18.2370 34.0291 30.7118 14.0358 25.1637 20.7997

Total line pack in main pipeline = 444.1785(million std.cu.ft)

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 271 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

271

NUMBER OF PIPE BRANCHES = 1 BRANCH TEMPERATURE AND PRESSURE PROFILE: Incoming Branch File: VALEBRANCH.TOT Branch Location: at 100 (mi) Distance Elevation Diameter Flow Velocity Press. Gas Temp. Amb Temp. Location (mi) (ft) (in) (MMSCFD) (ft/sec) (psig) (degF) (degF) 0.00 10.00 20.00 40.00 50.00 70.00 80.00

150.00 100.00 125.00 200.00 200.00 200.00 250.00

24.000 24.000 24.000 24.000 24.000 24.000 24.000

200.000 200.000 200.000 200.000 200.000 200.000 200.000

9.81 9.81 9.96 10.14 10.23 10.40 10.51

1163.79 1156.06 1146.13 1125.37 1115.97 1096.89 1085.79

70.00 63.97 61.56 60.24 60.09 60.01 60.01

60.00 60.00 60.00 60.00 60.00 60.00 60.00

Drake

Vale

*************** End of GASMOD Output Report ************* It can be seen from the GASMOD thermal hydraulic analysis report that the inlet pressure at Rockport is approximately 1224 psig, whereas the manual calculation considering isothermal flow yielded an inlet pressure of approximately 1223 psig at the Rockport compressor station. This difference is not very significant. However, in many cases the temperature variation along the pipeline will cause pressures calculated to be significantly different, especially in short pipelines. As an example, if we had a pipeline 100 mi long similar to the pipe section between Rockport and Vale, the thermal hydraulics will show a drastic temperature variation, from 132.24 to 60.65°F. Therefore, an isothermal analysis at 70°F for the entire 100 mi length will show considerable discrepancy in pressures. This is left as an exercise for the reader.

To recap, the manual calculations were based on an isothermal gas flow temperature of 70°F, whereas the thermal hydraulics shows variation of the gas temperature ranging from 132.24°F at the Rockport compressor discharge to 60.01°F at Concord, which is very close to the surrounding soil temperature of 60°F. The compression ratio is 1.52 at the Rockport compressor station, where the 800 psig inlet pressure of the gas is increased to the discharge pressure of 1224 psig. This, in accordance with our previous analysis in Chapter 4 under compressors, causes the discharge temperature of the gas to increase to 132.24°F. If the compression ratio were higher, the discharge temperature of the gas due to compression would have been still higher. The pipeline coating temperature limitation is 140°F and would then require gas cooling in order to avoid damage to the pipe coating. It can be seen from the GASMOD report that the gas flow temperature starts off at 132.24°F at milepost 0 (Rockport) and quickly drops to 67.38°F at milepost 50. This is the exponential temperature decay we discussed in an earlier section. Beginning at milepost 50, the gas temperature starts dropping off more gradually until it almost attains the soil temperature at milepost 240 (Concord). Also, the section of pipe between milepost 170 and Concord is at a fairly constant temperature, close to the soil temperature of 60°F. Therefore, the 70 mi pipeline section between milepost 170 and Concord can be considered to be in isothermal flow, for all practical purposes. A manual calculation of this last 70 mi pipe segment flowing

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 272 Friday, April 1, 2005 3:28 PM

272

GAS PIPELINE HYDRAULICS

at 620 MMSCFD will yield a pressure profile very close to the pressures shown in the GASMOD report. This will be demonstrated next. First, we calculate the Reynolds number from Equation 2.34: 14.7 0.6 × 620 × 10 6 = 21,657,837 R = 0.0004778 60 + 460 0.000008 × 29 Next, using Colebrook Equation 2.39, we calculate the friction factor as 0.0007 2.51 = −2Log10 + f 3.7 × 29 21, 657, 837 f

1

Solving by trial and error, we get f = 0.0096 Therefore, the transmission factor is, using Equation 2.42, F=

2 0.0096

= 20.41

To calculate the compressibility factor Z, the average pressure is required. Since the pressure at milepost 170 is unknown, we will calculate an approximate value of Z using a value of 850 psig for the average pressure. Using CNGA Equation 1.34, we calculate the value of the compressibility factor as Z=

1 = 0.8765 850×344 ,400 (10 )1.785× 0.6 1 + 5203.825

Since there is an elevation difference of 100 (800 − 700) ft between milepost 170 and Concord, we must apply the elevation correction according to Equation 2.7. Using Equation 2.10, the elevation adjustment parameter is 800 − 700 s = 0.0375 × 0.6 = 0.0049 520 × 0.8765 The equivalent length Le from Equation 2.9 is Le =

70(1.0049 − 1) = 70.17 mi 0.0049

Next, using Equation 2.7, we calculate the pressure at milepost 70 as follows: 520 P12 − 1.0049 × 514.72 620 × 10 = 38.77 × 20.41 14.7 0.6 × 520 × 70.17 × 0.8765 6

P1 = 851.59 psia = 836.89 psig

Copyright 2005 by Taylor & Francis Group, LLC

0.5

292.5

2785_C007.fm Page 273 Friday, April 1, 2005 3:28 PM

THERMAL HYDRAULICS

273

This compares with 850 psig we assumed earlier. Recalculating the average pressure based on P1 = 851.59 psia and P2 = 514.7 psia, we get, using Equation 2.14, Pavg =

2 851.59 × 514.7 = 696.99 psia 851.59 + 514.7 − 3 851.59 + 514.7

Next, using CNGA Equation 1.34, we recalculate the compressibility factor as Z=

1 682.29×344 ,400 (10 )1.785× 0.6 1 + 5203.825

= 0.8984

Next, using Equation 2.7, we calculate the pressure at milepost 70 as follows: 520 P12 − 1.0049 × 514.72 620 × 10 = 38.77 × 20.41 14.7 0.6 × 520 × 70.17 × 0.8984 6

0.5

292.5

P1 = 858.30 psia = 843.6 psig This value is less than 1% different from the previously calculated value of 836.39 psig. Therefore, we do not have to iterate any further. Comparing the value of 843.6 psig with the pressure of approximately 847 psig from the GASMOD report, we see that we are less than 0.5% apart. Thus, the assumption of isothermal flow in the last 70 mi section of the pipeline is a valid one. We would have been closer still if the 70 mi section had been subdivided into two or more segments and the upstream pressures calculated as discussed in an earlier section. In conclusion, we can state that calculating the pressures and HP in a gas pipeline based on the assumption of constant temperature throughout the pipeline will yield satisfactory answers if the pipeline is long. For shorter pipelines, calculations must be performed by subdividing the pipeline into short segments and taking into account heat transfer between the pipeline gas and the surrounding soil.

7.4 SUMMARY In this chapter we reviewed the thermal effects of pressure drop and horsepower required in a natural gas pipeline system. We pointed out the differences between the results obtained from isothermal and thermal hydraulic analysis. This was illustrated with example problems using an isothermal analysis compared to a more rigorous approach considering heat transfer between the pipeline gas and the surrounding soil. A popular gas pipeline hydraulic simulation software application was used to illustrate the calculation methodology.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C007.fm Page 274 Friday, April 1, 2005 3:28 PM

274

GAS PIPELINE HYDRAULICS

PROBLEMS 1. Apply the technique discussed in the temperature variation calculation section to calculate the temperature profile of a gas pipeline 4 mi long, NPS 20, with 0.375 in wall thickness, at a flow rate of 200 MMSCFD. 2. A 200 mile, NPS 24, 0.500 inch wall thickness pipeline from Mobile to Savannah is used for transporting 300 MMSCFD of natural gas (gravity = 0.65 and viscosity = 0.000008 lb/ft-s). The MOP is 1400 psig. The gas inlet temperature and pressure at Mobile are 80°F and 1200 psig, respectively. The soil temperature can be assumed to be 60°F throughout. The delivery pressure required at Savannah is 900 psig. Assume isothermal flow at 70°F. Using the Panhandle B equation with an efficiency of 0.95, calculate the free flow volume with no compressor stations. Compare these results with thermal hydraulic analysis using subdivided pipe segments and heat transfer calculations. The base pressure is 14.7 psia and base temperature is 60°F.

REFERENCES 1. Nayyar, M.L., Piping Handbook, 7th ed., McGraw-Hill, New York, 2000. 2. Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003. 3. McCain, W.D. Jr., The Properties of Petroleum Fluids, Petroleum Publishing Company, Tulsa, OK, 1973. 4. Holman, J.P., Thermodynamics, 2nd ed., McGraw-Hill, New York, 1974. 5. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 6. Engineering Data Book, 10th ed., Gas Processors Suppliers Association, Tulsa, OK, 1994. 7. Pipeline Design for Hydrocarbon Gases and Liquids, American Society of Civil Engineers, New York, 1975. 8. Steady Flow in Gas Pipelines, Contract Report No. 10, July 1965, Pipeline Research Council International, Houston, TX, 1965. 9. McAdams, W.H., Heat Transmission, McGraw-Hill, New York, 1954. 10. Katz, D.L. et al., Handbook of Natural Gas Engineering, McGraw-Hill, New York, 1959.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 275 Friday, April 1, 2005 3:27 PM

CHAPTER

8

Transient Analysis and Case Studies This chapter is devoted to transient pressure analysis and case studies of typical longdistance gas transmission pipelines. The subject of transient pressure analysis is quite complex, and understanding the theory behind it requires delving into differential equations and solution by the method of characteristics. Further, these calculations require some form of computer simulation to arrive at meaningful results. Nevertheless, we will discuss several scenarios that are typical of unsteady flow in gas pipelines that cause transient conditions. The objective is to determine how the pressure varies along the pipeline due to disturbances caused by transient conditions, such as a mainline valve closure and compressor station shutdown. If these transient conditions cause the pipeline pressures at some points to exceed the MOP, measures must be provided to ensure that the pressures do not violate the limits allowed by design codes. For detailed analysis of transient pressures in gas pipelines, the reader should refer to the texts listed in the Reference section of this chapter. We will also be reviewing several real-world pipeline transportation scenarios that encompass the concepts covered in the preceding chapters.

8.1 UNSTEADY FLOW In the preceding chapters, we analyzed pipelines that were in steady-state flow. This means that, at any point in time, the pipelines were operating at constant flow rates with a constant pressure and temperature profile. In other words, if the pressures, temperatures, and flow rates were measured at some point in time, say at 10 a.m. on a certain day, these parameters persisted in values throughout the period under investigation. Therefore, at some other time, such as 12 noon or 5 p.m., the pressure profile, the temperature profile, and the gas flow rates were all the same as that at 10 a.m. In reality, this does not happen. Due to one reason or another, the flow rates and pressures tend to change. This might be due to changes in delivery conditions, such as variation in the amount of gas being received at delivery stations due to changes in operation of facilities that require the gas. Further environmental conditions, such as

275

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 276 Friday, April 1, 2005 3:27 PM

276

GAS PIPELINE HYDRAULICS

atmospheric temperatures, can cause variation in compressor performance, resulting in flow and pressure changes at the compressor discharge. The latter will, in turn, cause changes in pressure and temperature of the pipeline gas. Another reason for unsteady flow condition can be a result of switching or closing valves to divert flow to different customers or shutdown or startup of compressor stations. Such unsteady operation causes transient pressures in the pipeline. Under unsteady or transient flow, the pressures, temperatures, and flow rates become time dependent. This means that it will no longer be correct to use results based upon steady-state flow calculations. 8.1.1

Transient Due to Mainline Valve Closure

To illustrate a transient condition, let us review a simple pipeline system with a head compressor station and a mainline valve at the end of the pipeline, as shown in Figure 8.1. The pipeline has been operating in steady-state condition for a long time. The valve at the end of the pipeline is suddenly closed due to malfunction or human error. Immediately, the pressure at the valve and at points upstream of it starts to rise, as shown by the dashed lines in Figure 8.1. Since gas is compressible, the compressor at the upstream end continues to pump gas without sensing the pressure rise downstream. This will result in an increase in line pack in the downstream section of the pipe, which will progress toward the upstream end. The transient pressure waves moving upstream will eventually reach the discharge of the compressor, causing the discharge pressure to rise. If the increased pressure attains the discharge shutdown setting, the compressor will trip and shut down, producing no further pumping pressure. The blocked-in gas in the pipeline will continue to undergo pressure variation from upstream to downstream as the pressure waves go back and forth at the speed of sound in gas. Eventually, the pressure surge dies out because of friction and loss of inertia resulting from reduction in gas velocity. This is illustrated in another

1400 psig

Hydr

Pressure

aulic

pressure rise with time after valve closure

pres

sure

grad

ient 600 psig

Pipe elevation profile

A Compressor station Figure 8.1

Distance

Transient due to valve closure.

Copyright 2005 by Taylor & Francis Group, LLC

B Terminus

2785_C008.fm Page 277 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

277

G C Head

B

K

H

E

he

s

Sy

ad he or

A

tem

s es pr om

C

ad

F

D

Q Gas flow rate, ACFM Figure 8.2

Compressor performance curve vs. pipeline system head curve.

way in Figure 8.2, using the compressor performance curve and the pipeline system curve. The steady-state system head curve is represented by AB and the compressor performance curve by CD. The steady-state operating point is therefore at E, where the compressor head H matches the pipeline system head required at the flow rate Q. As the valve at the end of the pipeline is closed, the system head curve shifts to the left, indicating reduction in gas flow due to increased pipeline resistance caused by the constriction in the valve. 8.1.2

Transient Due to Compressor Shutdown

Another transient condition that can occur in a simple pipeline described in Figure 8.2 is that of a compressor station shutdown from a steady-state operating condition. Suppose the compressor shuts down in 30 seconds after a long period of steady-state flow. Since there is no pressure being generated at the upstream end of the pipeline, but gas continues to be delivered at the downstream end, the line pack in the pipeline starts reducing starting at the downstream end. The pressures continue to fall along the pipeline and eventually stabilize at some blocked-in pressure. Another slightly complicated compressor station shutdown scenario that causes transient pressures is illustrated in Figure 8.3. In this case, a pipeline with two compressor stations is shown with a hydraulic pressure gradient under steady-state conditions. If the intermediate compressor station shuts down and the gas continues to be pumped from the first compressor station bypassing the second compressor station, the hydraulic pressure gradient will eventually be as indicated in Figure 8.3. However, before steady-state conditions are achieved with only one compressor operating, transient pressures are developed from the point of shutdown of the intermediate compressor station. Suppose the initial flow rate with both compressors operating under steady-state conditions is indicated by a flow rate of Q. The compressor

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 278 Friday, April 1, 2005 3:27 PM

278

GAS PIPELINE HYDRAULICS

Hydraulic pressure gradient with one compressor station 1400 psig

1400 psig

900 psig 800 psig

Flow

Sheridan

Compton

Beaumont

NPS 16 pipeline 200 mi long Figure 8.3

Transient due to compressor station shutdown.

Head

ead

K

H

st em

C

sor h

he

G

pres

B

Sy

Com

ad

performance curve superimposed on the pipeline system head curve for the pipe segment between the two compressor stations is shown in Figure 8.4. Initially, the system head curve AB for the pipe segment 1 between the two compressor stations results in a flow rate of Q with the operating point at E. At this point, the compressor head H of the first compressor station matches the pipeline system head required at the flow rate Q. When the second compressor station shuts down, it no longer provides the discharge pressure to boost the gas in pipe segment 2. Therefore, the first compressor station has to push the gas all the way to the end of the pipeline. It therefore has to contend with a longer pipe segment, which has a system head curve FG as shown in Figure 8.4. It can be seen that the new operating point K is at

E

00

,0

15 R

F

PM

D

00

,0

12

A

Q1

Q

Gas flow rate, ACFM Figure 8.4

Transient due to intermediate compressor shutdown.

Copyright 2005 by Taylor & Francis Group, LLC

PM

R

Q2

2785_C008.fm Page 279 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

279

a reduced flow rate Q1. If the point K on the compressor head curve is at too high a discharge pressure, the control mechanism will signal the compressor to slow down in speed. Thus, if the original compressor curve was based on 15,000 rpm, the compressor would slow down to a speed such as 12,000 rpm. This results in a new operating point, L, corresponding to a flow rate Q2, as shown in Figure 8.4. In summary, shutting down the second compressor station causes the operating point to move from point E on the compressor head curve at 15,000 rpm down to point L on the compressor head curve at 12,000 rpm. Correspondingly, the flow rate will decrease from Q at point E to Q2 at point L. 8.2 CASE STUDIES In the next few pages of this chapter, we are going to look at some real-life gas transmission pipeline systems. We will be applying the concepts learned in the previous chapters to determine the pressures and flow rates required in various scenarios. 8.2.1

Offshore Pipeline Case

Consider a gas production facility located offshore. The gas is compressed from the offshore platform through submarine pipelines that go ashore and subsequently connect to onshore pipelines for transportation of gas to industrial consumers. We will look at sizing such an offshore and onshore piping system for transporting a given quantity of gas. Calculations will be performed considering different options such as the AGA equation and Panhandle equations. We will illustrate this using an example. Case Study 1—Offshore/Onshore Pipeline A natural gas pipeline system originates at an offshore facility that compresses the gas through 200 mi of NPS 30, 0.625 in. wall thickness submarine pipelines to an onshore location, as depicted in Figure 8.5. A compressor station located onshore is used to compress the gas through a 120 mi, NPS 24, 0.500 inch wall thickness onshore buried pipeline for eventual delivery to a power plant. Determine the maximum flow rate possible under the following conditions. Neglect elevation effects. The compression ratio is 1.5. Use the Weymouth equation with 95% efficiency. Assume a base pressure of 14.7 psia and base temperature of 60°F. The gas flowing temperature is 60 °F and the compressibility factor is 0.88. The gas gravity is 0.65. 1480 psig

Compressor Station 500 psig 120 mNPS 24

200 mNPS 30

Figure 8.5

Offshore/onshore pipeline.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 280 Friday, April 1, 2005 3:27 PM

280

GAS PIPELINE HYDRAULICS

(a) Gas pressure at the platform equals 1480 psig and free flow occurs without use of any compression offshore or onshore. The delivery pressure at the power plant is 500 psig. (b) Considering the MOP at the platform and onshore equal to 1480 psig, determine the maximum throughput possible with compression offshore and onshore. Solution (a) Free flow with 1480 psig at the offshore platform. Using Weymouth Equation 2.52, neglecting elevation effects, calculate the pressure at the beginning of the NPS 24 onshore pipeline. The pipe inside diameter = 24 – 2 × 0.500 = 23 in. P12 − 514.72 60 + 460 Q = 433.5 × 0.95 14.7 0.65 × 520 × 120 × 0.88

0.5

232.667

(8.1)

28.752.667

(8.2)

Similarly, considering the NPS 30 pipeline 200 mi long, Pipe inside diameter = 30 − 2 × 0.625 = 28.75 in. 60 + 460 1494.72 − P12 Q = 433.5 × 0.95 14.7 0.65 × 520 × 200 × 0.88

0.5

Eliminating Q from both Equation 8.1 and Equation 8.2 by division, we get P 2 − 514.72 1= 1 2 2 1494.7 − P1

0.5

200 120

0.5

23.0 28.75

2.667

(8.3)

Solving for the pressure P1 at the junction of the two pipes onshore, P1 = 1253.7 psia = 1239 psig Next, substituting this value of P1 in Equation 8.1, we calculate the free flow volume flow rate as 520 1253.72 − 514.72 Q = 433.5 × 0.95 14.7 0.65 × 520 × 120 × 0.88

0.5

232.667

or Q = 377.53 MMSCFD Therefore, without any compression, the free flow possible is 377.53 MMSCFD.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 281 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

281

(b) With compressors installed at onshore and offshore locations, each location will be delivering at an MOP of 1480 psig. With a compression ratio of 1.5, the suction pressure at the onshore compressor is Ps =

1480 + 14.7 = 996.47 psia = 981.77 psig 1.5

First, calculate the capacity of the NPS 24 onshore pipeline, considering 1480 psig at the upstream end and 500 psig at the downstream end 120 mi away. Using Weymouth Equation 2.52, 520 1494.72 − 514.72 Q = 433.5 × 0.95 × 14.7 0.65 × 520 × 120 × 0.88

0.5

232.667

Q = 463.43 MMSCFD Next, we must determine if the offshore NPS 30 pipeline can transmit this flow starting at 1480 psig at the offshore platform and with a downstream pressure of 981.77 psig calculated earlier. 520 1494.72 − 996.472 Q = 433.5 × 0.95 × 14.7 0.65 × 520 × 200 × 0.88

0.5

(28.75)2.667

Solving for Q, we get Q = 516.76 MMSCFD Thus, the NPS 30 submarine pipeline has a capacity of 516.76 MMSCFD, whereas the onshore NPS 24 pipeline has a capacity of only 463.43 MMSCFD. Picking the lower of the two flow rates, the maximum throughput possible with the onshore compressor is 463.43 MMSCFD.

Case Study 2—Gas Gathering System and Trunk Line to Power Plant Natural gas gathered from the San Juan gas fields is collected at Chico and transported through a DN 800, 15 mm wall thickness pipeline system, 420 km long, that ties into another DN 800, 15 mm wall thickness gas transmission pipeline at Rio for eventual delivery to a power plant at Madera, as shown in Figure 8.6. Chico is at an elevation of 2100 m, whereas Rio and Madera are at 1650 m and 3100 m, respectively. The length of the pipeline from Rio to Madera is 280 km. The required delivery pressure at Madera is 35 Bar gauge. The gas gravity and viscosity are 0.65 and 0.012 cP, respectively. The gas inlet temperature at Chico is 20 °C, and the pressure is 40 Bar gauge. Assume a constant gas flow temperature of 20 °C. The pipeline MOP is 100 Bar gauge. The base temperature and base pressure are 15 °C and 1 Bar absolute, respectively. Use the Panhandle B equation with a pipeline efficiency of 95%. Assume a gas compressibility factor of 0.85 throughout.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 282 Friday, April 1, 2005 3:27 PM

282

GAS PIPELINE HYDRAULICS

1.5 Mm3/day Rio Elev: 1650 m 0 km

0B

DN

Flo

°C 20

ay

w

San Juan Chico

: mp

/d

40 Bar gauge 20°C

0

3

m

80

M

10 P:

MO

4

28

DN

g ar

6

m

k 20

ge

au

0

80

te

35 Bar gauge Madera Elev: 3100 m

Elev: 2100 m Figure 8.6

Chico–Rio pipeline to Madera power plant.

(a) Determine the compressor station power required to deliver 6 Mm3/day at Madera. Use 80% isentropic efficiency and a specific heat ratio of 1.4. (b) What modifications are required to provide gas volumes of 1.5 Mm3/day for an industrial consumer at Rio in addition to that required at Madera? (c) What pipeline capacity can be expected if all compressor stations are shut down and free flow occurs from Chico to Rio and Madera? Ignore deliveries at Rio and assume all gas flows to Madera. Solution Assume initially that one compressor station at Chico will be able to transport 6 Mm3/day to Madera. 1 Bar = 100 kPa Pipe inside diameter = 800 – 2 × 15 = 770 mm The elevation adjustment parameter from Equation 2.11 for Rio to Madera is 3100 − 1650 = 0.2589 s = 0.0684 × 0.65 (20 + 273)0.85 e s = e 0.2589 = 1.2954 The equivalent length from Equation 2.9 is Le = 280 ×

Copyright 2005 by Taylor & Francis Group, LLC

(1.2954 − 1) = 319.52 km 0.2589

2785_C008.fm Page 283 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

283

Using Panhandle B Equation 2.60, considering elevation difference, first for the Rio to Madera pipe segment, we get 15 + 273 6 × 10 6 = 1.002 × 10 −2 × 0.95 × 100

1.02

P12 − 1.2954(3600)2 × 0.961 × 293 × 319.52 × 0.85 0.65

0.51

770 2.53

Solving for the pressure at Rio, P1 = 4818 kPa Next, using this pressure as the downstream pressure for the 420 km pipe segment from Chico to Rio, we get 1650 − 2100 Elevation adjustment parameter s = 0.0684 × 0.65 = −0.0803 (20 + 273)0.85 e s = e −0.0803 = 0.9228 The equivalent length is Le = 420 ×

(0.9228 − 1) = 403.74 km −0.0803

15 + 273 6 × 10 6 = 1.002 × 10 −2 × 0.95 × 100

1.02

P12 − 0.9228 × 48182 × 0.961 × 293 × 403.74 × 0.85 0.65

0.51

770 2.53

Solving for the upstream pressure at Chico, we get P1 = 5435 kPa = 54.35 Bar absolute Since the inlet pressure at Chico is 40 Bar gauge, the compression ratio required at Chico is Compression ratio =

54.35 = 1.33 40 + 1

The compressor station power required is calculated from Equation 4.16 as follows: 0.4 1.40 1 + 0.85 1 Power = 4.0639 × 6 (293) (1.33) 1.4 − 1 0.4 2 0.8

or Power = 2455 KW

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 284 Friday, April 1, 2005 3:27 PM

284

GAS PIPELINE HYDRAULICS

When the Rio delivery of 1.5 Mm3/day is included, we calculate the upstream pressure at Chico for the 420 km segment as follows: 288 7.5 × 10 6 = 1.002 × 10 −2 × 0.95 × 100

1.02

P12 − 0.9228 × 48182 × 0.961 × 293 × 403.74 × 0.85 0.65

0.51

770 2.53

By proportion, we get 7.5 P 2 − 0.9228 × 48182 = 1 2 6.0 5435 − 0.9228 × 48182 Solving for P1, we get P1 = 5831 kPa = 58.31 Bar absolute The new compression ratio becomes Compression ratio =

58.31 = 1.42 40 + 1

The new power required at the Chico compressor station is 0.4 1.40 1 + 0.85 1 (293) Power = 4.0639 × 7.5 (1.42) 1.4 − 1 = 3808 KW 0.4 2 0.8

(c) When the compressor station at Chico is shut down, the pressure available is only 40 Bar or 4000 kPa. Using this upstream pressure and considering the entire (420 + 280) km = 700 km pipeline from Chico to Madera, the free flow capability is calculated using the Panhandle B equation by considering the elevation changes in two steps. From Equation 2.12 and Equation 2.13, for the 420 km segment the elevation falls from 2100 m to 1650 m and s = − 0.0803 and es = 0.9228 (as calculated earlier) From Equation 2.12, j1 =

0.9228 − 1 = 0.9614 − 0.0803

L1 = 420 km Similarly, for the 280 km second segment of the pipeline, the elevation rises from 2100 m to 3100 m, measured from Chico. s = 0.1785

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 285 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

285

and es = 1.1954 j2 =

1.1954 − 1 = 1.095 0.1785

L2 = 280 km From Equation 2.13, the equivalent length is Le = 0.9614 × 420 + 1.095 × 280 × 0.9228 = 686.72 km For the entire line, 3100 − 2100 = 0.1785 s = 0.0684 × 0.65 293 × 0.85 e s = e 0.1785 = 1.1954 Applying the Panhandle B equation for the entire pipeline, we get 288 Q = 1.002 × 10 −2 × 0.95 × 100

1.02

4100 2 − 1.1954 × 3600 2 0.961 × 293 × 686.72 × 0.85 0.65

51 0.5

770 2.53

Q = 1,926,314 m3/day = 1.93 Mm3/day Thus, with the Chico compressor station shut down, the free flow throughput is Q = 1.93 Mm3/day Obviously, this is inadequate to feed the Madera power plant that requires 6 Mm3/day.

Case Study 3—Fairfield to Beaumont and Travis Pipeline A natural gas pipeline, NPS 24, is being built from the gas fields at Fairfield (elevation 610 ft) to transport gas to a 400 MW power plant at Beaumont (elevation 350 ft) 280 mi away, as illustrated in Figure 8.7. Along the way at Mavis (milepost 50, elevation 1200 ft), an industrial consumer requires 10 MMSCFD, and a small community at Mayberry (milepost 110, elevation 1800 ft) requires natural gas for a municipal gas distribution system with a city gate pressure of 600 psig and 20 MMSCFD. During the first 2 years of operation, the gas flow requirements are as follows: Mavis: 10 MMSCFD at 300 psig Mayberry: 20 MMSCFD at 600 psig Beaumont: 100 MMSCFD at 400 psig Total: 130 MMSCFD

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 286 Friday, April 1, 2005 3:27 PM

286

GAS PIPELINE HYDRAULICS

400 psig D

F SC

M 0M

10

20 MMSCFD

20 m

i NPS

10 MMSCFD 0m

Mayberry 0 Elev: 1800 ft 3 S m.p. 110 NP

Beaumont Elev: 350 ft

60 MMSCFD 350 psig 16 Travis Elev: 420 ft

28

Mavis Elev: 1200 ft m.p. 50

500 psig 70°F Fairfield Elev: 610 ft Figure 8.7

Fairfield to Beaumont and Travis pipeline.

At the end of the second year, a 240 MW power plant at Travis (elevation 420 ft) will come on stream and require a gas delivery of 60 MMSCFD at 350 psig. This requires a total pipeline capacity of 190 MMSCFD out of Fairfield. The gas pressure and temperature at the inlet to the pipeline are 500 psig and 70°F. The soil temperatures can be assumed to be as follows: Fairfield to Mavis: 60°F Mavis to Mayberry: 50°F Mayberry to Beaumont: 70°F The branch pipe to Travis starts at the Travis junction (milepost 200, elevation 750 ft) and extends 20 mi to the Travis power plant. It is an NPS 16, 0.250 in. wall thickness pipe. It is anticipated that API 5LX-70 material will be used for the pipe. The cost of pipe material is $1200 per ton for pipe coated, wrapped, and delivered to the field. Construction cost of the pipeline can be estimated at $20,000 per in.-diameter mi. Compressor stations cost is $2000 per installed HP. Mainline valves are to be installed at 20 mi intervals and cost $100,000 per site. Receipt and delivery meters at Fairfield, Mavis, Mayberry, Travis, and Beaumont are expected to cost as follows: Fairfield meter: $500,000 Mavis meter: $200,000 Mayberry meter: $250,000 Travis meter: $300,000 Beaumont meter: $350,000 Fuel consumption can be estimated at 0.2 MCF per day per HP. Fuel gas cost is $4 per MCF. Assume base pressure = 14.7 psia and base temperature = 60°F. The MOP of the pipeline is 1440 psig. Use the General Flow equation with a Colebrook friction factor and the CNGA equation for the compressibility factor.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 287 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

(a) Determine the pipe wall thickness required for the specified MOP. (b) Determine the locations and HP of the compressor stations necessary for the first 2 years (phase 1) and after that (phase 2). (c) Estimate the total capital cost of pipeline, compressor stations, and other facilities for phase 2. Solution During the phase 1 operation, we will calculate the pressures and HP required, considering 100 MMSCFD delivery to the Beaumont power plant at 400 psig. Since the MOP is 1440 psig, the minimum wall thickness needed for the class 1 location is calculated from Equation 6.8: 1440 =

2t × 70, 000 × 0.72 24

Solving for t, we get Wall thickness t = 0.343 in. or Use 0.375 in. standard size pipe. Inside diameter D = 24 – 2 × 0.375 = 23.25 in. First, calculate the upstream pressure at milepost 110, assuming a gas flow temperature at 70°F and compressibility factor of 0.85. The Reynolds number from Equation 2.34 is 14.7 0.6 × 100 × 10 6 Re = 0.0004778 = 4, 357,109 520 8 × 10 −6 × 23.25 Assuming an internal roughness e = 0.0007 in. and using Equation 2.45, we get the transmission factor F as follows: 0.0007 1.255F F = −4 log10 + 3.7 × 23.25 4.3571 × 10 6 Solving by successive iteration, F = 19.45 The upstream pressure at milepost 110 is calculated from General Flow Equation 2.4, considering the elevation difference. 350 − 1800 Elevation adjustment parameter s = 0.0375 × 0.6 = −0.0724 530 × 0.85 e s = e −0.0724 = 0.9301

Copyright 2005 by Taylor & Francis Group, LLC

287

2785_C008.fm Page 288 Friday, April 1, 2005 3:27 PM

288

GAS PIPELINE HYDRAULICS

and Le = 170 ×

(0.9301 − 1) = 164.03 mi −0.0724

520 P12 − 0.9301 × 414.72 100 × 10 = 38.77 × 19.45 14.7 0.6 × 530 × 164.03 × 0.85

0.5

6

23.252.5

Solving for P1, we get P1 = 501.7 psia = 487.0 psig The average pressure in the pipe segment is, by Equation 2.14, Pavg =

501.7 × 414.7 2 501.7 + 414.7 − = 459.6 psia = 444.9 psig 501.7 + 414.7 3

We will confirm the value of the compressibility factor Z we used earlier, using CNGA Equation 1.34: Z=

1 (10)1.785× 0.6 1 + 444.9×344,4003×.825 (530)

or Z = 0.9359 This value of Z is way off compared to the 0.85 value we used in our calculations. Recalculating P1 using the recent value of Z, we obtain 520 P12 − 0.9301 × 414.72 100 × 10 6 = 38.77 × 19.45 14.7 0.6 × 530 × 164.03 × 0.9359

0.5

23.252.5

Solving for P1 by proportion, we get 501.72 − 0.9301 × 414.72 P12 − 0.9301 × 414.72 = 0.85 0.9359 or P1 = 510.86 psia Recalculating the average pressure and the new compressibility factor Z, we find Pavg =

510.86 × 414.7 2 510.86 + 414.7 − = 464.45 psia = 449.75 psig 510.86 + 414.7 3

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 289 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

289

and Z=

1

1+

449.75×344,400×(10)1.785× 0.6 (530)3.825

or Z = 0.9352 The percentage difference between this value of Z compared to the previously calculated value is 0.9352 − 0.9359 = −0.07 % 0.9359 This is good enough, and we won’t iterate any further. Therefore, Pressure at Mayberry takeoff (milepost 110) = 510.86 psia = 496.16 psig Next, calculate the pressure at milepost 50, considering the pipe segment between Mavis and Mayberry at 50°F flowing temperature and a flow rate of 120 MMSCFD. The Reynolds number, by proportion, is R = 4, 357,109 ×

120 = 5, 228, 531 100

The transmission factor F is calculated from 0.0007 1.255F F = −4 log + 3.7 × 23.25 5.2285 × 10 6 Solving by iteration, F = 19.57 The upstream pressure at milepost 50 is found from General Flow Equation 2.4, considering the elevation difference, as follows: 1800 − 1200 Elevation adjustment parameter s = 0.0375 × 0.6 = 0.0311 510 × 0.85 e s = e 0.0311 = 1.0316 and Le = 60 ×

(1.0316 − 1) = 60.96 mi 0.0311

520 P12 − 1.0316 × 510.862 120 × 10 = 38.77 × 19.57 14.7 0.6 × 510 × 60.96 × 0.85 6

Copyright 2005 by Taylor & Francis Group, LLC

0.5

23.252.5

2785_C008.fm Page 290 Friday, April 1, 2005 3:27 PM

290

GAS PIPELINE HYDRAULICS

Solving for P1, we get P1 = 562.03 psia = 547.33 psig Calculating the average pressure, Pavg =

562.03 × 510.86 2 562.03 + 510.86 − = 536.85 psia = 522.15 psig 562.03 + 510.86 3

Recalculating the compressibility factor Z using the new average pressure, we get Z=

1 1.785× 0.6

(10) 1 + 522.15×344,4003×.825 (510)

or Z = 0.9147 Next, we recalculate the pressure at milepost 50: s = 0.0311 ×

0.85 = 0.0289 0.9147

e s = 1.0293 and Le = 60 ×

0.0293 = 60.88 mi 0.0289

520 P12 − 1.0293 × 510.862 120 × 10 6 = 38.77 × 19.57 14.7 0.6 × 510 × 60.88 × 0.9147

0.5

23.252.5

Solving for P1 by proportion, we get 562.032 − 1.0316 × 510.862 P12 − 1.0293 × 510.862 = 0.85 × 60.96 0.9147 × 60.88 P1 = 564.59 psia = 549.9 psig The average pressure and Z are calculated next: Pavg =

564.59 × 510.86 2 564.59 + 510.86 − = 538.33 psia = 523.63 psig 564.59 + 510.86 3

and Z=

Copyright 2005 by Taylor & Francis Group, LLC

1 (10)1.785× 0.6 1 + 523.63×344,4003×.825 (510)

2785_C008.fm Page 291 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

291

or Z = 0.9145 This is not too far from the previously calculated Z value of 0.914. Therefore, we will not iterate any further. The pressure at milepost 50 is P1 = 564.59 psia = 549.9 psig Next, calculate the upstream pressure at Fairfield, considering the 50 mi pipe segment flowing 130 MMSCFD at 60°F. The Reynolds number from Equation 2.34 is 14.7 0.6 × 130 × 10 6 Re = 0.0004778 = 5, 664, 242 520 8 × 10 −6 × 23.25 Using Equation 2.45, we get the transmission factor F as follows: 0.0007 1.255F F = −4 log10 + 3.7 × 23.25 5.6642 × 10 6 Solving for F, we get F = 19.61 The elevation adjustment is s = 0.0375 × 0.6

1200 − 610 = 0.0284, where Z = 0.9 is assumed. 520 × 0.90 e s = 1.0288

and Le = 50 ×

0.0288 = 50.65 mi 0.0284

Using the General Flow equation, we calculate the upstream pressure P1 as 520 P12 − 1.0288 × 564.592 130 × 10 = 38.77 × 19.61 14.7 0.6 × 520 × 50.65 × 0.9 6

0.5

23.252.5

Solving for the pressure at Fairfield, we get P1 = 613.89 psia = 599.2 psig The average pressure and Z are calculated next: Pavg =

613.89 × 564.59 2 613.89 + 564.59 − = 589.58 psia = 574.88 psig 613.89 + 564.59 3

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 292 Friday, April 1, 2005 3:27 PM

292

GAS PIPELINE HYDRAULICS

and Z=

1 (10)1.785× 0.6 1 + 574.88×344,4003×.825 (520)

or Z = 0.913 Recalculating, the elevation adjustment is s=

0.0284 × 0.9 = 0.028 0.913

e s = 1.0284 and Le = 50 ×

1.0284 − 1 = 50.70 mi 0.028

Recalculating the pressure at Fairfield, using the General Flow equation, we get 520 P12 − 1.0284 × 564.592 130 × 10 6 = 38.77 × 19.61 14.7 0.6 × 520 × 50.7 × 0.913

0.5

23.252.5

By proportions, P12 − 1.0284 × 564.592 613.892 − 1.0288 × 564.592 = 50.7 × 0.913 50.65 × 0.90 or P1 = 614.40 psia = 599.7 psig Recalculating, the average pressure and Z are Pavg =

614.4 × 564.59 2 614.4 + 564.59 − = 589.85 psia = 575.15 psig 614.4 + 564.59 3

and Z=

1 (10)1.785× 0.6 1 + 575.15×344,4003×.825 (520)

or Z = 0.913, which is the same as before. Therefore, the pressure at Fairfield is P1 = 614.40 psia = 599.7 psig

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 293 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

293

The HP required is calculated from Equation 4.15: 0.4 1.4 1 + 0.913 1 614.4 1.4 1 HP = 0.0857 × 130 520 − 0.8 514.7 2 0.4

or HP = 1258 for phase 1 For phase 2, the inlet volume at Fairfield increases to 190 MMSCFD. The pressure at milepost 200 will be calculated considering the pipe segment from milepost 200 to Beaumont at 100 MMSCFD. From earlier calculations, F = 19.45 350 − 750 s = 0.0375 × 0.6 = −0.0189 530 × 0.9 e s = 0.9813 and Le = 80 ×

0.9813 − 1 = 79.15 mi −0.0189

We will assume Z = 0.9 throughout for simplicity. The pressure at milepost 200 is found from the General Flow equation 520 P12 − 0.9813 × 414.72 100 × 10 6 = 38.77 × 19.45 14.7 0.6 × 530 × 79.15 × 0.9

0.5

23.252.5

Solving for P1, we get P1 = 464.35 psia Next, calculate the pressure at milepost 110, considering 90 mi of pipe at 70°F with a flow of 160 MMSCFD. We will assume F = 19.45 and Z = 0.9 throughout for simplicity. 750 − 1800 s = 0.0375 × 0.6 = −0.0495 530 × 0.9 e s = 0.9517 and Le = 90 ×

Copyright 2005 by Taylor & Francis Group, LLC

0.9517 − 1 = 87.86 mi −0.0495

2785_C008.fm Page 294 Friday, April 1, 2005 3:27 PM

294

GAS PIPELINE HYDRAULICS

The pressure at milepost 110 is found from the General Flow equation 520 P12 − 0.9517 × 464.352 160 × 10 = 38.77 × 19.45 14.7 0.6 × 530 × 87.86 × 0.9

0.5

23.252.5

6

Solving for P1, we get P1 = 581.7 psia Next, calculate the pressure at milepost 50, considering 60 mi of pipe at 50°F with a flow of 180 MMSCFD. 1800 − 1200 s = 0.0375 × 0.6 = 0.0294 510 × 0.9 e s = 1.0298 and Le = 60 ×

1.0298 − 1 = 60.92 mi 0.0294

Therefore, the pressure at milepost 50 is found from the General Flow equation 520 P12 − 1.0298 × 581.72 180 × 10 6 = 38.77 × 19.45 14.7 0.6 × 510 × 60.92 × 0.9

0.5

23.252.5

Solving for P1, we get P1 = 678.04 psia Finally, calculate the pressure at Fairfield considering 50 mi of pipe at 60°F with a flow rate of 190 MMSCFD. 1200 − 610 s = 0.0375 × 0.6 = 0.0284 520 × 0.9 e s = 1.0288 and Le = 50 ×

Copyright 2005 by Taylor & Francis Group, LLC

1.0288 − 1 = 50.65 mi 0.0284

2785_C008.fm Page 295 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

295

Therefore, the pressure at Fairfield is found from the General Flow equation 520 P12 − 1.0288 × 678.04 2 190 × 10 6 = 38.77 × 19.45 14.7 0.6 × 520 × 50.65 × 0.9

0.5

23.252.5

Solving for P1, we get P1 = 761.04 psia The HP required at Fairfield for phase 2 is calculated from Equation 4.15: 0.4 1.4 1 + 0.9 1 761.04 1.4 1 HP = 0.0857 × 190 520 − 2 0.88 514.7 0.4

or HP = 4161 for phase 2 The capital cost is calculated next. The weight per foot of NPS 24 pipe is calculated using Equation 6.11: w = 10.68 × 0.375 × (24 – 0.375) = 94.62 lb/ft Similarly, for the Travis branch, the weight per foot of NPS 16 pipe is calculated using Equation 6.11: w = 10.68 × 0.25 × (16 – 0.25) = 42.05 lb/ft The tonnage for 280 mi of NPS 24 pipe is Tons =

94.62 × 5280 × 280 = 69, 943 2000

The tonnage for 20 mi of NPS 16 pipe is Tons =

42.05 × 5280 × 20 = 2220 2000

Total pipe cost = $1200 × (69,943 + 2220) = $86.61 million The installation cost of the pipe is calculated next: Installation cost = $20,000 × (24 × 280 + 16 × 20) = $140.8 million The installation cost of the compressor station for phase 2 is Compressor cost = $2000 × 4161 = $8.33 million

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 296 Friday, April 1, 2005 3:27 PM

296

GAS PIPELINE HYDRAULICS

Considering mainline valves at 20 mi intervals, the total number of valves required for both the main line and the Travis branch is Number of valves =

280 20 + 1+ + 1 = 17 20 20

Total cost of valves = $100,000 × 17 = $1.7 million Total cost of all meter stations = ($500 + $200 + $250 + $300 + $350) thousand = $1.6 million Therefore, Total capital cost of all facilities = ($140.8 + $8.33 + $1.7 + $1.6 + $86.61) million = $239.04 million To account for other items and indirect costs, increase the above by 30%: Total capital cost = $239.04 × 1.3 = $310.75 million.

8.3 SUMMARY This chapter reviewed some elementary concepts of transient pressures caused by valve closures and compressor station shutdown. Since the calculation methodology of transient pressures and flow rates requires the solution of partial differential equations and manual calculation is quite laborious, we refer the reader to an advanced text that specializes in this area of hydraulics. We also covered several real-world pipeline case studies. PROBLEMS 1. A natural gas pipeline system from an offshore facility is used to compress natural gas through 120 mi of NPS 24, 0.375 inch wall thickness pipe to an onshore location, similar to Figure 8.5. The compressor station located onshore is used to pump the gas through an 80 mi, NPS 20, 0.375 inch wall thickness pipe to a power plant. Determine the maximum flow rate possible under the following conditions. Neglect elevation effects. The compression ratio is 1.5. Use the Weymouth equation with 95% efficiency. Assume a base pressure of 14.7 psia and base temperature of 60°F. The gas flowing temperature is 60°F, and the compressibility factor is 0.9. The gas gravity is 0.6. a. The gas pressure at the platform equals 1440 psig and free flow occurs without use of any compression offshore or onshore. The delivery pressure at the power plant is 500 psig. b. Considering the MOP at the platform and onshore equal to 1440 psig, determine the maximum throughput possible with compression offshore and onshore.

2. Natural gas gathered from the Blanco fields is collected at Tapas and transported through a DN 600, 10 mm wall thickness pipeline system, 240 km long, that ties

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 297 Friday, April 1, 2005 3:27 PM

TRANSIENT ANALYSIS AND CASE STUDIES

297

into another DN 600, 10 mm wall thickness gas transmission pipeline at Rojas for eventual delivery to a power plant at Montecito. Tapas is at an elevation of 1500 m, whereas Rojas and Montecito are at 650 m and 300 m, respectively. The length of the pipeline from Rojas to Montecito is 140 km. The required delivery pressure at Montecito is 40 Bar gauge. The gas gravity and viscosity are 0.60 and 0.012 cP, respectively. The gas inlet temperature at Tapas is 20°C and the pressure is 40 Bar gauge. Assume a constant gas flow temperature of 20°C. The pipeline MOP is 100 Bar gauge. The base temperature and base pressure are 15°C and 1 Bar absolute, respectively. Use the Panhandle B equation with a pipeline efficiency of 95%. Assume a gas compressibility factor of 0.90 throughout. a. Determine the compressor station power required to deliver 4 Mm3/day at Montecito. Use 80% isentropic efficiency and a specific heat ratio of 1.4. b. What pipeline capacity can be expected if all compressor stations are shut down and free flow occurs from Tapas to Rojas and Montecito?

REFERENCES 1. Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003. 2. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 3. Engineering Data Book, 10th ed., Gas Processors Suppliers Association, Tulsa, OK, 1994. 4. Pipeline Design for Hydrocarbon Gases and Liquids, American Society of Civil Engineers, New York, 1975. 5. Steady Flow in Gas Pipelines, Contract Report No. 10, July 1965, Pipeline Research Council International, Houston, TX, 1965. 6. Katz, D.L. et al., Handbook of Natural Gas Engineering, McGraw-Hill, New York, 1959. 7. Wylie, E. B. and Streeter, V.L., Fluid Transients in Systems, Prentice Hall, New York, 1993.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C008.fm Page 298 Friday, April 1, 2005 3:27 PM

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 299 Friday, April 1, 2005 3:12 PM

CHAPTER

9

Valves and Flow Measurements

In this chapter we will discuss the various types of valves and flow measurements used on gas pipelines. The design and construction codes for valves, materials of construction, and application of the different types of valves and their performance characteristics will be explained. The importance of flow measurement in a gas pipeline and the accuracy of available instruments, codes, and standards used will be discussed. Various American National Standards Institute (ANSI), American Petroleum Institute (API), and American Gas Association (AGA) formulas used in connection with orifice meters will be reviewed. Since a small error in measurement in gas flow in a pipeline can translate to several thousand dollars of loss of revenue, it is important that industry strives to improve upon measurement methods. Accordingly, gas transportation companies and related industries have been researching better ways to improve flow measurement accuracy. For a detailed discussion of gas flow measurement, the reader is referred to the publications listed in the Reference section.

9.1 PURPOSE OF VALVES Valves are installed on pipelines and piping systems to isolate sections of piping for maintenance, to direct the fluid from one location to another, to shut down flow through pipe sections, and to protect pipe and prevent loss of fluid in the event of a rupture. On long-distance pipelines transporting natural gas and other compressible fluids, design codes and regulatory requirements dictate that sections of pipeline be isolated by installing mainline block valves at certain fixed spacing. For example, DOT 49 CFR, Part 192 requires that in class 1 locations, mainline valves be installed 20 mi apart. Class locations were discussed in Chapter 6.

299

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 300 Friday, April 1, 2005 3:12 PM

300

GAS PIPELINE HYDRAULICS

Grade Steel sleeve size as required

Pad or saddle

Sectionalizing or lateral control valve Figure 9.1

3 concrete cube

Mainline valve installation. (Reproduced from Katz et al., Handbook of Natural Gas Engineering, McGraw-Hill, New York, 1959. With permission.)

A typical mainline block valve installation on a gas transmission pipeline is illustrated in Figure 9.1.

9.2 TYPES OF VALVES The various types of valves used in the gas pipeline industry include the following: • • • • • • • • •

Gate valve Ball valve Plug valve Butterfly valve Globe valve Check valve Control valve Relief valve Pressure regulating valve

Each of these valves listed will be discussed in detail in the following sections. Valves can be of screwed design, welded ends, or flanged ends. In the gas industry, large valves are generally of the welded type, in which the valve is attached to the pipe on either side by a welded joint to prevent gas leakage to the atmosphere.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 301 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

Figure 9.2

301

Mainline block valve.

In smaller sizes, screwed valves are used. A typical welded end mainline valve, along with smaller valves on either side, is shown in Figure 9.2. Valves may be operated manually using a hand wheel or using an electric, pneumatic, or gas operator, as shown in Figure 9.3.

Bypass Operator Bypass ball valve

Grade

Riser Extension

Mainline valve

Figure 9.3

Valve with motor operator. (Reproduced from Nayyar, M.L., Piping Handbook, McGraw-Hill, New York, 2000. With permission.)

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 302 Friday, April 1, 2005 3:12 PM

302

GAS PIPELINE HYDRAULICS

Table 9.1 ANSI Pressure Ratings for Valves Class

Allowable Pressure, psi

150 300 400 600 900 1500

275 720 960 1440 2160 3600

9.3 MATERIAL OF CONSTRUCTION Most valves used in gas pipelines are constructed of steel and conform to specifications such as API, ASME, and ANSI standards. For certain gases that are corrosive and require certain special properties, some exotic materials can be used. The next section lists applicable standards and codes used in the design and construction of valves and fittings on gas pipelines. The valve trim material, which refers to the various working parts of a valve such as the stem, wedge, and disc, are constructed of many different materials depending upon the pressure rating and service. Valve manufacturers designate their products using some form of a proprietary numbering system. However, the purchaser of the valve must specify the type of material and operating conditions required. A typical gate valve specification might be as follows: NPS 12, ANSI 600 gate valve, cast steel flanged ends rising stem 13% CR, single wedge CS, stellite faced, seat rings SS 304, ABC company #2308. Valve operators may consist of a hand wheel or lever that is attached to the stem of the valve. Gear systems are used for larger valves. Electric motor operated valves are quite commonly used in gas pipeline systems, as are gas and pneumatic operators. Many valves can be buried, resulting in a portion of the valve and the operator above ground. The pressure rating of a valve represents the internal pressure that the valve can be subject to under normal operating conditions. For example, an ANSI 600 rating refers to a valve that can be safely operated at pressures up to 1440 psig. Most gas pipelines are operated around this pressure rating. Table 9.1 shows the ANSI pressure ratings for valves and pipes. If a valve is designated as an ANSI 600 rated valve, the manufacturer of the valve must hydrostatically test the valve at a higher pressure for a specified period of time, as required by the design code. Generally, the hydrotest pressure is 150% of the valve rating. This compares with a hydrotest pressure of 125% of MOP for pipelines, as discussed in Chapter 6. 9.4 CODES FOR DESIGN AND CONSTRUCTION The following is a list of applicable standards and codes used in the design and construction of valves and fittings on gas pipelines. ASME B31.8: Gas Transmission and Distribution Piping Systems ASME B16.3: Malleable Iron Threaded Fittings

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 303 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

303

ASME B16.5: Pipe Flanges and Flanged Fittings ASME B16.9: Factory Made Wrought Steel Butt Welding Fittings ASME B16.10: Face to Face and End to End Dimensions of Valves ASME B16.11: Forged Steel Fittings, Socket Welding and Threaded Fittings ASME B16.14: Ferrous Pipe Plugs, Bushing, etc. ASME B16.20: Metallic Gaskets ASME B16.21: Nonmetallic Gaskets ASME B16.25: Butt Welding Ends ASME B16.28: Wrought Steel, Butt Welding, Short Radius Elbows and Returns ASME B16.36: Orifice Flanges ANSI/ASTM A182: Forged or rolled alloy-steel pipe flanges, forged fittings, and valves and parts for high temperature service API 593: Ductile iron plug valves API 594: Wafer type check valves API 595: Cast iron gate valves API 597: Steel venturi gate valves API 599: Steel plug valves API 600: Steel gate valves API 602: Compact cast-steel gate valves API 603: Class 150 corrosion-resistant gate valves API 604: Ductile iron gate valves API 606: Compact carbon-steel gate valves (extended bodies) API 609: Butterfly valves to 150 psig and 150°F API 6D: Pipeline valves MSS DS-13: Corrosion resistant cast flanged valves MSS SP-25: Standard marking system for valves, fittings, and flanges

9.5 GATE VALVE A gate valve is generally used to completely shut off fluid flow or, in the fully open position, provide full flow in a pipeline. Thus, it is used either in the fully closed or fully open position. A gate valve consists of a valve body, seat, and disc; a spindle; a gland; and a wheel for operating the valve. The seat and the gate together perform the function of shutting off the flow of fluid. A typical gate valve is shown in Figure 9.4. Gate valves are generally not suitable for regulating flow or pressure or operating in a partially open condition. For this service, a plug valve or control valve should be used. It must be noted that, due to the type of construction, a gate valve requires many turns of the hand wheel to completely open or close the valve. When fully opened, gate valves offer little resistance to flow, and their equivalent length to diameter ratio (L/D) is approximately 8. The equivalent L/Ds for commonly used valves and fittings are listed in Table 9.2. This ratio represents the resistance of the valve. The gate valves used in the main lines carrying oil or gas must be of full bore or through conduit design to enable smooth passage of scrapers or pigs used for cleaning or monitoring pipelines. Such gate valves are referred to as full bore or through conduit gate valves.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 304 Friday, April 1, 2005 3:12 PM

304

Figure 9.4

GAS PIPELINE HYDRAULICS

Typical gate valve.

Table 9.2 Equivalent Lengths of Valves and Fittings Description

L/D

Gate valve Globe valve Angle valve Ball valve Plug valve straightway Plug valve 3-way thru-flo Plug valve branch flo Swing check valve Lift check valve Standard elbow–90° Standard elbow–45° Standard elbow long radius 90° Standard tee thru-flo Standard tee thru-branch Mitre bends—a = 0 Mitre bends—a = 30 Mitre bends—a = 60 Mitre bends—a = 90

8 340 55 3 18 30 90 50 600 30 16 16 20 60 2 8 25 60

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 305 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

305

Stem Seat ring Spherical plug

Body Figure 9.5

Typical ball valve. (Reproduced from Nayyar, M.L., Piping Handbook, McGrawHill, New York, 2000. With permission.)

9.6 BALL VALVE A ball valve consists of a valve body in which a large sphere with a central hole equal to the inside diameter of the pipe is mounted. As the ball is rotated, in the fully open position the valve provides the through conduit or full bore required for unrestricted flow of the fluid and scrapers or pigs. Compared to a gate valve, a ball valve has very little resistance to flow in the fully open position. When fully open, the L/D ratio for a ball valve is approximately 3.0. The ball valve, like the gate valve, is generally used in the fully open or fully closed position. A typical ball valve is shown in Figure 9.5. Unlike a gate valve, a ball valve requires one-quarter turn of the hand wheel to go from the fully open to the fully closed position. Such quick opening and closing of a ball valve can be important in some installations where isolating pipe sections quickly is needed in the event of emergency. 9.7 PLUG VALVE The plug valve traces its origin to the beginnings of the valve industry. It is a simple device for shutting off or allowing the flow of a fluid in a pipe by a simple quarter turn of the handle. In this sense, it is similar to the ball valve. Plug valves are generally used in screwed piping and in small pipe sizes. Plug valves can be hand wheel operated or operated using a wrench or gearing mechanism. The L/D ratio for this type of valve ranges from 18 to 90, depending upon the design. A typical plug valve is shown in Figure 9.6. 9.8 BUTTERFLY VALVE The butterfly valve was originally used where a tight closure was not absolutely necessary. However, over the years, this valve has been manufactured with fairly tight seals made of rubber or elastomeric materials that provide good shutoff similar

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 306 Friday, April 1, 2005 3:12 PM

306

GAS PIPELINE HYDRAULICS

Plug Body Washer Nut

Figure 9.6

Typical plug valve.

to other types of valves. Butterfly valves are used where space is limited. Unlike gate valves, butterfly valves can be used for throttling or regulating flow, as well as in the full open and fully closed positions. The pressure loss through a butterfly valve is higher in comparison with the gate valve. The L/D ratio for this type of valve is approximately 45. Butterfly valves are used in large and small sizes. They can be hand wheel operated or operated using a wrench or gearing mechanism. A typical butterfly valve is shown in Figure 9.7. 9.9 GLOBE VALVE Globe valves, so called because of their outside shape, are widely used in plant piping. They are suitable for manual and automatic operation. Unlike the gate valve, globe valves can be used for regulating flow or pressures as well as complete

Figure 9.7

Typical butterfly valve.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 307 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

Figure 9.8

307

Typical globe valve.

shutoff of flow. They can also be used as pressure relief valves or as check valves. Compared to a gate valve or ball valve, the globe valve has considerably higher pressure loss in the fully open position. This is due to the fact that the flow of fluid changes direction as it goes through the valve. The L/ D ratio for this type of valve is approximately 340. Globe valves are manufactured in sizes up to NPS 16. They are generally hand wheel operated. A typical globe valve is shown in Figure 9.8.

9.10 CHECK VALVE Check valves are normally in the closed position and are open when the fluid flows through them. They also have the capability of shutting off the flow in the event the pressure downstream exceeds the upstream pressure. In this respect, they are used for flow in one direction only. Thus, they prevent back flow through the valve. Since flow of the fluid through these valves is allowed to be in one direction only, check valves must be installed properly by noting the normal direction of flow. An arrow stamped on the outside of the valve body indicates the direction of flow. Check valves can be classified as swing check valves and lift check valves. The L/D ratios for check valves range from 50 for the swing check valve to as high as 600 for lift check valves. Examples of typical check valves are shown in Figure 9.9.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 308 Friday, April 1, 2005 3:12 PM

308

GAS PIPELINE HYDRAULICS

Flow

Flow

Swing check

Tilting disc check

Lift check Figure 9.9

Typical check valves. (Reproduced from Tullis, J.P., Hydraulics of Pipelines, John Wiley & Sons, New York, 1989. With permission.)

9.11 PRESSURE CONTROL VALVE A pressure control valve is used to automatically control the pressure at a certain point in a pipeline. In this respect it is similar to a pressure regulator discussed next. Whereas the pressure regulator is generally used to maintain a constant downstream pressure, a pressure control valve is used to control the upstream pressure. The upstream and downstream are relative to the location of the valve on the pipeline. Generally, a bypass piping system around the control valve is installed to isolate the control valve in the event of an emergency or for maintenance work on the control valve. This is illustrated in Figure 9.10. Upstream pressure

P1

Pressure drop ∆P

Downstream pressure P2

Flow Figure 9.10

Pressure control valve.

Copyright 2005 by Taylor & Francis Group, LLC

Q

2785_C009.fm Page 309 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

Figure 9.11

309

Pressure regulator. (Reproduced from Katz et al., Handbook of Natural Gas Engineering, McGraw-Hill, New York, 1959. With permission.)

9.12 PRESSURE REGULATOR A pressure regulator is a valve that is similar to a pressure control valve. Its function is to control or regulate the pressure in a certain section of a pipeline system. For example, on a lateral piping that comes off a main pipeline, used for delivering gas to a customer, a lower pressure might be required on the customer side. If the main pipeline pressure at the point of connection to the lateral pipeline is 800 psig, but the customer’s piping is limited to 600 psig, a pressure regulator is used to reduce the pressure by 200 psig, as shown in Figure 9.11. 9.13 PRESSURE RELIEF VALVE The pressure relief valve is used to protect a section of piping by relieving the pipeline pressure when it reaches a certain value. For example, if the MOP of a pipeline system is 1400 psig, a pressure relief valve may be set at 1450 psig. Any upset conditions

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 310 Friday, April 1, 2005 3:12 PM

310

GAS PIPELINE HYDRAULICS

that cause the pipeline pressure to exceed the normal 1400 psig will cause the relief valve to open at the set point of 1450 psig and expel the gas to the atmosphere or to a relief vessel, thereby protecting the pipeline from overpressure and, eventually, rupture. The difference between the relief valve set point (1450 psig) and the pipeline MOP (1400 psig) will depend on the actual application, the valve type, and expected fluctuations in pressure. Generally, the difference will range between 20 and 50 psig. Too close a difference will result in frequent operation of the relief valve, which will be a nuisance and, in many cases, a waste of valuable gas. A large difference between a relief valve set point and the pipeline MOP may render the valve ineffective.

9.14 FLOW MEASUREMENT Gas flow measurement in a pipeline is necessary for properly accounting for the amount of gas transported from one point to another along a gas pipeline. The owner of the gas and the customer who purchases the gas both require that the correct amount of gas be delivered for the agreed-upon price. Even a very small error in flow measurement on large-capacity pipelines can result in huge losses to either the owner or customer of the gas. For example, consider a gas pipeline transporting 300 MMSCFD at a tariff of 50 cents per MCF. An error of 1% in the gas flow measurement can translate to a loss of more than $500,000 per year to either the seller or the buyer. Hence, it is easy to appreciate the importance of good, accurate flow measurement in gas pipelines. Over the years, gas flow measurement technology have improved considerably. Many organizations have jointly developed standards and procedures for measurement of natural gas through orifice meters installed in pipelines. AGA, API, ANSI, and ASME have together endorsed standards for orifice metering of natural gas. The AGA Measurement Committee Report No. 3 is considered to be the leading publication in this regard. This standard is also endorsed by ANSI and API and is referred to as the ANSI/API 2530 standard. We will refer to sections of this standard when discussing orifice meters.

9.15 FLOW METERS Since the orifice meter is the main flow measurement instrument used in the gas industry, we will discuss this first. 9.15.1

Orifice Meter

The orifice meter is a flat steel plate that has a concentric machined hole with a sharp edge and is positioned inside the pipe, as shown in Figure 9.12. As the gas flows through the pipeline and then through the orifice plate, due to the reduction in cross-sectional area as the gas approaches the orifice, the velocity of flow increases and, correspondingly, the pressure drops. After the orifice, the crosssectional area increases again back to the full pipe diameter, which results in expansion of gas and decrease in flow velocity. This process of accelerating flow through

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 311 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

Pressure tap Section 1 D

D/2

311

Pressure tap Section 2 Orifice plate P2

P1

Flow

d

D

Figure 9.12

d

D

Orifice meter.

the orifice and subsequent expansion forms a vena contracta, or a throat, immediately past the orifice, as shown in the figure. Three different types of orifice meters are illustrated in Figure 9.13. The different types of orifice meters shown have different crest shapes, which affect the extent of contraction of the jet of gas as it flows through the orifice. The contraction coefficient Cc is defined in terms of the area of cross section of the vena contracta compared to the cross-sectional area of the orifice, as defined below: Cc =

Ac Ao

(9.1)

where Cc = contraction coefficient, dimensionless Ac = cross-sectional area of the vena contracta, in2 Ao = cross-sectional area of the orifice, in2

Flange P1

Pipe

Orifice P2

V

(a) Sharp-Crested Figure 9.13

Flange P1

Pipe

Orifice P2

Flange P1

Pipe

Orifice P2

V

V

(b) Round-Crested

(c) Nozzle-Crested

Different types of orifice meters. (Reproduced from Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003.)

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 312 Friday, April 1, 2005 3:12 PM

312

GAS PIPELINE HYDRAULICS

The discharge through the orifice meter is represented by the following basic equation: Q = CcCv Ao where Q = Cc = Cv = Ao = A = p1 = p2 = r = z1 = z2 = g =

2 ( p1 − p2 )/ρ + g( z1 − z2 ) 1 − Cc 2 ( Ao / A)2

(9.2)

flow rate, ft3/s contraction coefficient, dimensionless discharge coefficient, dimensionless cross-sectional area of the orifice, in2 cross-sectional area of pipe containing the orifice, in2 upstream pressure, psig downstream pressure, psig density of gas, lb/ft3 upstream elevation, ft downstream elevation, ft acceleration due to gravity

When the elevation difference between the upstream and downstream pressure taps is negligible, the discharge equation for the orifice meter can be simplified to Q = CcCv Ao

2( p1 − p2)/ρ 1 − Cc

2

( ) Ao A

2

(9.3)

where all symbols are as defined earlier. For round-crested and nozzle-crested orifice meters, shown in Figure 9.13, the value of Cc can be taken as 1.0. This indicates an absence of vena contracta for these types of orifices. For the sharp-crested orifice at high Reynolds numbers or for turbulent flow, Cc is calculated from the equation 5

A 2 Cc = 0.595 + 0.29 o A

(9.4)

where all symbols are as defined earlier. There are basically two types of pressure measurements in orifice meters. These are called flange taps and pipe taps. They relate to the locations where the pressure measurements are taken. A flange tap requires that the upstream tap be located at a distance of 1 in. upstream of the nearest plate face and that the downstream tap be located 1 in. downstream of the nearest plate face. Pipe taps are such that the upstream tap be located at a distance of 2.5 times the inside diameter of the pipe, upstream of the nearest plate face and that the downstream tap be located at a distance of 8 times the inside diameter of the pipe, downstream of the nearest plate face. Figure 9.14 illustrates the location of flange taps and pipe taps.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 313 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

313

Flange taps 1 in. (+/– 0.04 in for beta ratio < 0.6) (+/– 0.02 in for beta ratio > 0.6)

Same as downstream

Flow

8D +/– 0.1D

2 1/2D +/– 0.1D

2 1/2 D and 8D Taps (pipe taps) Figure 9.14

Flange taps and pipe taps.

Several terms used in the calculation of the orifice flow must be explained first. The differential pressure for an orifice is the pressure difference between the upstream and downstream taps. The orifice diameter is defined as the arithmetic average of four or more inside diameter measurements evenly spaced. Strict tolerances for the orifice diameters are specified in the AGA3/ANSI 2530 standard. Table 9.3 shows these tolerances taken from the standard. 9.15.1.1

Meter Tube

The meter tube is the piece of pipe in which the orifice plate is installed, along with straightening vanes as needed. A typical meter tube consisting of the orifice plate and straightening vanes is illustrated in Figure 9.15. The dimensions of the meter tube, such as A, B, C, and C′, depend upon the orifice to pipe diameter ratio, also known as beta ratio b, and are specified in AGA Report No. 3. For example, for beta = 0.5, A = 25

A′ = 10

B=4

C=5

C′ = 5.5

Table 9.3 Orifice Plate Diameter Tolerances

Copyright 2005 by Taylor & Francis Group, LLC

Orifice Diameter, in.

Tolerance /, in.

0.250 0.375 0.500 0.625 0.750 0.875 1.000 above 1.000

0.0003 0.0004 0.0005 0.0005 0.0005 0.0005 0.0005 0.0005 (per inch dia)

2785_C009.fm Page 314 Friday, April 1, 2005 3:12 PM

314

GAS PIPELINE HYDRAULICS

Meter tube A′ B C

C′

Straightening vanes Figure 9.15

Orifice

Meter tube installation.

These numbers are actually multiples of the pipe or meter tube diameter. The requirements of straightening vanes before the orifice plate depend on the specific installation. The main reason for straightening vanes is to reduce flow disturbance at the orifice plate from upstream fittings. Refer to AGA Report No. 3 for various meter tube configurations. The orifice flow rate is the mass flow rate or volume flow rate of gas per unit of time. The density is the mass per unit volume of gas at a specific temperature and pressure. 9.15.1.2

Expansion Factor

The expansion factor is a dimensionless factor used to correct the calculated flow rate to take into account the reduction in gas density as it flows through an orifice, which is caused by the increased velocity and corresponding reduced static pressure. Methods of calculating the expansion factor Y will be discussed in subsequent sections. The beta ratio is defined as the ratio of the orifice diameter to the meter tube diameter, as follows:

β=

d D

(9.5)

For orifice meters with flange taps, the beta ratio ranges between 0.15 and 0.70. For orifice meters with pipe taps, the beta ratio ranges between 0.20 and 0.67, where b = beta ratio, dimensionless d = orifice diameter, in. D = meter tube diameter, in. The fundamental orifice meter flow equation described in ANSI 2530/AGA Report No. 3 is as follows: qm =

Copyright 2005 by Taylor & Francis Group, LLC

π C Y d 2 (2 gρ f ∆P)0.5 (1 − β 4 )0.5 4

(9.6)

2785_C009.fm Page 315 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

315

or qm = KY

where qm rf C b d D Y g ∆P K

= = = = = = = = = =

π 2 d (2 gρ f ∆P)0.5 4

(9.7)

β=

d D

(9.8)

K=

C CD 2 = (1 − β 4 )0.5 ( D 4 − d 4 )0.5

(9.9)

mass flow rate of gas, lb/s density of gas, lb/ ft3 discharge coefficient beta ratio, dimensionless orifice diameter, in. meter tube diameter, in. expansion factor, dimensionless acceleration due to gravity, ft/s2 pressure drop across the orifice, psi flow coefficient, dimensionless

These equations were arrived at using the conservation of energy and mass equations with thermodynamics and the equation of state for the gas in question. It can be seen that, essentially, these formulas give the mass flow rate of gas. We need to convert these to the volume flow rate using the density. The coefficient of discharge C in the preceding equation is approximately 0.6, and the flow coefficient K is a value that is between 0.6 and 0.7. Both the flow coefficient K and the expansion factor Y are determined using test data. The volume flow rate at standard (base) conditions is calculated from the mass flow rate as follows: qv =

qm ρb

(9.10)

where q v = volume flow rate, ft3/s qm = mass flow rate, lb/s r b = gas density at base temperature, lb/ft3 The expansion factor Y for low-compressibility fluids, such as water at 60°F and 1 atmosphere pressure, is taken as 1.0. For gases, Y can be calculated as explained in the next section. The flow coefficient K is found to vary with the diameter of the meter tube D, orifice diameter d, mass flow rate qm, and fluid density and viscosity at the flowing temperature. For gases, K also varies with the ratio of

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 316 Friday, April 1, 2005 3:12 PM

316

GAS PIPELINE HYDRAULICS

the differential pressure to the static pressure and k, the ratio of specific heat of the gas. In many cases, the flow coefficient K is considered to be a function of the Reynolds number, acoustic ratio, meter tube diameter, and beta ratio. Rearranging Equation 9.7, we get KY =

4 qm

(9.11)

π d [2 gc ρ f ∆P]0.5 2

Several empirical equations are available to calculate the flow coefficient K. The following equation by Buckingham and Bean is endorsed by the National Bureau of Standards (NBS) and is listed in AGA Report No.3. For flange taps, 1 0.5 0.007 0.076 Ke = 0.5993 + + 0.364 + 0.5 β 4 + 0.4 1.6 − 0.07 + − β D D D D 5

65 0.034 − 0.009 + (0.5 − β )1.5 + 2 + 3 (β − 0.7)2.5 D D where Ke = D = d = b =

2.5

(9.12)

flow coefficient for Reynolds number Rd = d(106/15), dimensionless meter tube diameter, in. orifice diameter, in. beta ratio, dimensionless

For pipe taps, Ke = 0.5925 +

0.0182 0.06 2 0.225 5 + 0.440 − β + 0.935 + β D D D

+ 1.35β 14 +

1.43 (0.25 − β )2.5 D 0.5

(9.13)

where all symbols are as defined before. For flange taps and pipe taps, the value of Ko is calculated from Ko =

Ke

1+

15×10−6 E d

(9.14)

where the parameter E in Equation 9.14 is found from E = d (830 − 5000 β + 9000 β 2 − 4200 β 3 + B)

Copyright 2005 by Taylor & Francis Group, LLC

(9.15)

2785_C009.fm Page 317 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

317

The value of parameter B in Equation 9.15 is defined as follows: For flange taps, B = For pipe taps, B =

530 D 0.5

(9.16)

875 + 75 D

(9.17)

Finally, the flow coefficient K is calculated from E K = Ko 1 + Rd

(9.18)

where Ko = flow coefficient for infinitely large orifice Reynolds numbers, dimensionless Rd = Reynolds number at the inlet of orifice, dimensionless The Reynolds number used in the preceding equations is calculated from Rd = where Rd = Vf = d = rf = m =

Vf d ρ f µ

(9.19)

Reynolds number at the inlet of orifice, dimensionless velocity of fluid at inlet of orifice, ft/s orifice diameter, ft fluid density at flowing conditions, lb/ft3 dynamic viscosity of fluid, lb/ft.s

The values of flow coefficient K calculated using the preceding equations apply to orifice meters manufactured and installed in accordance with AGA Report No. 3, as long as the meter tube is greater than 1.6 in. inside diameter and the beta ratio is between 0.10 and 0.75. The uncertainties in flow coefficient K, in accordance with AGA Report No. 3, follow: For For For For

flange taps, the uncertainty is +/−0.5% for 0.15 < b < 0.70 flange taps, the uncertainty is greater than +/−1.0% for 0.15 > b > 0.70 pipe taps, the uncertainty is +/−0.75% for 0.20 < b < 0.67 pipe taps, the uncertainty is greater than +/−1.5% for 0.20 > b > 0.67

The expansion factor Y is calculated in two ways. In the first method, it is calculated using the upstream pressure, and in the second method, it is calculated using the downstream pressure. The following equation is used for the expansion factor Y1 with reference to upstream pressure.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 318 Friday, April 1, 2005 3:12 PM

318

GAS PIPELINE HYDRAULICS

For flange taps, Y1 = 1 − (0.41 + 0.35β 4 )

x1 k

(9.20)

For pipe taps, Y1 = 1 − [0.333 + 1.145(β 2 + 0.7β 5 + 12β 13 )]

x1 k

(9.21)

and the pressure ratio x1 is x1 =

Pf 1 − Pf 2 hw = 27.707 Pf 1 Pf 1

(9.22)

where Y1 = expansion factor based on upstream pressure x1 = ratio of differential pressure to absolute upstream static pressure hw = differential pressure between upstream and downstream taps in in. of water at 60°F Pf1 = static pressure at upstream tap, psia Pf2 = static pressure at downstream tap, psia x1/k = acoustic ratio, dimensionless k = ratio of specific heats of gas, dimensionless The value of Y 1 calculated using these equations is subject to a tolerance from 0 to +/− 0.5% for the range of x = 0 to 0.20. For larger values of x, the uncertainty is larger. For flange taps, the values of Y1 are valid for a beta ratio range of 0.10 to 0.80. For pipe taps, the beta ratio range is 0.10 to 0.70. With reference to the downstream pressure, the expansion factor Y2 is calculated using the following equations. For flange taps, 1 Y2 = Y1 1 − x1

0.5

Y2 = (1 + x 2 )0.5 − (0.41 + 0.35β 4 )

(9.23) x2 k (1 + x 2 )0.5

(9.24)

For pipe taps, Y2 = (1 + x 2 )0.5 − [0.333 + 1.145(β 2 + 0.7β 5 + 12β 13 )]

Copyright 2005 by Taylor & Francis Group, LLC

x2 k (1 + x 2 )0.5

(9.25)

2785_C009.fm Page 319 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

319

and the pressure ratio x2 is x2 =

Pf 1 − Pf 2 hw = Pf 2 27.707 Pf 2

(9.26)

where all symbols are as defined before. The density of the flowing gas used in Equation 9.6 must be obtained from the equation of state or from tables. It is important to use the correct density in the flow equations. Otherwise, the uncertainty in flow measurement could be as great as 10%. Generally, the density of the gas can be calculated from the perfect gas law discussed in Chapter 1, with the modification using the compressibility factor. The following equation is obtained by rearranging the real gas equation and using the gravity of gas (see Chapter 1 for details):

ρf =

m Gi MPf = V Z f RT f

(9.27)

ρf1 =

Gi MPf 1 Z f 1RT f

(9.28)

ρb =

MGi Pb RZ bTb

(9.29)

where m = mass of gas V = volume of gas Gi = gravity of gas (air = 1.00) M = molecular weight of gas Pf = absolute gas pressure Zf = compressibility factor at flowing temperature R = gas constant Tf = absolute flowing temperature subscript f 1 refers to upstream tap flowing conditions subscript f 2 refers to downstream tap flowing conditions subscript b refers to base conditions Two other equations, based on real gas specific gravity and taking the base conditions of 14.73 psia and 60°F, result in the gas densities at the upstream tap and at the base conditions as follows:

ρf1 =

MZ bGPf 1 0.99949 RZ f 1T f

(9.30)

MGPb 0.99949 RTb

(9.31)

ρb =

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 320 Friday, April 1, 2005 3:12 PM

320

GAS PIPELINE HYDRAULICS

Knowing the densities at the upstream tap and at the base condition, the following equation is used for the volume flow rate. This equation is derived from the equations listed in the preceding sections. qv =

π 4

(

2 g KY1d 2

)

( ρ f 1∆P)

ρb

(9.32)

Combining all equations we have reviewed so far, AGA Report No. 3 shows a compact equation for the flow of gas through an orifice meter as follows: Qv = C hw Pf

(9.33)

where Qv = gas flow rate at base conditions, ft3/h hw = differential pressure between upstream and downstream taps in in. of water at 60°F Pf = absolute static pressure, psia C = orifice flow constant For Pf , subscript 1 is used for upstream and subscript 2 for downstream pressure. The orifice flow constant C consists of the product of several factors that depend on the Reynolds number, expansion factor, base pressure, base temperature, flowing temperature, gas gravity, and supercompressibility factor of gas. It is given by the following equation: C = Fb Fr Fpb Ftb Ftf Fgr FpvY

(9.34)

where the dimensionless factors are Fb = basic orifice factor Fr = Reynolds number factor Fpb = pressure base factor Ftb = temperature base factor Ftf = flowing temperature factor Fgr = gas relative density factor Fpv = supercompressibility factor Y = expansion factor These values of the factors that constitute the orifice flow constant C are defined in AGA Report No. 3 and are listed in Appendix B of that publication. However, each of these factors can be calculated as follows. The basic orifice factor is Fb = 338.178d 2 K o where Ko is calculated using Equation 9.14.

Copyright 2005 by Taylor & Francis Group, LLC

(9.35)

2785_C009.fm Page 321 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

321

The Reynolds number factor is Fr = 1 +

E Rd

(9.36)

K = K o Fr

(9.37)

The pressure base factor is 14.73 Pb

(9.38)

Tb 519.67

(9.39)

Fpb = The temperature base factor is Ftb = The flowing temperature factor is

519.67 Ftf = Tf

0.5

(9.40)

The gas relative density factor is 1 Fgr = Gr

0.5

(9.41)

where all symbols in the preceding equations are as defined before.

9.16 VENTURI METER The venturi meter, shown in Figure 9.16, is based upon Bernoulli’s equation. It consists of a smooth gradual contraction from the main pipe size to a reduced section known as the throat, finally expanding back gradually to the original pipe diameter. This type of venturi meter is called the Herschel type. The angle of contraction from the main pipe to the throat section is in the range of 21° +/− 2°. The gradual expansion from the throat to the main pipe section is in the range of 5 to 15°. This design causes the least energy loss such that the discharge coefficient can be assumed at 1.0. Venturi meters range in size from 4.0 in. to 48.0 in. The beta ratio, equal to d/D, generally ranges between 0.30 and 0.75. The gas pressure in the main pipe section is represented by P1 and that at the throat is represented by P2. As gas flows through a venturi meter, it increases in flow velocity in the narrow throat section. Correspondingly, the pressure reduces in the throat section according to Bernoulli’s equation. After gas leaves the throat section,

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 322 Friday, April 1, 2005 3:12 PM

322

GAS PIPELINE HYDRAULICS

Pipe section 1 Throat section 2 P1

Pipe section 3

P2

D

Flow

d

D

h Manometer

Figure 9.16

Venturi meter.

it reduces in flow velocity due to the increase in pipe cross-sectional area, and it reaches the original flow velocity. The flow velocity in the main pipe section before the throat is calculated from the known pressures P1 and P2:

V1 =

2 g ( P1− P2 ) + ( Z1 − Z 2 ) − hL ρ

( ) −1 2

A1 A2

(9.42)

Neglecting the elevation difference Z1 – Z2 and the friction loss hL, this equation reduces to the following:

V1 = C

where V1 = r = A1 = A2 = C =

2 g ( P1− P2 ) ρ

( ) −1 A1 A2

2

(9.43)

velocity of gas in the main pipe section before the throat the average gas density cross-sectional area of the pipe cross-sectional area of the throat discharge coefficient, dimensionless

The volume flow rate is then calculated by multiplying the velocity by the crosssectional area, resulting in the following equation:

Q = CA1

Copyright 2005 by Taylor & Francis Group, LLC

2 g ( P1− P2 ) ρ

( ) −1 A1 A2

2

(9.44)

2785_C009.fm Page 323 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

323

Using the beta ratio, we simplify the above equation as follows:

Q = CA1

2 g ( P1− P2 ) ρ

(9.45)

( ) −1 1 β

4

The discharge coefficient C is a number less than 1.0, and it depends on the Reynolds number in the main pipe section. For a Reynolds number greater than 2 × 105, the value of C remains constant at 0.984. In smaller pipe sizes, such as 2 to 10 in., venturi meters are machined and, therefore, have a better surface finish than the larger rough cast meters. Smaller venturi meters have a C value of 0.995 for Reynolds numbers larger than 2 × 105. 9.17 FLOW NOZZLE The flow nozzle shown in Figure 9.17 is another device for measuring flow rate. It consists of a main pipe section, followed by a gradual reduction in cross-section area and a short cylindrical section, ending in a gradual expansion to the original pipe size. The discharge coefficient C for a flow nozzle is approximately 0.99 for Reynolds numbers greater than 106. At lower Reynolds numbers, due to greater energy loss subsequent to the nozzle throat, C values are lower. The discharge coefficient C depends on the beta ratio and Reynolds number. It is calculated using the following equation: C = 0.9975 − 6.53

Pressure tap Section 1 D

D/2

P1

P2

To manometer Flow nozzle.

Copyright 2005 by Taylor & Francis Group, LLC

(9.46)

Pressure tap Section 2

d Flow

D

Figure 9.17

β R

2785_C009.fm Page 324 Friday, April 1, 2005 3:12 PM

324

GAS PIPELINE HYDRAULICS

where b = d/D R = Reynolds number based on the pipe diameter D Example 1 An orifice meter with 4 in. diameter is installed in a pipe with an inside diameter of 12.09 in. The differential pressure is measured at 30 in. of water, and the static pressure upstream is 600 psig. The gas gravity = 0.6 and the gas flowing temperature = 70°F. The base temperature and the base pressure are 60°F and 14.7 psia, respectively. Assuming flange taps, calculate the flow rate in standard ft3/h. The barometric pressure is 14.5 psia. Solution The basic orifice factor F b is calculated from the AGA 3 appendix as follows:

Fb = 3258.5 (hP)0.5 = [30 × (600 + 14.5)]0.5 = 135.78 Fr = 1 +

0.0207 = 1.0002 135.78

Fpb =

14.73 = 1.002 14.7

Ftb =

60 + 460 = 1.006 519.67

519.67 Ftf = 70 + 460 1 Fgr = 0.6

0.5

= 0.9902

0.5

= 1.291

Fpv = 1.0463 30 h = = 0.0488 P 614.5

β=

4 = 0.3309 12.09

Y = 0.9995 C = 3258.5 × 1.0002 × 1.002 × 1.006 × 0.9902 × 1.291 × 1.00463 × 0.9995 = 4391.96 Using Equation 9.33, the flow rate is Qv = 4391.96 × 135.78 = 596, 340 ft 3 /h

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 325 Friday, April 1, 2005 3:12 PM

VALVES AND FLOW MEASUREMENTS

325

9.18 SUMMARY In this chapter we covered the topics of valves and flow measurement as they relate to gas pipeline transportation. The various types of valves used and their functions were reviewed. The importance of flow measurement in natural gas pipeline transaction was explained. The predominantly used measuring device known as an orifice meter was discussed in detail. The calculation methodology based on AGA Report No. 3 was reviewed. The venturi meter and the flow nozzle were also discussed.

PROBLEMS 1. An orifice meter 2 in. in diameter is installed in a pipe with an inside diameter of 12.09 in. The differential pressure is measured at 20 in. of water, and the static pressure upstream is 500 psig. The gas gravity = 0.65 and gas flowing temperature = 75°F. The base temperature and the base pressure are 60°F and 14.7 psia, respectively. Assuming pipe taps, calculate the flow rate in standard ft3/h. The barometric pressure is 14.6 psia. 2. An orifice meter has a bore size of 1 in. diameter and is installed in a pipe with an inside diameter of 6.125 in. The differential pressure is measured at 10 in. of water, and the static pressure upstream is 300 psig. The gas gravity = 0.6 and gas flowing temperature = 70°F. The base temperature and the base pressure are 60°F and 14.7 psia, respectively. Assuming flange taps, calculate the flow rate in standard ft3/h. The barometric pressure is 14.6 psia.

REFERENCES 1. Miller, R.W., Flow Measurement Engineering Handbook, McGraw-Hill, New York, 1983. 2. Upp, E.L., Fluid Flow Measurement, Gulf Publishing Company, Houston, TX, 1993. 3. Flow of Fluids through Valves, Fittings and Pipe, Crane Company, New York, 1976. 4. Cheremisinoff, N., Applied Fluid Flow Measurement, Marcel Dekker, Inc., New York, 1979. 5. Nayyar, M.L., Piping Handbook, 7th ed., McGraw-Hill, New York, 2000. 6. Liu, H., Pipeline Engineering, CRC Press, Boca Raton, FL, 2003. 7. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 8. Engineering Data Book, 10th ed., Gas Processors Suppliers Association, Tulsa, OK, 1994. 9. Steady Flow in Gas Pipelines, Contract Report No. 10, July 1965, Pipeline Research Council International, Houston, TX, 1965. 10. Katz, D.L. et al., Handbook of Natural Gas Engineering, McGraw-Hill, New York, 1959. 11. Orifice Metering of Natural Gas, AGA Report No. 3, ANSI/API 2530, American Gas Association, Arlington, VA, June 1987. 12. Mendel, O., Practical Piping Handbook, PennWell Books, Tulsa, OK, 1981.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C009.fm Page 326 Friday, April 1, 2005 3:12 PM

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 327 Friday, April 1, 2005 3:32 PM

CHAPTER

10

Pipeline Economics In the previous chapters we explored different scenarios of pipe sizes and pressures to transport natural gas through pipelines from one location to another. Various pressure drop formulas, compression requirements, and HP required were calculated without delving too much into costs of facilities. In this chapter the economic aspects of pipelines will be reviewed. The economic pipe size required for a particular throughput will be arrived at considering the various costs that make up a pipeline system. The initial capital cost of pipeline and ancillary facilities will be discussed, along with the annual operating and maintenance costs. Since pipelines are generally designed to transport gas belonging to one company by another company, a methodology for determining transportation cost or tariff will be analyzed. A pipeline can be constructed to transport natural gas for the owner of the pipeline, to sell gas to another company, or to transport some other company’s gas. These three scenarios represent three major uses of pipeline transportation of natural gas. The economics involved in the selection of pipe diameter, compressor station, and related facilities will vary slightly for each scenario. As an owner company transporting its own gas, minimal facilities will probably be built. However, Department of Transportation (DOT) codes and other regulatory requirements will still have to be met to ensure a safe pipeline operation that will not endanger humans or the environment. In the second scenario, in which a company builds a pipeline to transport its gas and sells the gas at the end of the pipeline to a customer, minimal facilities will be constructed without too much regulatory control. In the third scenario, a pipeline company constructs and operates a pipeline for the purpose of transporting gas belonging to other companies. This will be under the jurisdiction of the Federal Energy Regulatory Commission (FERC) or a state agency such as the Public Utilities Commission (PUC) in California or the Texas Railroad Commission in Texas. An interstate pipeline in which the pipeline crosses one or more state boundaries will be regulated by the FERC. A pipeline that is intrastate, such as wholly within California, will be subject to PUC rules and not FERC. Such regulatory requirements impose strict guidelines on the type and number of facilities and costs that may be passed on to the customer requesting gas transportation. These regulatory requirements will dictate that excessive capital facilities not be built and the amortized cost passed on 327

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 328 Friday, April 1, 2005 3:32 PM

328

GAS PIPELINE HYDRAULICS

to the customers. Whereas a private pipeline company transporting its own gas may build in extra compressor units as spares to ensure uninterrupted operation in the event of equipment failure, FERC-regulated pipelines may not be able to do so. Thus, pipeline economics will differ slightly from case to case. In this chapter we will not discuss other modes of transportation of gas, such as truck transport of pressurized gas containers. The general economic principles discussed here are applicable to private unregulated pipelines as well as FERC-regulated pipelines used for interstate transportation of natural gas. 10.1 COMPONENTS OF COST In a gas pipeline system the major components that contribute to the initial capital cost are the pipeline, compressor stations, mainline valve stations and metering facilities, telecommunications, and supervisory control and data acquisition (SCADA). Other costs include environmental and permitting costs, right of way (ROW), acquisition cost, engineering and construction management, legal and regulatory costs, contingency, and allowance for funds used during construction (AFUDC). The recurring annual costs will include operating and maintenance (O&M) costs, fuel, energy and utility costs, rental, permitting, and annual right of way costs. The O&M costs will include payroll and general and administrative (G&A) costs. In any pipeline system constructed to provide transportation of gas, there will be capital costs and annual operating costs. If we decide on a useful life of the pipeline (say, 30 or 40 years) we can annualize all costs and also determine the revenue stream necessary to amortize the total investment in the pipeline project. The revenue earned after expenses and taxes plus a percentage for proﬁt divided by the volume transported will give the transportation tariff necessary. The calculation of capital cost, operating cost, and transportation tariff will be illustrated using an example. Throughout this chapter we will need to convert annual cash ﬂows or expenses into present value and vice versa. A useful equation relating the present value of a series of annual payments over a number of years at a speciﬁed interest rate is as follows: PV = where PV R i n

= = = =

R 1 1− i (1 + i)n

(10.1)

present value, $ series of cash ﬂows, $ interest rate, decimal value number of periods, years

For example, $10,000 in annual payments for 20 years at an annual interest rate of 10% results in a present value of PV =

Copyright 2005 by Taylor & Francis Group, LLC

10, 000 1 = $85,136 1− 20 0.10 (1 + 0.10)

2785_C010.fm Page 329 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

329

Similarly, we can convert a present value of $10 million into an annualized cost based on 8% interest for 30 years as follows. From Equation 10.1, 10, 000, 000 =

R 1 1− 0.08 (1.08)30

Solving for the annual cost R, we get R = $888,274 Next, we will calculate the cost of service and transportation tariff using a simple example. Example 1 A natural gas pipeline transports 100 MMSCFD at a load factor of 95%. The capital cost is estimated at $60 million and the annual operating cost is $5 million. Amortizing the capital at 10% for a project life of 25 years, calculate the cost of service and transportation tariff for this pipeline. Solution All costs will be converted to annualized values for a 25-year project life and 10% interest rate. This will be the cost of service on an annual basis. When this cost is divided by the annual pipeline throughput, we obtain the transportation tariff. The capital cost of $60 million is ﬁrst converted to annual cash ﬂow at a 10% interest rate for a period of 25 years. Using Equation 10.1, Annualized capital cost =

60 × 0.10 = $6.61 million 1 − 1 25 (1.10)

This assumes zero salvage value at the end of the 25-year useful life of the pipeline. Therefore, for a project life of 25 years and a discount rate of 10%, the capital cost of $60 million is equivalent to annual cost of $6.61 million. Adding the annual operating cost of $5 million, the total annual cost is $6.61 + $5 = $11.61 million per year. This annual cost is deﬁned as the cost of service incurred each year. Actually, to be accurate, we should take into account several other factors such as the tax rate, depreciation of assets, and proﬁt margin to arrive at a true cost of service. The transportation tariff is deﬁned as the cost of service divided by the annual volume transported. At a 95% load factor and ﬂow rate of 100 MMSCFD, the transportation tariff is Tariff =

Copyright 2005 by Taylor & Francis Group, LLC

$11.61 × 10 6 × 10 3 = $0.3348 per MCF 100 × 10 6 × 365 × 0.95

2785_C010.fm Page 330 Friday, April 1, 2005 3:32 PM

330

GAS PIPELINE HYDRAULICS

In other words, for this pipeline, every MCF of gas transported requires a payment of approximately 33.5 cents to the pipeline owner that provides the transportation. This is a very rough and simplistic calculation of an example of tariff. In reality, we must take into account many other factors to arrive at an accurate cost of service. For example, the annual operating cost will vary from year to year over the life of the pipeline, due to inﬂation and other reasons. Taxes, depreciation of assets, and salvage value at the end of the life of the pipeline must also be considered. Nevertheless, the preceding analysis gives a quick overview of the approach used to calculate a rough value of the transportation cost.

10.2 CAPITAL COSTS The capital cost of a pipeline project consists of the following major components: • • • • • • • • •

Pipeline Compressor stations Mainline valve stations Meter stations Pressure regulator stations SCADA and telecommunication Environmental and permitting Right of way acquisitions Engineering and construction management

In addition, there are other costs such as allowance for funds used during construction (AFUDC) and contingency. Each of the preceding major categories of capital cost will be discussed next. 10.2.1

Pipeline

The pipeline cost consists of those costs associated with the pipe material, coating, pipe ﬁttings, and the actual installation or labor cost. In Chapter 6, we introduced a simple formula to calculate the weight of pipe per unit length. From this and the pipe length, the total tonnage of pipe can be calculated. Given the cost per ton of pipe material, the total pipe material cost can be calculated. Knowing the construction cost per unit length of pipe, we can also calculate the labor cost for installing the pipeline. The sum of these two costs is the pipeline capital cost. Using Equation 6.11 for pipe weight, the cost of pipe required for a given pipeline length is found from PMC =

10.68( D − T )TLC × 5280 2000

where PMC = pipe material cost, $ L = length of pipe, mi D = pipe outside diameter, in.

Copyright 2005 by Taylor & Francis Group, LLC

(10.2)

2785_C010.fm Page 331 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

331

= pipe wall thickness, in. = pipe material cost, $/ton

T C

In SI units, PMC = 0.0246( D − T )TLC where PMC L D T C

= = = = =

(10.3)

pipe material cost, $ length of pipe, km pipe outside diameter, mm pipe wall thickness, mm pipe material cost, $/metric ton

Generally, pipe will be supplied externally coated and wrapped. Therefore, we must add this cost or a percentage to the bare pipe cost to account for the extra cost and the delivery cost to the construction site. In the absence of actual cost, we may increase the bare pipe cost by a small percentage, such as 5%. For example, using Equation 10.2 for a 100 mi pipeline, NPS 20 with 0.500 in. wall thickness, the total pipe cost, based on $800 per ton, is PMC =

10.68(20 − 0.5)0.5 × 100 × 800 × 5280 = $21.99 million 2000

If the pipe is externally coated and wrapped and delivered to the ﬁeld at an extra cost of $5 per ft, this cost can be added to the bare pipe cost as follows: Pipe coating and wrapping cost = $5 × 5280 × 100 = $2.64 million Therefore, the total pipe cost becomes $21.99 + $2.64 = $24.63 million The labor cost to install the pipeline can be represented in dollars per unit length of pipe. For example, the labor cost might be $60 per ft or $316,800 per mi of pipe for a particular size pipe in a certain construction environment. This number will depend on whether the pipeline is installed in open country, ﬁelds, or city streets. Such numbers are generally obtained from contractors who will take into consideration the difﬁculty of trenching, installing pipe, and backﬁlling in the area of construction. For estimation purposes, there is a wealth of historical data available for construction cost for various pipe sizes. Sometimes the pipe installation cost is expressed in terms of dollars per in. diameter per mi of pipe. For example, an NPS 16 pipe might have an installation cost of $15,000 per in.-diameter-mile. Thus, if 20 mi of NPS 16 pipe are to be installed, we estimate the labor cost as follows: Pipe installation cost = $15,000 × 16 × 20 = $4.8 million

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 332 Friday, April 1, 2005 3:32 PM

332

GAS PIPELINE HYDRAULICS

Table 10.1

Typical Pipeline Installation Costs

Pipe Diameter, in.

Average Cost, $/in.-dia/mi

8 10 12 16 20 24 30 36

18,000 20,000 22,000 14,900 20,100 33,950 34,600 40,750

If we convert this cost on a unit length basis, we get Pipe installation cost =

4.8 × 10 6 = $45.45 per ft 20 × 5280

Table 10.1 shows typical installation costs for pipelines. These numbers must be veriﬁed by discussions with construction contractors who are familiar with the construction location. Several other construction costs must be added to the installation costs for straight pipe. These expeditures include costs for road, highway, and railroad crossings and stream and river crossings. These costs can be provided as lump sum numbers, which can be added to the pipeline installation costs to come up with a total pipeline construction cost. For example, a pipeline might include two road and highway crossings that total $300,000 in addition to a couple of river crossings costing $1 million. Compared to the installation cost of a long-distance pipeline, the road and river crossings total might be a small percentage. 10.2.2

Compressor Stations

In order to provide transportation of gas through a pipeline, we have to install one or more compressor stations to provide the necessary gas pressure. Once we decide on the details of the compressor station equipment and piping, a detailed bill of materials can be developed from the engineering drawings. Based upon quotations from equipment vendors, a detailed cost estimate of the compressor stations can be developed. In the absence of vendor data and in situations where a rough order of magnitude costs for compressor stations is desired, we can use an all-inclusive price of dollars per installed HP. For example, using an installed cost of $2000 per HP, for a 5000 HP compressor station, the capital cost will be estimated as follows: Compressor station cost = 2000 × 5000 = $10 million In the above calculations, the all-inclusive number of $2000 per installed HP is expected to include material and equipment cost and the labor cost for installing the compressor equipment, piping, valves, instrumentation, and controls within the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 333 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

333

compressor stations. Generally, the $/HP number decreases as the size of the compressor HP increases. Thus, a 5000 HP compressor station might be estimated on the basis of $2000 per HP, whereas a 20,000 HP compressor station will be estimated at an installed cost of $1500 per HP. These numbers are mentioned for illustration purposes only. Actual $/HP values must be obtained from historical pipeline cost data and in consultation with compressor station construction contractors and compressor station equipment vendors. Generally, the pipeline and compressor station costs constitute the bulk of the total pipeline project cost. 10.2.3

Mainline Valve Stations

Mainline block valves are installed to isolate sections of a pipeline for safety reasons and maintenance and repair. In the event of a pipeline rupture, the damaged pipeline section can be isolated by closing off the mainline valves on either side of the rupture location. For mainline valve stations installed at speciﬁed intervals along the pipeline, the cost of facilities can be speciﬁed as a lump sum ﬁgure that includes the mainline valve and operator, blowdown valves and piping, and other pipe and ﬁttings that constitute the entire block valve installation. Generally, a lump sum ﬁgure can be obtained for a typical mainline block valve installation from a construction contractor. For example, an NPS 16 mainline valve installation might be estimated at $100,000 per site. In a 100 mi, NPS 16 pipeline, DOT code requirements might dictate that a mainline valve be installed every 20 mi. Therefore, in this case there would be six mainline valves for a 100 mi pipeline. At $100,000 per site, the total installed cost of all mainline valve stations will be $600,000. This will be added to the capital cost of the pipeline facilities. 10.2.4

Meter Stations and Regulators

Meter stations are installed for measuring the gas ﬂow rate through the pipeline. These meter stations will consist of meters, valves, ﬁttings, instrumentation, and controls. Meter stations can also be estimated as a ﬁxed price, including material and labor for a particular site. For example, a 10 in. meter station might cost $300,000 lump sum. If there are four such meter stations on a 100 mi gas pipeline, the total meter station cost will be $1.2 million. The meter station costs, like the mainline valve station costs, will be added to the pipeline cost. Pressure regulating stations are installed at some locations on a gas pipeline to reduce the pressure for delivery to a customer or to protect a section of a pipeline with a lower MOP. Such pressure regulating stations can also be estimated as a lump sum per site and added to the capital cost of the pipeline. 10.2.5

SCADA and Telecommunication System

Typically on a gas pipeline, the pressures, ﬂow rates, and temperatures are monitored along the pipeline by means of electronic signals sent from remote terminal units (RTU) on various valves and meters to a central control center via telephone lines or

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 334 Friday, April 1, 2005 3:32 PM

334

GAS PIPELINE HYDRAULICS

microwave or satellite communication systems. The term supervisory control and data acquisition (SCADA) is used to refer to these facilities. SCADA is used to remotely monitor, operate, and control a gas pipeline system from a central control center. In addition to monitoring valve status, ﬂows, temperatures, and pressures along a pipeline, SCADA also monitors the compressor stations. In many cases, starting and stopping of compressor units are performed remotely using SCADA. The cost of SCADA facilities range from $2 million to $5 million or more, depending on the pipeline length, number of compressor stations, and the number of mainline valves and meter stations. Sometimes this category is estimated as a percentage of the total project cost, such as 2 to 5%. 10.2.6

Environmental and Permitting

The environmental and permitting costs are those costs that are associated with the modiﬁcations to pipeline, compressor stations, and valve and meter stations to ensure that these facilities do not pollute the atmosphere, streams, and rivers or damage ecosystems including the ﬂora and fauna, ﬁsh and game, and endangered species. Many sensitive areas, such as Native American religious and burial sites, must be considered and allowances must be made for mitigation of habitat in certain areas. Permitting costs can include those costs associated with changes needed to compression equipment, pipeline alignment such that toxic emissions from pipeline facilities do not endanger the environment, humans, and plant and animal life. In many cases, these costs include acquisition of land to compensate for the areas that were disturbed due to pipeline construction. Such lands acquired will be allocated for public use, such as parks and wildlife preserves. Permitting costs will also include an environmental study, the preparation of an environmental impact report, and permits for road crossings, railroad crossings, and stream and river crossings. These environmental and permitting costs on a gas pipeline project may range between 10 and 15% of total project costs. 10.2.7

Right of Way Acquisitions

The right of way (ROW) for a pipeline is acquired from private parties and state and local government and federal agencies for a fee. This fee might be a lump sum payment at the time of acquisition with additional annual fees to be paid for a certain duration. For example, the right of way can be acquired from private farms, cooperatives, the Bureau of Land Management (BLM), and railroads. The initial cost for acquiring the ROW will be included in the capital cost of the pipeline. The annual rent or lease payment for land will be considered an expense. The latter will be included in the annual costs, such as operating costs. As an example, the ROW acquisition costs for a gas pipeline might be $30 million. This cost would be added to the total capital costs of the gas pipeline. Also, there might be annual ROW lease payments of $300,000 a year, which would be added to other annual costs such as operating and maintenance costs and administrative costs. For most gas pipelines, the initial ROW costs will be in the range of 6 to 10% of the total project costs.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 335 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

10.2.8

335

Engineering and Construction Management

Engineering costs are those costs that pertain to the design and preparation of drawings for the pipeline, compressor stations, and other facilities. This will include both preliminary and detailed engineering design costs, including development of speciﬁcations, manuals, purchase documents, equipment inspection, and other costs associated with materials and equipment acquisition for the project. The construction management costs include ﬁeld personnel cost, rental facilities, ofﬁce equipment, transportation, and other costs associated with overseeing and managing the construction effort for the pipeline and facilities. On a typical pipeline project, engineering and construction management costs range from 15 to 20% of the total pipeline project cost. 10.2.9

Other Project Costs

In addition to the major cost categories discussed in the preceding sections, there are other costs that should be included in the total pipeline project cost. These include legal and regulatory costs necessary for ﬁling an application with the FERC and state agencies that have jurisdiction over interstate and intrastate transportation of natural gas, as well as a contingency costs intended to cover categories not considered or not envisioned when the project was conceptualized. As the project is engineered, new issues and problems might surface that require additional funds. These are generally included in the category of contingency cost. The ﬁnal category of cost, referred to as allowance for funds used during construction (AFUDC), is intended to cover the cost associated with ﬁnancing the project during various stages of construction. Contingency and AFUDC costs can range between 15 and 20% of the total project cost. Table 10.2 shows a cost breakdown for a typical natural gas pipeline project.

Table 10.2

Cost Breakdown for a Typical Natural Gas Pipeline Project

Description 1 2 3 4 5 6 7 8 9 10

Pipeline Compressor stations Mainline valve stations Meter stations Pressure regulator stations SCADA and telecommunications Environmental and permitting Right of way acquisiton Engineering and construction management Contingency

11 12

Working capital AFUDC

Million $

2 to 5% 10 to 15% 6 to 10% 15 to 20% 10%

Sub-Total

Total

Copyright 2005 by Taylor & Francis Group, LLC

160.00 20.00 1.20 1.20 0.10 5.48 21.90 14.60 36.50 26.10 287.08

5%

5.00 14.35 306.43

2785_C010.fm Page 336 Friday, April 1, 2005 3:32 PM

336

GAS PIPELINE HYDRAULICS

10.3 OPERATING COSTS Once the pipeline, compressor stations, and ancillary facilities are constructed and the pipeline is put into operation, there will be annual operating costs over the useful life of the pipeline, which might be 30 to 40 years or more. These annual costs consist of the following major categories: • Compressor station fuel or electrical energy cost • Compressor station equipment maintenance and repair costs • Pipeline maintenance costs, such as pipe repair, relocation, aerial patrol, and monitoring • SCADA and telecommunication • Valve, regulator, and meter station maintenance • Utility costs, such as water and natural gas • Annual or periodic environmental and permitting costs • Lease, rental, and other recurring right of way costs • Administrative and payroll costs

Compressor station costs include periodic equipment maintenance and overhaul costs. For example, a gas turbine–driven compressor unit may have to be overhauled every 18 to 24 months. Table 10.3 shows the annual operating cost of a typical gas pipeline. Example 2 A new pipeline is being constructed to transport natural gas from a gas processing plant to a power plant 100 mi away Two project phases are envisioned. During the ﬁrst phase lasting 10 years, the amount of gas shipped is expected to be a constant

Table 10.3

Annual Operating Cost of a Typical Gas Pipeline

Description 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Salaries Payroll overhead (20%) Admin and general (50%) Vehicle expense Office expenses (6%) Misc materials and tools Compressor station maintenance Consumable materials Periodic maintenance ROW payments Utilities Gas control SCADA contract install and maintenance Internal corrosion inspection ($750,000/3 years) Cathodic protection survey Total O&M

Copyright 2005 by Taylor & Francis Group, LLC

$ per year 860,000 172,000 516,000 72,800 92,880 100,000 50,000 150,000 350,000 150,000 100,000 200,000 250,000 100,000 3,163,680

2785_C010.fm Page 337 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

337

volume of 120 MMSCFD at a 95% load factor. A pipe size of NPS 16, 0.250 in. wall thickness is required to handle the volumes with two compressor stations with a total of 5000 HP. The total pipeline cost can be estimated at $800,000 per mi and compressor station cost at $2000 per HP installed. The annual operating costs are estimated at $8 million. The pipeline construction project will be ﬁnanced by borrowing 80% of the required capital at an interest rate of 6%. The regulatory rate of return allowed on equity capital is 14%. Consider a project life of 20 years and an overall tax rate of 40%. (a) Calculate the annual cost of service for this pipeline and the transportation tariff in $/MCF. (b) The second phase, lasting the next 10 years, is projected to increase throughput to 150 MMSCFD. Calculate the transportation tariff for the second phase, considering the capital cost to increase by $20 million and the annual cost to increase to $10 million, with the same load factor as phase 1. Solution (a) First, calculate the total capital cost of facilities of phase 1. Pipeline cost = $800,000 × 100 = $80 million Compressor station cost = $2000 × 5000 = $10 million Total capital cost = $80 + $10 = $90 million 80% of this capital of $90 million will be borrowed at 6% interest for 20 years. From Equation 10.1, the annual cost to amortize the loan is Loan amortization cost =

90 × 0.8 × 0.06 1−

( 1.106 )

20

= $6.28 million

Therefore, we need to build into the cost of service the annual payment of $6.28 million to retire the debt of $72 million (80% of $90 million) over the project life of 20 years. On the remaining capital (equity) of ($90 – $72) million or $18 million, a 14% rate of return per year is allowed. Therefore, 14% of $18 million can be included in the cost of service to account for the equity capital. Annual revenue on equity capital = 0.14 × $18 million = $2.52 million Since the tax rate is 40%, the adjusted annual revenue on equity capital is $2.52 million = $4.2 million 1 − 0. 4 Next, add the operating cost of $8 million per year to the annual costs for debt and equity just calculated to arrive at the annual cost of service as follows: Annual payment to retire debt = $6.28 million Annual revenue on equity capital = $4.2 million Annual operating cost = $8 million

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 338 Friday, April 1, 2005 3:32 PM

338

GAS PIPELINE HYDRAULICS

Therefore, Annual cost of service = 6.28 + 4.2 + 8 = $18.48 million The transportation tariff at 120 MMSCFD and 95% load factor is Tariff =

18.48 × 10 6 × 10 3 = $0.4441 per MCF. 120 × 10 6 × 365 × 0.95

(b) In the second phase, which lasts 10 years, the capital cost increases by $20 million. The extra $20 million will be assumed to be ﬁnanced by 80% debt and 20% equity as before. The annual cost to amortize the debt is Loan amortization cost =

20 × 0.8 × 0.06 1−

( 1.106 )

10

= $2.17 million

The remaining capital of ($20 – $16) or $4 million is equity that, according to regulatory guidelines, can earn 14% interest. It must be noted that the interest rate and ROR used in this example are approximate and only for the purpose of illustration. The actual ROR allowed on a particular pipeline will depend on various factors such as the state of the economy, current FERC regulations, or state laws, and can range from as low as 8% to as high as 16% or more. Similarly, the interest rate of 6% used for debt amortization is also an illustrative number. The actual interest rate on debt will depend on various factors such as the state of the economy, money supply, and the federal interest rate charged by Federal Reserve (prime rate). This rate will vary with the country where the pipeline is built and the multinational bank that might ﬁnance the pipeline project. For phase 2, the annual revenue on equity capital is 4 × 0.14 = $0.56 million Accounting for a 40% tax rate, the adjusted annual revenue on equity capital is $0.56 = $0.93 million 1 − 0. 4 Therefore, for phase 2, the increase in capital of $20 million and operating cost of $2 million will result in an increase in cost of service of Annual cost of service = $2.17 + 0.93 + 2 = $5.1 million In summary, for phase 2, the total cost of services is $18.48 + $5.1 = $23.58 million At a ﬂow rate of 150 MMSCFD and 95% load factor, the tariff for phase 2 is 23.58 × 10 6 × 10 3 = $0.4534 per MCF 150 × 10 6 × 365 × 0.95

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 339 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

339

10.4 DETERMINING ECONOMIC PIPE SIZE For a particular pipeline transportation application, there is an economic or optimum pipe diameter that will result in the lowest cost of facilities. For example, a pipeline that requires 100 MMSCFD gas to be transported from a source location to a destination location may be constructed of a wide range of pipe materials and diameters. We may choose to use NPS 14, NPS 16, or NPS 18 pipe or any other pipe size for this application. Using the smallest-diameter pipe will cause the greatest pressure drop and the highest HP requirement for a given volume ﬂow rate. The largest pipe size will result in the lowest pressure drop and, hence, require the least HP. Therefore, the NPS 14 system will be the lowest in pipe material cost and highest in HP required. On the other hand, the NPS 18 system will require the least HP but considerably more pipe material cost due to the difference in pipe weight per unit length. Determining the optimum pipe size for an application will be illustrated in the next example. Example 3 A gas pipeline is to be constructed to transport 150 MMSCFD of natural gas from Dixie to Florence, 120 mi away. Consider three pipe sizes—NPS 14, NPS 16, and NPS 18—all having 0.250 in. wall thickness. Determine the most economical pipe diameter, taking into account the pipe material cost, cost of compressor stations, and fuel costs. The selection of pipe size may be based on a 20-year project life and a present value (PV ) of discounted cash ﬂow at 8% per year. Use $800 per ton for pipe material and $2000 per installed HP for compressor station cost. Fuel gas can be estimated at $3 per MCF. The following information from hydraulic analysis is available: NPS 14 pipeline: Two compressor stations, 8196 HP total. Fuel consumption is 1.64 MMSCFD. NPS 16 pipeline: One compressor station, 3875 HP. Fuel consumption is 0.78 MMSCFD. NPS 18 pipeline: One compressor station, 2060 HP. Fuel consumption is 0.41 MMSCFD. Solution First, calculate the capital cost of 120 mi of pipe for each case. From Equation 10.2, the cost of NPS 14 pipe is PMC =

10.68(14 − 0.250) × 0.250 × 120 × 800 × 5280 = $9.3 million 2000

Similarly, the cost of NPS 16 pipe is PMC =

10.68(16 − 0.250) × 0.250 × 120 × 800 × 5280 = $10.66 million 2000

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 340 Friday, April 1, 2005 3:32 PM

340

GAS PIPELINE HYDRAULICS

and the cost of NPS 18 pipe is PMC =

10.68(18 − 0.250) × 0.250 × 120 × 800 × 5280 = $12.01 million 2000

Next, calculate the installed cost of compressor stations for each pipe size. For NPS 14 pipe, the compressor station cost is 8196 × 2000 = $16.39 million For NPS 16 pipe, the compressor station cost is 3875 × 2000 = $7.75 million For NPS 18 pipe, the compressor station cost is 2060 × 2000 = $4.12 million The operating fuel cost for each case will be calculated next, considering fuel gas at $3 per MCF and 24-hour-a-day operation for 350 days a year. A shutdown for 15 days per year is allowed for maintenance and any operational upset conditions. For NPS 14 pipe, the fuel cost is 1.64 × 10 3 × 350 × 3 = $1.72 million per year For NPS 16 pipe, the fuel cost is 0.78 × 10 3 × 350 × 3 = $0.82 million per year For NPS 18 pipe, the fuel cost is 0.41 × 10 3 × 350 × 3 = $0.43 million per year The actual operating cost includes many other items besides the fuel cost. For simplicity, in this example we will only consider the fuel cost. The annual fuel cost for the project life of 20 years will be discounted at 8% in each case. This will then be added to the sum of the pipeline and compressor station capital cost to arrive at a present value (PV). The present value of a series of cash ﬂows, each equal to R for a period of n years at an interest rate of i%, is given by Equation 10.1. The PV of NPS 14 fuel cost is, from Equation 10.1, PV =

1 1.72 1− = 1.72 × 9.8181 = $16.89 million 20 0.08 (1 + 0.08)

The PV of NPS 16 fuel cost is PV = 0.82 × 9.8181 = $8.05 million

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 341 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

341

The PV of NPS 18 fuel cost is PV = 0.43 × 9.8181 = $4.22 million Therefore, adding up all costs, the PV for NPS 14 is PV14 = 9.3 + 16.39 + 16.89 = $42.58 million Adding up all costs, the PV for NPS 16 is PV16 = 10.66 + 7.75 + 8.05 = $26.46 million and adding up all costs, the PV for NPS 18 is PV18 = 12.01 + 4.12 + 4.22 = $20.35 million Therefore, we see that the lowest cost option is NPS 18 pipeline with a PV of $20.35 million.

In the preceding example, if the ﬂow rate had been lower or higher, the result may be different. For each pipe size, if we were to calculate the HP required at various ﬂow rates and the corresponding fuel consumption, we could generate a graph showing the variation of total cost with ﬂow rate. Obviously, as ﬂow rate is increased, the HP required and fuel consumption also increase. Performing these calculations for different pipe sizes will yield a graph similar to that shown in Figure 10.1. In the next example, we will consider three pipe sizes (NPS 16, NPS 18, and NPS 20) and calculate the capital cost and O&M cost for a range of ﬂow rates to develop curves similar to those shown in Figure 10.1.

NPS 20

Cost

NPS 18 NPS 16

Minimum cost Flow rate Figure 10.1

Pipeline cost vs. flow rate for various pipe sizes.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 342 Friday, April 1, 2005 3:32 PM

342

GAS PIPELINE HYDRAULICS

Example 4 For a natural gas pipeline 120 mi long, three pipe sizes were analyzed for ﬂow rate ranges of 50 to 500 MMSCFD using a hydraulic simulation software application, GASMOD (www.systek.us). The following are the pipe sizes and ﬂow rates studied: NPS 16 pipe: ﬂow rates—50 to 200 MMSCFD NPS 18 pipe: ﬂow rates—50 to 300 MMSCFD NPS 20 pipe: ﬂow rates—100 to 500 MMSCFD The wall thickness was 0.250 in. for NPS 16 and NPS 18 and 0.500 in. for NPS 20. From the hydraulic simulation, the number of compressor stations required, HP, and fuel consumption were obtained as shown in Table 10.4.

Table 10.4 Flow Rate, MMSCFD 50 100 150 175 200

Flow Rate, MMSCFD 50 100 150 175 200 250 300

Flow Rate, MMSCFD 100 150 175 200 250 300 400 500

Copyright 2005 by Taylor & Francis Group, LLC

Hydraulic Simulation Results for Three Pipe Sizes NPS 16 Compressor Stations 1 1 1 2 2 NPS 18 Compressor Stations 1 1 1 1 1 2 2 NPS 20 Compressor Stations 1 1 1 1 1 2 2 2

Total HP

Fuel, MMSCFD

49 1072 3875 5705 9203

0.01 0.21 0.78 1.14 1.84

Total HP

Fuel, MMSCFD

49 209 2060 3394 4954 9348 17902

0.01 0.04 0.41 0.68 1 1.87 3.58

Total HP

Fuel, MMSCFD

98 1053 2057 3281 6312 10519 31401 73207

0.02 0.21 0.41 0.66 1.26 2.1 6.28 14.64

2785_C010.fm Page 343 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

343

Develop annualized costs for each pipe size and ﬂow rate using the following assumptions: The capital cost of the pipe material is based on $800 per ton. For pipe installation cost, use the following: NPS 16: $50 per foot NPS 18: $60 per foot NPS 20: $80 per foot For compressor station capital cost, use $2000 per installed HP. Fuel gas can be assumed to be $3 per MCF. The project life is 20 years, and the interest rate for discounting cash ﬂow is 8%. Add 40% to the pipe and compressor capital costs to account for miscellaneous costs such as meter stations; valves; ROW; environmental, engineering, and construction management; and contingency. The pipeline is assumed to be operational 350 days a year. Solution From the given hydraulic simulation data, using the assumptions listed, we develop the total capital cost of pipe, compressor station, and miscellaneous costs. The pipe material cost is calculated from Equation 10.2, using $800 per ton for pipe material cost, as follows: For NPS 16 pipe, Pipe material cost =

10.68(16 − 0.25)0.25 × 120 × $800 × 5280 = $10.66 million 2000

Similarly, for NPS 18 pipe, Pipe material cost =

10.68(18 − 0.25)0.25 × 120 × $800 × 5280 = $12.01 million 2000

and NPS 20 pipe material cost is Pipe material cost =

10.68(20 − 0.5)0.5 × 120 × $800 × 5280 = $26.39 million 2000

These costs are shown in Table 10.5 through Table 10.7.

Copyright 2005 by Taylor & Francis Group, LLC

NPS 16 Pipe Cost Summary

Number of Flow Compressor Rate Stations

Total HP

Fuel, MMSCFD

50 100 150 175 200

49 1072 3875 5705 9203

0.01 0.21 0.78 1.14 1.84

1 1 1 2 2

Fuel, $/yr

Pipe Material, $

Pipe Labor, $

0.01 0.22 0.82 1.20 1.93

10.66 10.66 10.66 10.66 10.66

31.68 31.68 31.68 31.68 31.68

Copyright 2005 by Taylor & Francis Group, LLC

42.34 42.34 42.34 42.34 42.34

0.098 2.144 7.75 11.41 18.406

16.97 17.79 20.04 21.50 24.30

Total Capital, $

O&M Cost, $/yr

Annualized Capital, $/yr

Total Annual Cost, $

Annual Cost, $/MCF

59.41 62.27 70.12 75.25 85.04

2.00 2.00 2.00 3.00 3.00

6.05 6.34 7.14 7.66 8.66

8.06 8.56 9.96 11.86 13.59

0.4607 0.2447 0.1897 0.1937 0.1942

GAS PIPELINE HYDRAULICS

Notes: Pipe material cost = $800/ton Pipe labor cost = $50/ft for NPS 16 Compressor station cost = $2000 per installed HP Miscellaneous cost is 40% of pipe and compressor station cost Operating cost based on 350 days per year Fuel cost is $3 per MCF Capital cost is annualized at 8% interest for 20-year project life Table values in millions of dollars

NPS 16 Total Compressor Misc Pipe Station Cost, Cost, $ Cost, $ $

2785_C010.fm Page 344 Friday, April 1, 2005 3:32 PM

344

Table 10.5

Number of Flow Compressor Rate Stations

Total HP

Fuel, MMSCFD

50 100 150 175 200 250 300

49 209 2060 3394 4954 9348 17902

0.01 0.04 0.41 0.68 1.00 1.87 3.58

1 1 1 1 1 2 2

Fuel, $/yr

Pipe Material, $

Pipe Labor, $

0.01 0.04 0.43 0.71 1.05 1.96 3.76

12.01 12.01 12.01 12.01 12.01 12.01 12.01

38.02 38.02 38.02 38.02 38.02 38.02 38.02

NPS 18 Total Compressor Misc Pipe Station Cost, Cost, $ Cost, $ $ 50.03 50.03 50.03 50.03 50.03 50.03 50.03

0.098 0.418 4.12 6.788 9.908 18.696 35.804

20.05 20.18 21.66 22.73 23.97 27.49 34.33

Total Capital, $

O&M Cost, $/yr

Annualized Capital, $/yr

Total Annual Cost, $

Annual Cost, $/MCF

70.18 70.62 75.81 79.54 83.91 96.21 120.16

2.00 2.00 2.00 2.00 2.00 3.00 3.00

7.15 7.19 7.72 8.10 8.55 9.80 12.24

9.16 9.24 10.15 10.82 11.60 14.76 19.00

0.5233 0.2639 0.1934 0.1766 0.1657 0.1687 0.1809

Notes: Pipe material cost = $800/ton Pipe labor cost = $60/ft for NPS 18 Compressor station cost = $2000 per installed HP Miscellaneous cost is 40% of pipe and compressor station cost Operating cost based on 350 days per year Fuel cost is $3 per MCF Capital cost is annualized at 8% interest for 20-year project life Table values in millions of dollars

345

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 345 Friday, April 1, 2005 3:32 PM

NPS 18 Pipe Cost Summary

PIPELINE ECONOMICS

Table 10.6

NPS 20 Pipe Cost Summary

Number of Flow Compressor Rate Stations

Total HP

Fuel, MMSCFD

Fuel, $/yr

Pipe Material, $

100 150 175 200 250 300 400 500

98 1053 2057 3281 6312 10519 31401 73207

0.02 0.21 0.41 0.66 1.26 2.1 6.28 14.64

0.02 0.22 0.43 0.69 1.32 2.21 6.59 15.37

26.39 26.39 26.39 26.39 26.39 26.39 26.39 26.39

1 1 1 1 1 2 2 2

NPS 20 Pipe Total Compressor Misc Labor, Pipe Station Cost, $ Cost, $ Cost, $ $

Total Capital, $

O&M Cost, $/yr

50.69 50.69 50.69 50.69 50.69 50.69 50.69 50.69

108.18 110.86 113.67 117.10 125.58 137.36 195.83 312.89

2.00 2.00 2.00 2.00 2.00 3.00 3.00 3.00

Copyright 2005 by Taylor & Francis Group, LLC

0.196 2.106 4.114 6.562 12.624 21.038 62.802 146.414

30.91 31.67 32.48 33.46 35.88 39.25 55.95 89.40

11.02 11.29 11.58 11.93 12.79 13.99 19.95 31.87

13.04 13.51 14.01 14.62 16.11 19.20 29.54 50.24

0.3726 0.2574 0.2287 0.2089 0.1842 0.1828 0.2110 0.2871

GAS PIPELINE HYDRAULICS

Notes: Pipe material cost = $800/ton Pipe labor cost = $80/ft for NPS 20 Compressor station cost = $2000 per installed HP Miscellaneous cost is 40% of pipe and compressor station cost Operating cost based on 350 days per year Fuel cost is $3 per MCF Capital cost is annualized at 8% interest for 20-year project life Table values in millions of dollars

77.08 77.08 77.08 77.08 77.08 77.08 77.08 77.08

Total Annual Annualized Annual Cost, Capital, $/yr Cost, $ $/MCF

2785_C010.fm Page 346 Friday, April 1, 2005 3:32 PM

346

Table 10.7

2785_C010.fm Page 347 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

347

The labor cost for installing pipe is calculated as follows: For NPS 16 pipe, Pipe installation cost = $50 × 5280 × 120 = $31.68 million Similarly, for NPS 18 pipe, Pipe installation cost = $60 × 5280 × 120 = $38.02 million and for NPS 20 pipe, Pipe installation cost = $80 × 5280 × 120 = $50.69 million Next, we calculate the installation cost of compressor stations using $2000 per installed HP. For the NPS 16 pipe at 100 MMSCFD ﬂow rate, the HP required is 1072 and the installation cost is $2000 × 1072 = $2.14 million Similarly, the installation costs of each compressor station for all cases are calculated and tabulated as shown in Table 10.5 through Table 10.7. The miscellaneous cost is 40% of the sum of the pipe cost and compressor station cost, as follows: Pipe material cost = $10.66 million Pipe installation cost = $31.68 million Compressor station cost = $2.14 million Thus, for NPS 16 pipe at 100 MMSCFD, Miscellaneous cost = 0.40 × (10.66 + 31.68 + 2.14) = $17.79 million The operation and maintenance costs are added to the annual fuel cost to obtain the total annual cost. The total capital cost is annualized at 8% interest for 20 years and added to the O&M and fuel costs. For example, for NPS 16 pipe at 100 MMSCFD ﬂow rate, the total capital cost of $62.27 million is annualized at $6.34 million and added to the O&M and fuel costs to obtain the total annual cost of $8.56 million. Dividing this annual cost by the gas transported per year, we obtain the annual cost per MCF as follows: Annual cost per MCF =

8.56 × 10 6 × 10 3 = $0.2447 100 × 10 6 × 350

Similarly, for each pipe size and ﬂow rate, the values are tabulated as shown in Table 10.5 through Table 10.7. Upon reviewing Table 10.5 for NPS 16 pipe, we see that the annual cost per MCF decreases from $0.4607 to $0.1897 as the ﬂow rate increases from 50 to 150 MMSCFD.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 348 Friday, April 1, 2005 3:32 PM

348

GAS PIPELINE HYDRAULICS

After that, it increases with ﬂow rate and reaches a value of $0.1942 at 200 MMSCFD. Therefore, for NPS 16 pipe, 150 MMSCFD is the optimum ﬂow rate that results in the lowest transportation cost. Similarly, from Table 10.6, for NPS 18 pipe, the annual cost per MCF decreases from $0.5233 to $0.1657 as the ﬂow rate increases from 50 to 200 MMSCFD. After that, it increases with ﬂow rate and reaches a value of $0.1809 at 300 MMSCFD. Therefore, for NPS 18 pipe, 200 MMSCFD is the optimum ﬂow rate that results in the lowest transportation cost. Finally, from Table 10.7, for NPS 20 pipe, the annual cost per MCF decreases from $0.3726 to $0.1828 as the ﬂow rate increases from 100 to 300 MMSCFD. After that, it increases with ﬂow rate and reaches a value of $0.2871 at 500 MMSCFD. Therefore, for NPS 20 pipe, 300 MMSCFD is the optimum ﬂow rate that results in the lowest transportation cost. A plot of the annual cost per MCF vs. ﬂow rate for the three pipe sizes is shown in Figure 10.2.

In the preceding calculations, to simplify matters, several assumptions were made. Miscellaneous costs were estimated as a percentage of the pipeline and compressor station costs. Also, we considered the annual costs to be constant from year to year. A more nearly accurate calculation would be to escalate the annual costs by a percentage each year to account for inﬂation, using the Consumer Price Index (CPI). Nevertheless, the preceding calculations illustrate a methodology of economic analysis to determine the most optimum pipe size. Example 5 In Chapter 5, we compared expanding the capacity of the gas pipeline from Windsor to Cardiff using two options—installing intermediate compressor stations or installing pipe loops. Using the results of Example 1 in Chapter 5, compare the two options, 0.6000 0.5000

$/MCF

0.4000 NPS 20 NPS 18 NPS 16

0.3000 0.2000 0.1000 0.0000

50

100

150

175

200

250

300

Flow rate, MMSCFD Figure 10.2

Annualized cost vs. flow rate for three pipe sizes.

Copyright 2005 by Taylor & Francis Group, LLC

400

2785_C010.fm Page 349 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

349

taking into account the capital cost, operating cost, and fuel cost and considering a project life of 25 years. The capital will be ﬁnanced with 70% debt at 8% interest. The regulatory return allowed on the 30% equity is 12%. The tax rate can be assumed to be 35%. Fuel consumption is 0.2 MCF per day per HP and fuel gas cost is $3 per MCF. Assume 350 days of operation per year. Calculate the annualized cost of service and transportation tariff for both options. It is expected that the annual O&M cost will increase by $2 million for phase 1 and an additional $3 million for phase 2 compressor station options. For the looping option, the incremental O&M cost is $0.5 million for phase 1 and $0.75 million for phase 2. Solution Phase 1 expansion This expansion results in a ﬂow rate of 238.41 MMSCFD, and the compressor station option requires installing the following HP: Windsor compressor station—8468 HP Avon compressor station—3659 HP Total HP = 8468 + 3659 = 12,127 HP The incremental HP for phase 1 was calculated as ∆HP = 12,127 – 7064 = 5063 HP The cost of this incremental HP based on $2000 per installed HP is ∆Capital cost = 5063 × 2000 = $10.13 million The incremental fuel cost for 5063 HP is ∆Fuel cost = 5063 × 0.2 × $3 × 350 = $1.06 million per year. The incremental capital cost of $10.13 million will be funded by 70% debt and 30% equity. The debt capital = 10.13 × 0.7 = $7.09 million. Loan amortization cost =

10.13 × 0.7 × 0.08 1−

( 1.108 )

25

= $0.66 million per year

The remaining capital of ($10.13 – $7.09) million or $3.04 million is equity that, according to regulatory guidelines, can earn 12% interest. The annual revenue allowed on equity capital is 3.04 × 0.12 = $0.36 million Accounting for a 35% tax rate, the adjusted annual revenue on equity capital is $0.36 = $0.55 million 1 − 0.35

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 350 Friday, April 1, 2005 3:32 PM

350

GAS PIPELINE HYDRAULICS

Next, add the O&M cost increase of $2 million per year and the fuel cost of $1.06 million to the annual costs for debt and equity just calculated to arrive at the annual cost of service as follows: Annual Annual Annual Annual

payment to retire debt = $0.66 million revenue on equity capital = $0.36 million operating cost = $2.0 million fuel cost = $1.06 million

Therefore, the incremental annual cost of service for phase 1 expansion compressor station option is $0.66 + 0.36 + 2.0 + 1.06 = $4.08 million. This amount is the incremental annual cost of service over and above the cost of service for the initial ﬂow rate of 188.41 MMSCFD. The incremental tariff for an incremental ﬂow rate of 50 MMSCFD for phase 1 expansion is Incremental tariff =

4.08 × 10 6 × 10 3 = $0.2331 per MCF 50 × 10 6 × 350

Next, we calculate the cost of service and tariff considering the looping option. In Example 1 of Chapter 5, for phase 1 expansion, we required installation of 50.03 mi of loop at a cost of $25.02 million. In addition to this cost of pipe loop, we must include the cost of the increased horsepower requirement at Windsor for the phase 1 ﬂow rate, which was calculated at 1404 HP. At $2000 per installed HP, the extra cost for incremental HP is $2.81 million. Thus, for phase 1 the total cost of looping pipe upstream of Cardiff and increased HP cost at the Windsor compressor station is calculated as $25.02 + $2.81 = $27.83 million The incremental fuel cost for the extra 1404 HP is ∆Fuel cost = 1404 × 0.2 × $3 × 350 = $0.30 million per year The incremental capital of $27.83 million for the looping option would also be funded by 70% debt and 30% equity. Debt capital = 27.83 × 0.7 = $19.48 million Loan amortization cost =

19.48 × 0.08 1−

( 1.108 )

25

= $1.82 million per year

The remaining capital of ($27.83 – $19.48) million or $8.35 million is equity that, according to regulatory guidelines, can earn 12% interest.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 351 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

351

The annual revenue allowed on equity capital is 8.35 × 0.12 = $1.0 million Accounting for a 35% tax rate, the adjusted annual revenue on equity capital is $1.0 = $1.54 million 1 − 0.35 Next, add the O&M cost increase of $0.5 million per year and the fuel cost of $0.30 million to the annual costs for debt and equity just calculated to arrive at the incremental annual cost of service as follows: Annual Annual Annual Annual

payment to retire debt = $1.82 million revenue on equity capital = $1.0 million operating cost = $0.5 million fuel cost = $0.3 million

Therefore, the incremental annual cost of service for phase 1 expansion looping option is $1.82 + 1.0 + 0.5 + 0.3 = $3.62 million. This amount is the incremental annual cost of service over and above the cost of service for the initial ﬂow rate of 188.41 MMSCFD. The incremental tariff for an incremental ﬂow rate of 50 MMSCFD for the phase 1 expansion looping option is Incremental tariff =

3.62 × 10 6 × 10 3 = $0.2069 per MCF 50 × 10 6 × 350

We can summarize the calculations as follows: For the phase 1 expansion, compressor station option: Incremental annual cost of service = $4.08 million Incremental tariff = $0.2331 per MCF For the phase 1 expansion, looping option: Incremental annual cost of service = $3.62 million Incremental tariff = $0.2069 per MCF It can be seen that the incremental annual cost of service and the incremental tariff for phase 1 expansion are less in the looping option than the compressor station option. Therefore, for phase 1 expansion, the looping option is the preferred choice. For phase 2 expansion, the throughput increase of 50 MMSCFD will be on top of phase 1 expansion. Since the preferred choice for phase 1 expansion is the

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 352 Friday, April 1, 2005 3:32 PM

352

GAS PIPELINE HYDRAULICS

looping option, we must consider the increase in facilities required for phase 2 with 50.03 mi of pipe loop already installed. In Example 1 of Chapter 5, for phase 2, the loop required was calculated to be 76.26 mi. The incremental HP at Windsor was calculated as 1775 HP. Also, the incremental looping required and cost of increased HP at Windsor over the phase 1 values were calculated to be $16.66 million. The incremental fuel cost for the extra 1775 HP is ∆Fuel cost = 1775 × 0.2 × $3 × 350 = $0.37 million per year The incremental capital of $16.66 million for the phase 2 looping option would also be funded by 70% debt and 30% equity. Debt capital = 16.66 × 0.7 = $11.66 million Loan amortization cost =

11.66 × 0.08 1−

( 1.108 )

25

= $1.09 million per year

The remaining capital of ($16.66 – $11.66) million or $5.0 million is equity that, according to regulatory guidelines, can earn 12% interest. The annual revenue allowed on equity capital is $5.0 × 0.12 = $0.6 million Accounting for a 35% tax rate, the adjusted annual revenue on equity capital is $0.6 = $0.92 million 1 − 0.35 Next, add the O&M cost increase of $0.75 million per year and the fuel cost of $0.37 million to the annual costs for debt and equity just calculated to arrive at the incremental annual cost of service for the phase 2 looping expansion as follows: Annual Annual Annual Annual

payment to retire debt = $1.09 million revenue on equity capital = $0.6 million operating cost = $0.75 million fuel cost = $0.37 million

Therefore, the incremental annual cost of service for phase 2 expansion looping option is $1.09 + 0.6 + 0.75 + 0.37 = $2.81 million. This amount is the incremental annual cost of service over and above the cost of service for the phase 1 ﬂow rate of 238.41 MMSCFD.

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 353 Friday, April 1, 2005 3:32 PM

PIPELINE ECONOMICS

353

The incremental tariff for an incremental ﬂow rate of 50 MMSCFD for the phase 2 expansion looping option is Incremental tariff =

2.81 × 10 6 × 10 3 = $0.1606 per MCF 50 × 10 6 × 350

In summary, For the phase 2 expansion, looping option: Incremental annual cost of service = $2.81 million Incremental tariff = $0.1606 per MCF These incremental costs are over and above the phase 1 numbers. It must be noted that we did not consider a compressor station option for phase 2 expansion. This is because the preferred option for phase 1 expansion was installing loop. Since approximately 50 mi of pipe loop was already installed for phase 1, we simply looked at adding approximately 26 mi of extra loop for phase 2. For comparison, we could determine additional compressor station requirements for phase 2 instead of extending the loop. This is left as an exercise for the reader.

10.5 SUMMARY In this chapter the economic aspects of a natural gas pipeline transportation were reviewed. A method for determining the optimum pipe size necessary to transport a certain ﬂow rate was discussed. We introduced concepts of the capital cost of pipeline and compressor stations and the annual operating and maintenance costs. The fuel consumption calculations were also explained. Taking into account time value of money and the rate of return allowed on an equity investment in pipeline facilities, we calculated an annual cost of transporting gas. From this annual cost, the transportation tariff was calculated. The economic pipe size for a particular application was illustrated using three different pipe sizes and estimating the initial capital cost and annual operating costs. A typical pipeline expansion scenario with the option of installing compressors vs. pipe loops was also explained using economic principles. Additionally, the major components of the capital cost of a typical pipeline system were reviewed. PROBLEMS 1. A natural gas pipeline transports 120 MMSCFD at a load factor of 95%. The capital cost is estimated at $70 million and the annual operating cost is $6 million. Amortizing the capital at 8% for a project life of 20 years, calculate the cost of the service and transportation tariffs for this pipeline. 2. A new pipeline is being constructed to transport natural gas from a processing plant to a power plant 150 mi away. An initial phase and an expansion phase

Copyright 2005 by Taylor & Francis Group, LLC

2785_C010.fm Page 354 Friday, April 1, 2005 3:32 PM

354

GAS PIPELINE HYDRAULICS

are contemplated. During the initial phase lasting 10 years, the amount of gas shipped is expected to be a constant volume of 100 MMSCFD at a 95% load factor. A pipe size of NPS 18, 0.250 in. wall thickness, is required to handle the volumes with two compressor stations of 5000 HP total. The total pipeline cost can be estimated at $750,000 per mi and the compressor station cost at $2000 per HP installed. The annual operating costs are estimated at $6 million. The construction project will be ﬁnanced by borrowing 75% of the required capital at an interest rate of 6%. The regulatory rate of return allowed on equity is 13%. Consider a project life of 25 years and an overall tax rate of 36%. a. Calculate the annual cost of service for this pipeline and the transportation tariff in $/MCF. b. The second phase, lasting the next 10 years, is projected to increase throughput to 150 MMSCFD. Calculate the transportation tariff for the expansion phase, considering the capital cost to increase by $30 million and the annual cost to increase by $4 million, with the same load factor as before.

3. A gas pipeline is to be constructed to transport 200 MMSCFD of natural gas from Jackson to Columbus, 180 mi away. Consider three pipe sizes—NPS 18, NPS 20, and NPS 24—all constructed of API 5L-X52 pipe with suitable wall thickness for an MOP of 1400 psig. Determine the most economical pipe diameter, taking into account the pipe material cost, cost of compressor stations, and fuel costs. The selection of pipe size can be based on a 30-year project life and a present value of discounted cash ﬂow at 6% per year. Use $750 per ton for pipe material and $2000 per installed HP for compressor station cost. Fuel gas can be estimated at $3 per MCF.

REFERENCES 1. Nayyar, M.L., Piping Handbook, 7th ed., McGraw-Hill, New York, 2000. 2. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction, 2nd ed., ASME Press, New York, 2003. 3. Pipeline Design for Hydrocarbon Gases and Liquids, American Society of Civil Engineers, New York, 1975. 4. Katz, D.L. et al., Handbook of Natural Gas Engineering, McGraw-Hill, New York, 1959.

Copyright 2005 by Taylor & Francis Group, LLC

2785_A001.fm Page 355 Friday, March 11, 2005 5:31 PM

APPENDIX

A

Units and Conversions

355

Copyright 2005 by Taylor & Francis Group, LLC

millimeter (mm) 1 meter (m) = 1,000 mm 1 kilometer (km) = 1,000 m square meter (m2) 1 hectare = 10,000 m2 cubic meter (m3) kilogram/cubic meter (kg/m3) Newton per cubic meter (N/m3) 1 1 1 1 m2/s Stoke (S), centiStoke (cSt)

pound (lb) inch (in.) 1 foot (ft) = 12 in. 1 mile (mi) = 5,280 ft square foot (ft2) 1 acre = 43,560 ft2 cubic foot (ft3) slug per cubic foot (slug/ft3) pound per cubic foot (lb/ft3) lb/ft-s lb-s/ft2

ft2/s cubic foot/hour (ft3/h) cubic foot/day (ft3/day) million Std ft3/day (MMSCFD) pound (lb)

Weight

Copyright 2005 by Taylor & Francis Group, LLC

Length

Area

Volume

Density

Specific weight

Viscosity (absolute or dynamic)

Viscosity (kinematic)

Flow rate

Force

1 N/m3 = 0.0064 lb/ft3

1 lb/ft3 = 157.09 N/m3

1 m3/h = 35.3134 ft3/h

1 ft3/h = 0.02832 m3/h

1 lb = 4.4482 N

cubic meter/hour (m3/h) cubic meter/day (m3/day) million std m3/day (Mm3/day) Newton (N) = kg-m/s2

1 N = 0.2248 lb

1 m2/s = 10.7639 ft2/s 1 cSt = 1.076 × 10−5 ft2/s

1 lb-s/ft = 47.88 N-s/m 1lb-s/ft2 = 478.8 Poise

1 kg/m3 = 0.0019 slug/ft3

1 slug/ft3 = 515.38 kg/m3

1 cP = 6.7197 × 10-4 lb/ft-s 1 N-s/m2 = 0.0209 lb-s/ft2 1 Poise = 0.00209 lb-s/ft2

1 m3 = 35.3134 ft3

1 ft3 = 0.02832 m3

2

1 m2 = 10.764 ft2 1 hectare = 2.4711 acre

1 ft2 = 0.0929 m2 1 acre = 0.4047 hectare

2

1 mm = 0.0394 in 1 m = 3.2808 ft 1 km = 0.6214 mi

kg = 0.0685 slug kg = 2.205 lb t = 1.1023 U.S. ton t = 0.9842 long ton

1 in = 25.4 mm 1 ft = 0.3048 m 1 mi = 1.609 km

1 1 1 1

1 N = 0.2248 lb

lb = 0.45359 kg slug = 14.594 kg U.S. ton = 0.9072 t long ton = 1.016 t

SI to USCS Conversion

1 lb = 4.4482 N

1 1 1 1

USCS to SI Conversion

356

1 ft2/s = 0.092903 m2/s

Poise (P) = 0.1 Pa-s centiPoise (cP) = 0.01 P Poise = 1 dyne-s/cm2 Poise = 0.1 N-s/m2

Newton (N) = kg-m/s2

metric tonne (t) = 1,000 kg

kilogram (kg)

SI Units

slug (slug) pound mass (lbm) 1 U.S. ton = 2,000 lb 1 long ton = 2,240 lb

USCS Units

Mass

Item

USCS Units—U.S. Customary System of Units SI Units—Systeme International Units (modified metric)

2785_A001.fm Page 356 Friday, March 11, 2005 5:31 PM

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

W/m/°C W/m2/°C

Btu/hr/ft/°F Btu/hr/ft2/°F Btu/lb/°F

Thermal conductivity

Heat transfer coefficient

Specific heat

kJ/kg/°C

degree Celsius (°C) 1 degree Kelvin (K) = °C + 273

degree Fahrenheit (°F) 1 degree Rankin (°R) = °F + 460

1°C = (°F − 32)/1.8 1 K = °R/1.8 1 W/m/°C = 0.5778 Btu/hr/ft/F 1 W/m2/°C = 0.1761 Btu/hr/ft2/°F 1 kJ/kg/°C = 0.2388 Btu/lb/F

1°F = 9/5°C + 32 1°R = 1.8 K 1 Btu/hr/ft/°F = 1.7307W/m/°C 1 Btu/hr/ft2/°F = 5.6781 W/m2/°C 1 Btu/lb/°F = 4.1869 kJ/kg/°C

1 kW = 1.3405 HP

1 HP = 0.746 kW

Temperature

1 W = 3.4121 Btu/hr

1 Btu/hr = 0.2931 W

Joule/second (J/s) watt (W) = J/s 1 kilowatt (kW) = 1,000 W

ft-lb/min Btu/hour horsepower (HP) 1 HP = 33,000 ft-lb/min

Power

1 kJ = 0.9478 Btu

1 Btu = 1055.0 J

Joule (J) = N-m

foot-pound (ft-lb) British thermal unit (Btu) 1 Btu = 778 ft-lb

Work and energy

1 m/s = 3.281 ft/s

1 ft/s = 0.3048 m/s

meter/second (m/s)

foot/second (ft/s) mile/hour (mi/h) = 1.4667 ft/s

1 Bar = 14.5 psi 1 kg/cm2 = 14.22 psi

1 psi = 0.069 Bar 1 psi = 0.0703 kg/cm2

Velocity

1 kPa = 0.145 psi

1 psi = 6.895 kPa

Pascal (Pa) = N/m2 1 kiloPascal (kPa) = 1,000 Pa 1 megaPascal (MPa) = 1,000 kPa 1 Bar = 100 kPa kilogram/sq. centimeter (kg/cm2)

pound/square inch, lb/in2 (psi) 1lb/ft2 = 144 psi

Pressure

2785_A001.fm Page 357 Friday, March 11, 2005 5:31 PM

APPENDIX A 357

2785_A002.fm Page 359 Friday, March 11, 2005 5:32 PM

APPENDIX

B

Physical Properties of Various Gases

359

Copyright 2005 by Taylor & Francis Group, LLC

Methane Ethane Propane Isobutane n-butane Iso-pentane n-pentane Neo-pentane n-hexane 2-methyl pentane 3-methyl pentane Neo hexane 2,3-dimethylbutane n-Heptane 2-Methylhexane 3-Methylhexane 3-Ethylpentane 2,2-Dimethylpentane 2,4-Dimethylpentane 3,3-Dimethylpentane Triptane n-octane Di Isobutyl Isooctane n-Nonane n-Decane Cyclopentane Methylcyclopentane Cyclohexane Methylcyclohexane

Gas CH4 C 2H 6 C 3H 8 C4H10 C4H10 C5H12 C5H12 C5H12 C6H14 C6H14 C6H14 C6H14 C6H14 C7H16 C7H16 C7H16 C7H16 C7H16 C7H16 C7H16 C7H16 C8H18 C8H18 C8H18 C9H20 C10H22 C5H10 C6H12 C6H12 C7H14

Formula 16.0430 30.0700 44.0970 58.1230 58.1230 72.1500 72.1500 72.1500 86.1770 86.1770 86.1770 86.1770 86.1770 100.2040 100.2040 100.2040 100.2040 100.2040 100.2040 100.2040 100.2040 114.2310 114.2310 114.2310 128.2580 142.2850 70.1340 84.1610 84.1610 98.1880

Molecular Weight 5,000 800 188.65 72.581 51.706 20.443 15.575 36.72 4.9596 6.769 6.103 9.859 7.406 1.621 2.273 2.13 2.012 3.494 3.294 2.775 3.376 0.5371 1.1020 1.7090 0.17155 0.06088 9.917 4.491 3.267 1.609

Vapor Pressure, psia at 100F 666.0 707.0 617.0 527.9 548.8 490.4 488.1 464.0 436.9 436.6 452.5 446.7 454.0 396.8 396.0 407.6 419.2 401.8 397.4 427.9 427.9 360.7 361.1 372.7 330.7 304.6 653.8 548.8 590.7 503.4

−116.66 90.07 205.93 274.4 305.52 368.96 385.7 321.01 453.8 435.76 448.2 419.92 440.08 512.8 494.44 503.62 513.16 476.98 475.72 505.6 496.24 564.15 530.26 519.28 610.72 652.1 461.1 499.28 536.6 570.2 0.0988 0.0783 0.0727 0.0714 0.0703 0.0684 0.0695 0.0673 0.0688 0.0682 0.0682 0.0667 0.0665 0.0682 0.0673 0.0646 0.0665 0.0665 0.0667 0.0662 0.0636 0.0673 0.0676 0.0657 0.0693 0.0702 0.0594 0.0607 0.0586 0.0600

Critical Constants Pressure, Temp., Volume, F ft3/lb psia 0.5539 1.0382 1.5226 2.0068 2.0068 2.4912 2.4912 2.4912 2.9755 2.9755 2.9755 2.9755 2.9755 3.4598 3.4598 3.4598 3.4598 3.4598 3.4598 3.4598 3.4598 3.9441 3.9441 3.9441 4.4284 4.9127 2.4215 2.9059 2.9059 3.3902

23.654 12.620 8.6059 6.5291 6.5291 5.2596 5.2596 5.2596 4.4035 4.4035 4.4035 4.4035 4.4035 3.7872 3.7872 3.7872 3.7872 3.7872 3.7872 3.7872 3.7872 3.322 3.322 3.322 2.9588 2.6671 5.411 4.509 4.509 3.8649

Ideal Gas, 14.696 psia, 60F Spgr (air = 1.00) ft3/lb-gas 0.52676 0.40789 0.38847 0.38669 0.39500 0.38448 0.38831 0.39038 0.38631 0.38526 0.37902 0.38231 0.37762 0.38449 0.38170 0.37882 0.38646 0.38651 0.39627 0.38306 0.37724 0.38334 0.37571 0.38222 0.38248 0.38181 0.27122 0.30027 0.29012 0.31902

Specific Heat, Btu/lb/F 14.696 psia, 60F Ideal Gas

2785_A002.fm Page 360 Friday, March 11, 2005 5:32 PM

360 GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

Ethylene Propylene Butylene Cis-2-butene Trans-2-butene Isobutene 1-Pentene 1,2-Butadene 1,3-Butadene Isoprene Acetylene Benzene Toluene Ethyl-benzene o-Xylene m-Xylene p-Xylene Styrene Isopropylbenzene Methyl alcohol Ethyl alcohol Carbon monoxide Carbon dioxide Hydrogen sulfide Sulfur dioxide Ammonia Air Hydrogen Oxygen Nitrogen Chlorine Water Helium Hydrogen Chloride

C 2H 4 C 3H 6 C 4H 8 C 4H 8 C 4H 8 C 4H 8 C5H10 C 4H 6 C 4H 6 C 5H 8 C 2H 2 C 6H 6 C 7H 8 C8H10 C8H10 C8H10 C8H10 C 8H 8 C9H12 CH4 O C 2H 6 O CO CO2 H 2S SO2 NH3 N2+O2 H2 O2 N2 Cl2 H 2O He HCl

28.0540 42.0810 56.1080 56.1080 56.1080 56.1080 70.1340 54.0920 54.0920 68.1190 26.0380 78.1140 92.1410 106.1670 106.1670 106.1670 106.1670 104.1520 120.1940 32.0420 46.0690 28.0100 44.0100 34.0820 64.0650 17.0305 28.9625 2.0159 31.9988 28.0134 70.9054 18.0153 4.0026 36.4606

Copyright 2005 by Taylor & Francis Group, LLC

906.71

157.3 0.95

394.59 85.46 211.9

3.225 1.033 0.3716 0.2643 0.3265 0.3424 0.2582 0.188 4.631 2.313

1400 232.8 62.55 45.97 49.88 64.95 19.12 36.53 59.46 16.68

731.0 676.6 586.4 615.4 574.9 580.2 509.5 656.0 620.3 582.0 890.4 710.4 595.5 523 541.6 512.9 509.2 587.8 465.4 1174 891.7 506.8 1071 1306 1143 1647 546.9 187.5 731.4 493 1157 3200.1 32.99 1205

48.54 198.31 296.18 324.31 311.8 292.49 376.86 354 306 403 95.29 552.15 605.5 651.22 674.85 650.95 649.47 703 676.2 463.01 465.31 −220.51 87.73 212.4 315.7 270.2 −221.29 −400.3 −181.4 −232.48 290.69 705.1 −450.31 124.75

0.0746 0.0717 0.0683 0.0667 0.0679 0.0681 0.0674 0.0700 0.0653 0.0660 0.0693 0.0531 0.0549 0.0564 0.0557 0.0567 0.0572 0.0534 0.0569 0.0590 0.0581 0.0527 0.0342 0.0461 0.0305 0.0681 0.0517 0.5101 0.0367 0.0510 0.0280 0.04975 0.2300 0.0356

0.9686 1.4529 1.9373 1.9373 1.9373 1.9373 2.4215 1.8677 1.8677 2.3520 0.8990 2.6971 3.1814 3.6657 3.6657 3.6657 3.6657 3.5961 4.1500 1.1063 1.5906 0.9671 1.5196 1.1768 2.2120 0.5880 1.0000 0.06960 1.1048 0.9672 2.4482 0.62202 0.1382 1.2589

13.527 9.0179 6.7636 6.7636 6.7636 6.7636 5.411 7.0156 7.0156 5.571 14.574 4.8581 4.1184 3.5744 3.5744 3.5744 3.5744 3.6435 3.1573 11.843 8.2372 13.548 8.6229 11.134 5.9235 22.283 13.103 188.25 11.859 13.546 5.3519 21.065 94.814 10.408

0.35789 0.35683 0.35535 0.33275 0.35574 0.36636 0.35944 0.34347 0.34223 0.35072 0.39754 0.24295 0.26005 0.27768 0.28964 0.27427 0.27470 0.26682 0.30704 0.32429 0.33074 0.24847 0.19909 0.23838 0.14802 0.49678 0.2398 3.4066 0.21897 0.24833 0.11375 0.44469 1.24040 0.19086

2785_A002.fm Page 361 Friday, March 11, 2005 5:32 PM

APPENDIX B 361

2785_A003.fm Page 363 Friday, March 11, 2005 5:33 PM

APPENDIX

C

Pipe Properties—U.S. Customary System of Units

363

Copyright 2005 by Taylor & Francis Group, LLC

Copyright 2005 by Taylor & Francis Group, LLC

21/2

0.84 0.84 0.84 0.84 0.84 0.84 1.05 1.05 1.05 1.05 1.05 1.05 1.315 1.315 1.315 1.315 1.315 1.315 1.900 1.900 1.900 1.900 1.900 1.900 2.375 2.375 2.375 2.375 2.375 2.375 2.875 2.875

Outside Dia, in.

40 80 160

40 80 160

40 80 160

40 80 160

40 80 160

a

XXS

Std XS

XXS

Std XS

XXS

Std XS

XXS

Std XS

XXS

Std XS

5S 10S

5S 10S 40S 80S

5S 10S 40S 80S

5S 10S 40S 80S

5S 10S 40S 80S

5S 10S 40S 80S

Schedule b c 0.065 0.083 0.109 0.147 0.187 0.294 0.065 0.083 0.113 0.154 0.218 0.308 0.065 0.109 0.330 0.179 0.250 0.358 0.065 0.109 0.145 0.200 0.281 0.400 0.065 0.109 0.154 0.218 0.343 0.436 0.083 0.12

Wall Thickness, in. 0.710 0.674 0.622 0.546 0.466 0.252 0.920 0.884 0.824 0.742 0.614 0.434 1.185 1.097 0.655 0.957 0.815 0.599 1.770 1.682 1.610 1.500 1.338 1.100 2.245 2.157 2.067 1.939 1.689 1.503 2.709 2.635

Inside Dia, in. 0.3957 0.3566 0.3037 0.2340 0.1705 0.0499 0.6644 0.6134 0.5330 0.4322 0.2959 0.1479 1.1023 0.9447 0.3368 0.7189 0.5214 0.2817 2.4593 2.2209 2.0348 1.7663 1.4053 0.9499 3.9564 3.6523 3.3539 2.9514 2.2394 1.7733 5.7609 5.4504

Inside Area, in.2 0.22 0.22 0.22 0.22 0.22 0.22 0.27 0.27 0.27 0.27 0.27 0.27 0.34 0.34 0.34 0.34 0.34 0.34 0.50 0.50 0.50 0.50 0.50 0.50 0.62 0.62 0.62 0.62 0.62 0.62 0.75 0.75

Surface Area, ft2/ft 0.0027 0.0025 0.0021 0.0016 0.0012 0.0003 0.0046 0.0043 0.0037 0.0030 0.0021 0.0010 0.0077 0.0066 0.0023 0.0050 0.0036 0.0020 0.0171 0.0154 0.0141 0.0123 0.0098 0.0066 0.0275 0.0254 0.0233 0.0205 0.0156 0.0123 0.0400 0.0379

Volume, ft3/ft 0.54 0.67 0.85 1.09 1.30 1.71 0.68 0.86 1.13 1.47 1.94 2.44 0.87 1.40 3.47 2.17 2.84 3.66 1.27 2.08 2.72 3.63 4.86 6.41 1.60 2.64 3.65 5.02 7.44 9.03 2.47 3.53

Pipe Weight, lb/ft 0.02 0.02 0.02 0.01 0.01 0.00 0.04 0.04 0.03 0.03 0.02 0.01 0.06 0.05 0.02 0.04 0.03 0.02 0.14 0.13 0.12 0.10 0.08 0.06 0.23 0.21 0.19 0.17 0.13 0.10 0.33 0.32

Water Weight, lb/ft

364

2

11/2

1

3/4

1/2

Nominal Pipe Size, NPS

2785_A003.fm Page 364 Friday, March 11, 2005 5:33 PM

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

8

6

4

3

2.875 2.875 2.875 2.875 3.5 3.5 3.5 3.5 3.5 3.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 6.625 6.625 6.625 6.625 6.625 6.625 6.625 8.625 8.625 8.625 8.625 8.625 8.625 8.625 8.625 8.625 8.625 20 30 40 60 80 100 120 140

40 80 120 160

40 80 120 160

40 80 160

40 80 160

40S 80S

XS

5S 10S

5S 10S 40S 80S

5S 10S 40S 80S

5S 10S 40S 80S

Std

XXS

Std XS

XXS

Std XS

XXS

Std XS

XXS

Std XS

0.203 0.276 0.375 0.552 0.083 0.120 0.216 0.300 0.437 0.600 0.083 0.120 0.237 0.337 0.437 0.531 0.674 0.109 0.134 0.280 0.432 0.562 0.718 0.864 0.109 0.148 0.250 0.277 0.322 0.406 0.500 0.593 0.718 0.812

2.469 2.323 2.125 1.771 3.334 3.260 3.068 2.900 2.626 2.300 4.334 4.260 4.026 3.826 3.626 3.438 3.152 6.407 6.357 6.065 5.761 5.501 5.189 4.897 8.407 8.329 8.125 8.071 7.981 7.813 7.625 7.439 7.189 7.001

4.7853 4.2361 3.5448 2.4621 8.7257 8.3427 7.3889 6.6019 5.4133 4.1527 14.7451 14.2459 12.7238 11.4910 10.3211 9.2786 7.7991 32.2240 31.7230 28.8756 26.0535 23.7549 21.1367 18.8248 55.4820 54.4572 51.8223 51.1357 50.0016 47.9187 45.6404 43.4409 40.5702 38.4760

0.75 0.75 0.75 0.75 0.92 0.92 0.92 0.92 0.92 0.92 1.18 1.18 1.18 1.18 1.18 1.18 1.18 1.73 1.73 1.73 1.73 1.73 1.73 1.73 2.26 2.26 2.26 2.26 2.26 2.26 2.26 2.26 2.26 2.26

0.0332 0.0294 0.0246 0.0171 0.0606 0.0579 0.0513 0.0458 0.0376 0.0288 0.1024 0.0989 0.0884 0.0798 0.0717 0.0644 0.0542 0.2238 0.2203 0.2005 0.1809 0.1650 0.1468 0.1307 0.3853 0.3782 0.3599 0.3551 0.3472 0.3328 0.3169 0.3017 0.2817 0.2672

5.79 7.66 10.01 13.69 3.03 4.33 7.58 10.25 14.30 18.58 3.92 5.61 10.79 14.98 18.96 22.51 27.54 7.59 9.29 18.97 28.57 36.39 45.30 53.16 9.91 13.40 22.36 24.70 28.55 35.64 43.39 50.87 60.63 67.76

APPENDIX C

(continued )

0.28 0.25 0.21 0.14 0.51 0.48 0.43 0.38 0.31 0.24 0.85 0.83 0.74 0.67 0.60 0.54 0.45 1.87 1.84 1.67 1.51 1.38 1.22 1.09 3.21 3.15 3.00 2.96 2.90 2.78 2.64 2.52 2.35 2.23

2785_A003.fm Page 365 Friday, March 11, 2005 5:33 PM

365

Copyright 2005 by Taylor & Francis Group, LLC

8.625 8.625 10.75 10.75 10.75 10.75 10.75 10.75 10.75 10.75 10.75 10.75 10.75 10.75 12.75 12.75 12.75 12.75 12.75 12.75 12.75 12.75 12.75 12.75 12.75 12.75 12.75 14.00 14.00 14.00 14.00 14.00

Outside Dia, in.

10 20 30

60 80 100 120 140 160

40

20 30

30 40 60 80 100 120 140 160

20

160

a

Std

XS

Std

Std XS

XXS

5S 10S

80S

40S

5S 10S

40S 80S

5S 10S

Schedule b c 0.875 0.906 0.134 0.165 0.250 0.279 0.307 0.365 0.500 0.593 0.718 0.843 1.000 1.125 0.156 0.180 0.250 0.330 0.375 0.406 0.500 0.562 0.687 0.843 1.000 1.125 1.312 0.156 0.188 0.250 0.312 0.375

Wall Thickness, in. 6.875 6.813 10.482 10.420 10.250 10.192 10.136 10.020 9.750 9.564 9.314 9.064 8.750 8.500 12.438 12.390 12.250 12.090 12.000 11.938 11.750 11.626 11.376 11.064 10.750 10.500 10.126 13.688 13.624 13.500 13.376 13.250

Inside Dia, in. 37.1035 36.4373 86.2498 85.2325 82.4741 81.5433 80.6497 78.8143 74.6241 71.8040 68.0992 64.4925 60.1016 56.7163 121.4425 120.5070 117.7991 114.7420 113.0400 111.8749 108.3791 106.1036 101.5895 96.0935 90.7166 86.5463 80.4907 147.0787 145.7065 143.0663 140.4501 137.8166

Inside Area, in.2 2.26 2.26 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.34 3.67 3.67 3.67 3.67 3.67

Surface Area, ft2/ft 0.2577 0.2530 0.5990 0.5919 0.5727 0.5663 0.5601 0.5473 0.5182 0.4986 0.4729 0.4479 0.4174 0.3939 0.8434 0.8369 0.8181 0.7968 0.7850 0.7769 0.7526 0.7368 0.7055 0.6673 0.6300 0.6010 0.5590 1.0214 1.0119 0.9935 0.9754 0.9571

Volume, ft3/ft 72.42 74.69 15.19 18.65 28.04 31.20 34.24 40.48 54.74 64.33 76.93 89.20 104.13 115.64 20.98 24.16 33.38 43.77 49.56 53.52 65.42 73.15 88.51 107.20 125.49 139.67 160.27 23.07 27.73 36.71 45.61 54.57

Pipe Weight, lb/ft 2.15 2.11 5.00 4.94 4.78 4.72 4.67 4.57 4.32 4.16 3.94 3.74 3.48 3.29 7.03 6.98 6.82 6.65 6.55 6.48 6.28 6.15 5.88 5.57 5.26 5.01 4.66 8.52 8.44 8.29 8.14 7.98

Water Weight, lb/ft

366

14

12

10

Nominal Pipe Size, NPS

2785_A003.fm Page 366 Friday, March 11, 2005 5:33 PM

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

18

16

14.00 14.00 14.00 14.00 14.00 14.00 14.00 14.00 14.00 14.00 14.00 14.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 16.00 18.00 18.00 18.00 XS

40

10

100 120 140 160

80

60

Std

XS

10 20 30

100 120 140 160

80

60

40

5S 10S

5S 10S

0.437 0.500 0.562 0.593 0.625 0.687 0.750 0.875 0.937 1.093 1.250 1.406 0.165 0.188 0.250 0.312 0.375 0.437 0.500 0.562 0.625 0.656 0.687 0.750 0.843 0.875 1.031 1.218 1.437 1.593 0.165 0.188 0.250

13.126 13.000 12.876 12.814 12.750 12.626 12.500 12.250 12.126 11.814 11.500 11.188 15.670 15.624 15.500 15.376 15.250 15.126 15.000 14.876 14.750 14.688 14.626 14.500 14.314 14.250 13.938 13.564 13.126 12.814 17.670 17.624 17.500

135.2491 132.6650 130.1462 128.8959 127.6116 125.1415 122.6563 117.7991 115.4263 109.5629 103.8163 98.2595 192.7559 191.6259 188.5963 185.5908 182.5616 179.6048 176.6250 173.7169 170.7866 169.3538 167.9271 165.0463 160.8391 159.4041 152.5003 144.4259 135.2491 128.8959 245.0997 243.8252 240.4063

3.67 3.67 3.67 3.67 3.67 3.67 3.67 3.67 3.67 3.67 3.67 3.67 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.19 4.71 4.71 4.71

0.9392 0.9213 0.9038 0.8951 0.8862 0.8690 0.8518 0.8181 0.8016 0.7609 0.7209 0.6824 1.3386 1.3307 1.3097 1.2888 1.2678 1.2473 1.2266 1.2064 1.1860 1.1761 1.1662 1.1462 1.1169 1.1070 1.0590 1.0030 0.9392 0.8951 1.7021 1.6932 1.6695

63.30 72.09 80.66 84.91 89.28 97.68 106.13 122.65 130.72 150.67 170.21 189.11 27.90 31.75 42.05 52.27 62.58 72.64 82.77 92.66 102.63 107.50 112.35 122.15 136.46 141.34 164.82 192.29 223.50 245.11 31.43 35.76 47.39

APPENDIX C

(continued )

7.83 7.69 7.54 7.47 7.39 7.25 7.11 6.82 6.69 6.35 6.01 5.69 11.17 11.10 10.93 10.75 10.58 10.40 10.23 10.06 9.89 9.81 9.73 9.56 9.32 9.23 8.83 8.37 7.83 7.47 14.20 14.12 13.93

2785_A003.fm Page 367 Friday, March 11, 2005 5:33 PM

367

Copyright 2005 by Taylor & Francis Group, LLC

20

Nominal Pipe Size, NPS

18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00

Outside Dia, in.

XS

30

80 100 120

60

5S 10S

0.312 0.375 0.437 0.500 0.562 0.625 0.687 0.750 0.875 0.937 1.156 1.375 1.562 1.781 0.188 0.218 0.250 0.312 0.375 0.437 0.500 0.562 0.593 0.625 0.687 0.750 0.812 0.875 1.031 1.281 1.500

Wall Thickness, in. 17.376 17.250 17.126 17.000 16.876 16.750 16.626 16.500 16.250 16.126 15.688 15.250 14.876 14.438 19.624 19.564 19.500 19.376 19.250 19.126 19.000 18.876 18.814 18.750 18.626 18.500 18.376 18.250 17.938 17.438 17.000

Inside Dia, in. 237.0114 233.5866 230.2404 226.8650 223.5675 220.2416 216.9927 213.7163 207.2891 204.1376 193.1990 182.5616 173.7169 163.6378 302.3046 300.4588 298.4963 294.7121 290.8916 287.1560 283.3850 279.6982 277.8638 275.9766 272.3384 268.6663 265.0767 261.4541 252.5909 238.7058 226.8650

Inside Area, in.2 4.71 4.71 4.71 4.71 4.71 4.71 4.71 4.71 4.71 4.71 4.71 4.71 4.71 4.71 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24 5.24

Surface Area, ft2/ft 1.6459 1.6221 1.5989 1.5755 1.5526 1.5295 1.5069 1.4841 1.4395 1.4176 1.3417 1.2678 1.2064 1.1364 2.0993 2.0865 2.0729 2.0466 2.0201 1.9941 1.9680 1.9424 1.9296 1.9165 1.8912 1.8657 1.8408 1.8157 1.7541 1.6577 1.5755

Volume, ft3/ft 58.94 70.59 81.97 93.45 104.67 115.98 127.03 138.17 160.03 170.75 207.96 244.14 274.22 308.50 39.78 46.06 52.73 65.60 78.60 91.30 104.13 116.67 122.91 129.33 141.70 154.19 166.40 178.72 208.87 256.10 296.37

Pipe Weight, lb/ft 13.73 13.53 13.34 13.14 12.95 12.76 12.57 12.38 12.01 11.83 11.19 10.58 10.06 9.48 17.51 17.41 17.29 17.07 16.85 16.63 16.42 16.20 16.10 15.99 15.78 15.56 15.36 15.15 14.63 13.83 13.14

Water Weight, lb/ft

368

40

Std

XS

Std

Schedule b c

20

10

80 100 120 140 160

60

40

30

20

a

2785_A003.fm Page 368 Friday, March 11, 2005 5:33 PM

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

26

24

22

20.00 20.00 22.00 22.00 22.00 22.00 22.00 22.00 22.00 22.00 22.00 22.00 22.00 22.00 22.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 26.00 26.00 10

60 80 100 120 140 160

40

30

20

10

60 80 100 120 140 160

10 20 30

140 160

XS

Std

Std XS

5S 10S

5S 10S

1.750 1.968 0.188 0.218 0.250 0.375 0.500 0.625 0.750 0.875 1.125 1.375 1.625 1.875 2.125 0.188 0.218 0.250 0.312 0.375 0.437 0.500 0.562 0.593 0.625 0.812 1.031 1.281 1.500 1.750 1.968 0.250 0.312

16.500 16.064 21.624 21.564 21.500 21.250 21.000 20.750 20.500 20.250 19.750 19.250 18.750 18.250 17.750 23.624 23.564 23.500 23.376 23.250 23.126 23.000 22.876 22.814 22.750 22.376 21.938 21.438 21.000 20.500 20.064 25.500 25.376

213.7163 202.5709 367.0639 365.0298 362.8663 354.4766 346.1850 337.9916 329.8963 321.8991 306.1991 290.8916 275.9766 261.4541 247.3241 438.1033 435.8807 433.5163 428.9533 424.3416 419.8273 415.2650 410.7994 408.5757 406.2866 393.0380 377.8015 360.7765 346.1850 329.8963 316.0128 510.4463 505.4940

5.24 5.24 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.28 6.81 6.81

1.4841 1.4067 2.5491 2.5349 2.5199 2.4616 2.4041 2.3472 2.2910 2.2354 2.1264 2.0201 1.9165 1.8157 1.7175 3.0424 3.0270 3.0105 2.9789 2.9468 2.9155 2.8838 2.8528 2.8373 2.8214 2.7294 2.6236 2.5054 2.4041 2.2910 2.1945 3.5448 3.5104

341.09 379.00 43.80 50.71 58.07 86.61 114.81 142.68 170.21 197.41 250.81 302.88 353.61 403.00 451.06 47.81 55.37 63.41 78.93 94.62 109.97 125.49 140.68 148.24 156.03 201.09 252.91 310.82 360.45 415.85 463.07 68.75 85.60

APPENDIX C

(continued )

12.38 11.73 21.26 21.15 21.02 20.53 20.05 19.58 19.11 18.65 17.74 16.85 15.99 15.15 14.33 25.38 25.25 25.11 24.85 24.58 24.32 24.06 23.80 23.67 23.54 22.77 21.89 20.90 20.05 19.11 18.31 29.57 29.28

2785_A003.fm Page 369 Friday, March 11, 2005 5:33 PM

369

Copyright 2005 by Taylor & Francis Group, LLC

26.00 26.00 26.00 26.00 26.00 26.00 26.00 28.00 28.00 28.00 28.00 28.00 28.00 28.00 28.00 28.00 30.00 30.00 30.00 30.00 30.00 30.00 30.00 30.00 30.00 32.00 32.00 32.00 32.00 32.00 32.00

Outside Dia, in.

20 30 40

10

20 30 40

10

20 30

10

20

a

Std XS

Std XS

Std XS

Std XS

5S 10S

Schedule b c 0.375 0.500 0.625 0.750 0.875 1.000 1.125 0.250 0.312 0.375 0.500 0.625 0.750 0.875 1.000 1.125 0.250 0.312 0.375 0.500 0.625 0.750 0.875 1.000 1.125 0.250 0.312 0.375 0.500 0.625 0.688

Wall Thickness, in. 25.250 25.000 24.750 24.500 24.250 24.000 23.750 27.500 27.376 27.250 27.000 26.750 26.500 26.250 26.000 25.750 29.500 29.376 29.250 29.000 28.750 28.500 28.250 28.000 27.750 31.500 31.376 31.250 31.000 30.750 30.624

Inside Dia, in. 500.4866 490.6250 480.8616 471.1963 461.6291 452.1600 442.7891 593.6563 588.3146 582.9116 572.2650 561.7166 551.2663 540.9141 530.6600 520.5041 683.1463 677.4153 671.6166 660.1850 648.8516 637.6163 626.4791 615.4400 604.4991 778.9163 772.7959 766.6016 754.3850 742.2666 736.1961

Inside Area, in.2 6.81 6.81 6.81 6.81 6.81 6.81 6.81 7.33 7.33 7.33 7.33 7.33 7.33 7.33 7.33 7.33 7.85 7.85 7.85 7.85 7.85 7.85 7.85 7.85 7.85 8.38 8.38 8.38 8.38 8.38 8.38

Surface Area, ft2/ft 3.4756 3.4071 3.3393 3.2722 3.2058 3.1400 3.0749 4.1226 4.0855 4.0480 3.9741 3.9008 3.8282 3.7564 3.6851 3.6146 4.7441 4.7043 4.6640 4.5846 4.5059 4.4279 4.3506 4.2739 4.1979 5.4092 5.3667 5.3236 5.2388 5.1546 5.1125

Volume, ft3/ft 102.63 136.17 169.38 202.25 234.79 267.00 298.87 74.09 92.26 110.64 146.85 182.73 218.27 253.48 288.36 322.90 79.43 98.93 118.65 157.53 196.08 234.29 272.17 309.72 346.93 84.77 105.59 126.66 168.21 209.43 230.08

Pipe Weight, lb/ft 28.99 28.42 27.86 27.30 26.74 26.19 25.65 34.39 34.08 33.77 33.15 32.54 31.93 31.33 30.74 30.15 39.57 39.24 38.91 38.24 37.59 36.94 36.29 35.65 35.02 45.12 44.77 44.41 43.70 43.00 42.65

Water Weight, lb/ft

370

32

30

28

Nominal Pipe Size, NPS

2785_A003.fm Page 370 Friday, March 11, 2005 5:33 PM

GAS PIPELINE HYDRAULICS

Copyright 2005 by Taylor & Francis Group, LLC

42

36

34

32.00 32.00 32.00 32.00 34.00 34.00 34.00 34.00 34.00 34.00 34.00 34.00 34.00 34.00 36.00 36.00 36.00 36.00 36.00 36.00 36.00 36.00 36.00 42.00 42.00 42.00 42.00 42.00 42.00 42.00 42.00 20 30 40

20 30 40

10

20 30 40

10

Std XS

Std XS

Std XS

0.750 0.875 1.000 1.125 0.250 0.312 0.375 0.500 0.625 0.688 0.750 0.875 1.000 1.125 0.250 0.312 0.375 0.500 0.625 0.750 0.875 1.000 1.125 0.250 0.375 0.500 0.625 0.750 1.000 1.250 1.500

30.500 30.250 30.000 29.750 33.500 33.376 33.250 33.000 32.750 32.624 32.500 32.250 32.000 31.750 35.500 35.376 35.250 35.000 34.750 34.500 34.250 34.000 33.750 41.500 41.250 41.000 40.750 40.500 40.000 39.500 39.000

730.2463 718.3241 706.5000 694.7741 880.9663 874.4565 867.8666 854.8650 841.9616 835.4954 829.1563 816.4491 803.8400 791.3291 989.2963 982.3972 975.4116 961.6250 947.9366 934.3463 920.8541 907.4600 894.1641 1351.9663 1335.7266 1319.5850 1303.5416 1287.5963 1256.0000 1224.7963 1193.9850

8.38 8.38 8.38 8.38 8.90 8.90 8.90 8.90 8.90 8.90 8.90 8.90 8.90 8.90 9.42 9.42 9.42 9.42 9.42 9.42 9.42 9.42 9.42 11.00 11.00 11.00 11.00 11.00 11.00 11.00 11.00

5.0712 4.9884 4.9063 4.8248 6.1178 6.0726 6.0269 5.9366 5.8470 5.8021 5.7580 5.6698 5.5822 5.4954 6.8701 6.8222 6.7737 6.6780 6.5829 6.4885 6.3948 6.3018 6.2095 9.3887 9.2759 9.1638 9.0524 8.9417 8.7222 8.5056 8.2916

250.31 290.86 331.08 370.96 90.11 112.25 134.67 178.89 222.78 244.77 266.33 309.55 352.44 394.99 95.45 118.92 142.68 189.57 236.13 282.35 328.24 373.80 419.02 111.47 166.71 221.61 276.18 330.41 437.88 544.01 648.81

42.30 41.61 40.93 40.25 51.03 50.66 50.27 49.52 48.77 48.40 48.03 47.30 46.57 45.84 57.31 56.91 56.50 55.71 54.91 54.13 53.34 52.57 51.80 78.32 77.38 76.44 75.51 74.59 72.76 70.95 69.17

2785_A003.fm Page 371 Friday, March 11, 2005 5:33 PM

APPENDIX C 371

2785_A004.fm Page 373 Wednesday, April 13, 2005 2:36 PM

APPENDIX

D

GASMOD Output Report The following is a report from GASMOD—Gas Pipeline Hydraulics Simulation software (www.systek.us)—for a pipeline transporting natural gas in an NPS 16 pipeline 450 miles long from Compton to Harvard. ******* GASMOD - GAS PIPELINE HYDRAULIC SIMULATION ******** ************ 32-bit Version 5.00.200 ************

DATE:

21-September-2004

TIME: 09:45:06

PROJECT DESCRIPTION: Pipeline from Compton to Harvard 16" pipeline Case Number:

1264

Pipeline data file:

C:\GASMOD32\MYPIPE001.TOT

Pressure drop formula: Pipeline efficiency: Compressibility Factor Method: Inlet Gas Gravity(Air=1.0): Inlet Gas Viscosity:

AGA Turbulent 1.00 AGA NX19 0.67883 0.0000070(lb/ft-sec)

CALCULATION OPTIONS: Branch pipe calculations: Loop pipe calculations: Compressor Fuel Calculated: Joule Thompson effect included: Customized Output: Holding Delivery Pressure at terminus

NO NO YES NO NO

373

Copyright 2005 by Taylor & Francis Group, LLC

2785_A004.fm Page 374 Wednesday, April 13, 2005 2:36 PM

374

GAS PIPELINE HYDRAULICS

***************** Calculations Based on Specified Thermal Conductivities of Pipe, Soil and Insulation ************** Origin suction temperature: Base temperature: Base pressure: Origin suction pressure: Delivery pressure: Minimum pressure: Gas specific heat ratio: Maximum gas velocity:

70.00(degF) 60.00(degF) 14.700(psig) 800.00(psig) 500.33(psig) 200.00(psig) 1.26 50.00(ft/sec)

Inlet Flow rate: Outlet Flow rate:

150.0000(MMSCFD) 128.2744(MMSCFD)

**************** PIPELINE PROFILE DATA *********** Distance (mi)

Elevation (ft)

Diameter (in)

Thickness (in)

Roughness (in)

0.00 15.00 20.00 30.00 42.00 45.00 48.90 85.00 128.00 130.00 140.00 150.00 154.00 155.00 160.00 180.00 238.40 240.00 250.00 260.00 290.00 292.00 300.00 320.00 330.00 346.00 350.00 356.00 360.00 400.00 420.00 450.00

620.00 969.70 1086.26 1319.39 1599.14 1669.08 1760.00 2929.00 1260.00 1338.46 1730.77 2123.08 2280.00 2263.53 2181.18 1851.80 890.00 883.18 840.54 797.91 670.00 727.00 955.00 1525.00 1563.33 1624.67 1640.00 1543.40 1479.00 835.00 881.00 950.00

16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000

0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375 0.375

0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700 0.000700

Copyright 2005 by Taylor & Francis Group, LLC

2785_A004.fm Page 375 Wednesday, April 13, 2005 2:36 PM

APPENDIX D

375

******** THERMAL CONDUCTIVITY AND INSULATION DATA ********

Distance (mi)

Cover (in)

0.000 15.000 20.000 30.000 42.000 45.000 48.900 85.000 128.000 130.000 140.000 150.000 154.000 155.000 160.000 180.000 238.400 240.000 250.000 260.000 290.000 292.000 300.000 320.000 330.000 346.000 350.000 356.000 360.000 400.000 420.000 450.000

36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000 36.000

Thermal Conductivity (Btu/hr/ft/degF) Pipe Soil Insulation 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000 29.000

****************

0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.800 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750

Insul.Thk (in)

Soil Temp (degF)

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00

0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200

COMPRESSOR STATION DATA **************

FLOW RATES, PRESSURES AND TEMPERATURES:

Name Compton Sta-2 Sta-4 Sta-5

Flow Suct. Disch. Suct. Disch. Suct. Disch. MaxPipe Rate Press. Press. Compr. Loss. Loss. Temp. Temp Temp (MMSCFD) (psig) (psig) Ratio (psig) (psig) (degF) (degF) (degF) 149.39 98.90 98.45 98.27

Copyright 2005 by Taylor & Francis Group, LLC

795.00 752.05 768.45 953.88

1210.00 1210.00 1210.00 1149.13

1.5125 1.5973 1.5638 1.2016

5.00 5.00 5.00 5.00

10.00 10.00 10.00 10.00

70.00 80.00 80.00 80.00

129.17 148.62 145.37 106.12

140.00 140.00 140.00 140.00

2785_A004.fm Page 376 Wednesday, April 13, 2005 2:36 PM

376

GAS PIPELINE HYDRAULICS

INACTIVE COMPRESSOR STATIONS:

Name

Distance (mi)

Dimpton Jackson

180 420

********* COMPRESSOR EFFICIENCY, HP AND FUEL USED *********

Name Compton Sta-2 Sta-4 Sta-5

Distance (mi)

Compr Effy. (%)

Mech. Effy. (%)

Overall Effy. (%)

0.00 85.00 250.00 330.00

80.00 80.00 80.00 80.00

98.00 98.00 98.00 98.00

78.40 78.40 78.40 78.40

Horse Power

Fuel Factor (MCF/ day/HP)

Fuel Used (MMSCFD)

3,059.55 2,419.97 2,284.08 864.48

0.2000 0.2000 0.2000 0.2000

0.6119 0.4840 0.4568 0.1729

Total Compressor Station Horsepower:

8,628.08

Total Fuel consumption:

1.7256(MMSCFD)

**************** LOCATIONS AND FLOW RATES **************** Flow Distance in/out Viscosity Pressure GasTemp. Location (mi) (MMSCFD) Gravity (lb/ft-sec) (psig) (degF) GasName Compton

0.00

150.0000 0.6788

0.00000700

1200.00

129.17

Harvard

30.00 350.00 450.00

-50.0000 0.6788 30.0000 0.6000 -128.2744 0.6606

0.00000700 0.00000800 0.00000723

986.29 1083.08 500.33

83.52 82.05 80.00

SAN JUAN GAS

****** REYNOLD'S NUMBER AND HEAT TRANSFER COEFFICIENT ***** HeatTransCoeff CompressibilityDistance FrictFactor Transmission (Btu/hr/ Factor (mi) Reynold'sNum. (Darcy) Factor ft2/degF) (AGA NX19) 0.000 15.000 20.000 30.000 42.000 45.000 48.900 85.000 128.000 130.000 140.000

12,831,042. 12,831,042. 12,831,042. 8,536,437. 8,536,437. 8,536,437. 8,536,437. 8,493,784. 8,493,784. 8,493,784. 8,493,784.

Copyright 2005 by Taylor & Francis Group, LLC

0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104

19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63

0.4999 0.4999 0.4999 0.4992 0.4992 0.4992 0.4992 0.4991 0.4991 0.4991 0.4991

0.8027 0.7887 0.7913 0.7979 0.8024 0.8052 0.8241 0.7672 0.7518 0.7574 0.7670

2785_A004.fm Page 377 Wednesday, April 13, 2005 2:36 PM

APPENDIX D

150.000 154.000 155.000 160.000 180.000 238.400 240.000 250.000 260.000 290.000 292.000 300.000 320.000 330.000 346.000 350.000 356.000 360.000 400.000 420.000 450.000

377

8,493,784. 8,493,784. 8,493,784. 8,493,784. 8,493,784. 8,493,784. 8,493,784. 8,451,888. 8,451,888. 8,451,888. 8,451,888. 8,451,888. 8,451,888. 8,435,063. 8,435,063. 10,716,727. 10,716,727. 10,716,727. 10,716,727. 10,716,727. 10,716,727.

0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104 0.0104

19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63 19.63

0.4991 0.4991 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4681 0.4685 0.4685 0.4685 0.4685 0.4685 0.4685

0.7738 0.7761 0.7777 0.7849 0.8094 0.8298 0.8345 0.8069 0.7629 0.7635 0.7676 0.7797 0.7923 0.7808 0.7706 0.7892 0.7925 0.8124 0.8451 0.8805 0.8805

******** PIPELINE TEMPERATURE AND PRESSURE PROFILE ******** Distance Diameter Flow Velocity Press. GasTemp. SoilTemp. (mi) (in) (MMSCFD) (ft/sec) (psig) (degF) (degF)

MAOP (psig) Location

0.00 15.00 20.00 30.00 42.00 45.00 48.90 85.00

16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000

149.3881 149.3881 149.3881 99.3881 99.3881 99.3881 99.3881 98.9041

16.53 18.08 18.68 13.34 13.97 14.15 14.38 17.22

1200.00 1095.78 1060.08 986.29 940.95 929.42 914.30 757.05

129.17 93.61 88.70 83.52 80.68 80.45 80.26 80.00

80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00

1400.00 Compton 1400.00 1400.00 1400.00 1400.00 1400.00 Branch1 1400.00 1400.00 Sta-2

85.00 128.00 130.00 140.00 150.00 154.00 155.00 160.00 180.00 238.40 240.00 250.00

16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000 16.000

98.9041 98.9041 98.9041 98.9041 98.9041 98.9041 98.9041 98.9041 98.9041 98.9041 98.9041 98.4473

10.94 11.42 11.50 11.91 12.36 12.55 12.58 12.72 13.34 16.01 16.12 16.78

1200.00 1149.26 1141.23 1100.95 1060.36 1044.01 1041.70 1030.05 981.36 815.32 809.68 773.45

140.00 80.23 80.17 80.04 80.01 80.01 80.01 80.00 80.00 80.00 80.00 80.00

80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00 80.00

1400.00 Sta-2 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 Dimpton 1400.00 1400.00 1400.00 Sta-4

250.00 260.00 290.00 292.00 300.00 320.00 330.00

16.000 16.000 16.000 16.000 16.000 16.000 16.000

98.4473 98.4473 98.4473 98.4473 98.4473 98.4473 98.2744

10.89 11.12 11.85 11.93 12.25 13.17 13.56

1200.00 1175.10 1101.69 1094.39 1065.01 990.04 958.88

140.00 98.11 80.39 80.30 80.11 80.01 80.00

80.00 80.00 80.00 80.00 80.00 80.00 80.00

1400.00 Sta-4 1400.00 1400.00 1400.00 1400.00 1400.00 1400.00 Sta-5

330.00 346.00

16.000 16.000

98.2744 98.2744

11.44 11.91

1139.13 1094.37

106.12 83.45

80.00 80.00

1400.00 Sta-5 1400.00

Copyright 2005 by Taylor & Francis Group, LLC

2785_A004.fm Page 378 Wednesday, April 13, 2005 2:36 PM

378

350.00 356.00 360.00 400.00 420.00 450.00

GAS PIPELINE HYDRAULICS

16.000 16.000 16.000 16.000 16.000 16.000

128.2744 128.2744 128.2744 128.2744 128.2744 128.2744

15.70 16.05 16.30 19.78 23.01 33.47

1083.08 1058.92 1042.45 856.54 734.40 500.33

82.05 81.11 80.74 80.01 80.00 80.00

80.00 80.00 80.00 80.00 80.00 80.00

1400.00 1400.00 1400.00 1400.00 1400.00 Jackson 1400.00 Harvard

************ LINE PACK VOLUMES AND PRESSURES ************** Distance (mi)

Pressure (psig)

Line Pack (million std.cu.ft)

0.00 15.00 20.00 30.00 42.00 45.00 48.90 85.00 128.00 130.00 140.00 150.00 154.00 155.00 160.00 180.00 238.40 240.00 250.00 260.00 290.00 292.00 300.00 320.00 330.00 346.00 350.00 356.00 360.00 400.00 420.00 450.00

1200.00 1095.78 1060.08 986.29 940.95 929.42 914.30 757.05 1149.26 1141.23 1100.95 1060.36 1044.01 1041.70 1030.05 981.36 815.32 809.68 773.45 1175.10 1101.69 1094.39 1065.01 990.04 958.88 1094.37 1083.08 1058.92 1042.45 856.54 734.40 500.33

0.0000 9.6000 3.0086 5.7524 6.4677 1.5581 1.9909 16.4027 23.2986 1.3544 6.5846 6.2702 2.4211 0.5982 2.9652 11.4175 28.9974 0.7002 4.2420 5.1998 20.1395 1.2790 5.0078 11.7426 5.4816 9.2580 2.5092 3.6990 2.4088 21.2972 8.5521 9.6571

Total line pack in main pipeline = 239.8614(million std.cu.ft) Started simulation at: 09:44:17 Finished simulation at: 09:45:06 Time elapsed: 49 seconds DATE: 21-September-2004

Copyright 2005 by Taylor & Francis Group, LLC

2785_A005.fm Page 379 Friday, April 1, 2005 3:06 PM

APPENDIX

E

Summary of Formulas CHAPTER 1 1. Density

ρ=

m V

(1.1)

where r = density of gas m = mass of gas V = volume of gas 2. Gas gravity G=

ρg ρair

(1.2)

where G = gas gravity, dimensionless rg = density of gas rair = density of air G=

Mg 29

where Mg = molecular weight of gas Mair = molecular weight of air = 28.9625

379

Copyright 2005 by Taylor & Francis Group, LLC

(1.4)

2785_A005.fm Page 380 Friday, April 1, 2005 3:06 PM

380

GAS PIPELINE HYDRAULICS

3. Kinematic viscosity

ν=

µ ρ

(1.5)

where, in USCS units, n = kinematic viscosity, ft2/s m = dynamic viscosity, lb/ft-s r = density, lb/ft3 and, in SI units, n = kinematic viscosity, cSt m = dynamic viscosity, cP r = density, kg/m3 4. Viscosity of mixture

µ=

where m = mi = yi = Mi =

(

Σ µi yi Mi

(

Σ yi Mi

)

)

(1.6)

dynamic viscosity of gas mixture dynamic viscosity of gas component i mole fraction or percent of gas component i molecular weight of gas component i

5. Ideal gas law or perfect gas equation PV = nRT where P = V = n = R = T =

(1.8)

absolute pressure, pounds per square inch absolute (psia) gas volume, ft3 number of lb moles as deﬁned in Equation 1.7 universal gas constant, psia ft3/lb mole °R absolute temperature of gas, °R (°F + 460)

6. Absolute pressure Pabs = Pgauge + Patm

(1.10)

P1 V2 = or P1V1 = P2V2 P2 V1

(1.13)

7. Boyle’s law

Copyright 2005 by Taylor & Francis Group, LLC

2785_A005.fm Page 381 Friday, April 1, 2005 3:06 PM

APPENDIX E

381

8. Charles’s law V1 T1 = V2 T2

at constant pressure

(1.14)

P1 T1 = P2 T2

at constant volume

(1.15)

9. Modiﬁed ideal gas equation PV = ZnRT (USCS units) where P = V = Z = T = n = R =

(1.16)

absolute pressure of gas, psia volume of gas, ft3 gas compressibility factor, dimensionless absolute temperature of gas, °R number of lb moles as deﬁned in Equation 1.7 universal gas constant, 10.73 psia ft3/lb mole °R

10. Reduced temperature and reduced pressure Tr =

T Tc

(1.17)

Pr =

P Pc

(1.18)

11. Pseudo-reduced temperature and pseudo-reduced pressure

where P T Tpr Ppr Tpc Ppc

= = = = = =

Tpr =

T Tpc

(1.19)

Ppr =

P Ppc

(1.20)

absolute pressure of gas mixture, psia absolute temperature of gas mixture, °R pseudo-reduced temperature, dimensionless pseudo-reduced pressure, dimensionless pseudo-critical temperature, °R pseudo-critical pressure, psia

12. Apparent molecular weight of gas mixture M a = Σyi Mi

Copyright 2005 by Taylor & Francis Group, LLC

(1.21)

2785_A005.fm Page 382 Friday, April 1, 2005 3:06 PM

382

GAS PIPELINE HYDRAULICS

where Ma = apparent molecular weight of gas mixture yi = mole fraction of gas component i Mi = molecular weight of gas component i 13. Kay’s rule to calculate the average pseudo-critical properties of the gas mixture Tpc = ΣyiTc

(1.22)

Ppc = Σyi Pc

(1.23)

14. Pseudo-critical properties from gas gravity Tpc = 170.491 + 307.344 G

(1.24)

Ppc = 709.604 − 58.718 G

(1.25)

where G = gas gravity (air = 1.00) Tpc = pseudo-critical temperature, °R Ppc = pseudo-critical pressure, psia 15. Supercompressibility factor Z=

1 ( Fpv )2

(1.30)

16. Dranchuk, Purvis, and Robinson method A A ρ5 A7 ρr3 A A A Z = 1 + A + 2 + 33 ρr + A4 + 5 ρr2 + 5 6 r + 2 Tpr Tpr Tpr Tpr Tpr3 1 + A8 ρr2 e( − A8ρr )

(

)

(1.31) where

ρr = and A1 A3 A5 A7 Ppr Tpr

= = = = = =

0.27 Ppr ZTpr

0.31506237; A2 = −1.04670990; −0.57832729; A4 = 0.53530771; −0.61232032; A6 = −0.10488813; 0.68157001; A8 = 0.68446549; pseudo-reduced pressure pseudo-reduced temperature

Copyright 2005 by Taylor & Francis Group, LLC

(1.32)

2785_A005.fm Page 383 Friday, April 1, 2005 3:06 PM

APPENDIX E

383

17. CNGA method Z=

1 Pavg 344 ,400 (10 )1.785G 1 + T f3.825

(1.34)

for the average gas pressure Pavg > 100 psig. For Pavg < 100 psig, Z = 1.00 where Pavg = average gas pressure, psig Tf = average gas temperature, °R G = gas gravity (air = 1.00) 18. Average pressure in a pipe segment Pavg =

2 P13 − P23 3 P12 − P2 2

(1.36)

19. Heating value H m = Σ( yi Hi )

(1.37)

where Hm = gross heating value of mixture, Btu/ft3 yi = mole fraction or percent of gas component i Hi = heating value of gas component, Btu/ft3

CHAPTER 2 1. General Flow equation using friction factor T P 2 − P22 Q = 77.54 b 1 Pb GT f LZf where Q = f = Pb = Tb = P1 = P2 = G = Tf =

0.5

D 2.5

(USCS units)

gas ﬂow rate, measured at standard conditions, ft3/day (SCFD) friction factor, dimensionless base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂowing temperature, °R (460 + °F)

Copyright 2005 by Taylor & Francis Group, LLC

(2.2)

2785_A005.fm Page 384 Friday, April 1, 2005 3:06 PM

384

GAS PIPELINE HYDRAULICS

L = pipe segment length, mi Z = gas compressibility factor at the ﬂowing temperature, dimensionless D = pipe inside diameter, in.

(

)

2 2 T P1 − P2 Q = 1.1494 × 10 −3 b Pb GT f LZf

0.5

D 2.5

(SI units)

(2.3)

where Q = gas ﬂow rate, measured at standard conditions, m3/day f = friction factor, dimensionless Pb = base pressure, kPa Tb = base temperature, K (273 + °C) P1 = upstream pressure, kPa P2 = downstream pressure, kPa G = gas gravity (air = 1.00) Tf = average gas ﬂowing temperature, K (273 + °C) L = pipe segment length, km Z = gas compressibility factor at the ﬂowing temperature, dimensionless D = pipe inside diameter, mm 2. General Flow equation using transmission factor T P 2 − e s P22 Q = 38.77 F b 1 Pb GT f Le Z

Q = 5.747 × 10 −4

0.5

(

D 2.5

2 s 2 T P1 − e P2 F b Pb GT f Le Z

)

(USCS units)

(2.7)

D 2.5

(2.8)

0.5

(SI units)

where the elevation correction is as follows: Le =

L (e s − 1) s

H −H 1 s = 0.0375 G 2 Tf Z where s = H1 = H2 = e =

(2.9)

(USCS units)

elevation adjustment parameter, dimensionless upstream elevation, ft downstream elevation, ft base of natural logarithms

Copyright 2005 by Taylor & Francis Group, LLC

(2.10)

2785_A005.fm Page 385 Friday, April 1, 2005 3:06 PM

APPENDIX E

385

and H −H 1 s = 0.0684 G 2 Tf Z

(SI units)

(2.11)

where H1 = upstream elevation, m H2 = downstream elevation, m 3. The equivalent length Le = j1L1 + j2L2es1 + j3L3es2 + …

(2.13)

Q P ZT u = 0.002122 b2 b D Tb P

(2.28)

4. The gas velocity

where u = gas velocity, ft/s Qb = gas ﬂow rate, measured at standard conditions, ft3/day (SCFD) D = pipe inside diameter, in. Pb = base pressure, psia Tb = base temperature, °R (460 + °F) P = upstream pressure, psia T = upstream gas temperature, °R (460 + °F) Z = gas compressibility factor at upstream conditions, dimensionless In SI units, the gas velocity at any point in a gas pipeline is given by Q P ZT u = 14.7349 b2 b D Tb P where u Qb D Pb Tb P T Z

(SI units)

(2.29)

= gas velocity, m/s = gas ﬂow rate, measured at standard conditions, m3/day = pipe inside diameter, mm = base pressure, kPa = base temperature, K (273 + °C) = pressure, kPa = average gas ﬂowing temperature, K (273 + °C) = gas compressibility factor at the ﬂowing temperature, dimensionless

5. Maximum velocity umax = 100

Copyright 2005 by Taylor & Francis Group, LLC

ZRT 29GP

(USCS units)

(2.31)

2785_A005.fm Page 386 Friday, April 1, 2005 3:06 PM

386

GAS PIPELINE HYDRAULICS

where Z = R = T = G = P =

compressibility factor of gas, dimensionless gas constant = 10.73 ft3 psia/lb-moleR gas temperature, °R gas gravity (air = 1.00) gas pressure, psia

6. Reynolds number P GQ Re = 0.0004778 b Tb µ D

(USCS units)

(2.34)

(SI units)

(2.35)

where Pb = base pressure, psia Tb = base temperature, °R (460 + °F) G = speciﬁc gravity of gas (air = 1.0) Q = gas ﬂow rate, standard ft3/day (SCFD) D = pipe inside diameter, in. m = viscosity of gas, lb/ft-s In SI units, the Reynolds number is P GQ Re = 0.5134 b Tb µ D

where Pb = base pressure, kPa Tb = base temperature, °K (273 + °C) G = speciﬁc gravity of gas (air = 1.0) Q = gas ﬂow rate, m3/day (standard conditions) D = pipe inside diameter, mm m = viscosity of gas, Poise 7. Colebrook-White equation e 2.51 = − 2Log10 + f 3.7 D Re f

1

where f D e Re

= = = =

friction factor, dimensionless pipe inside diameter, in. absolute pipe roughness, in. Reynolds number of ﬂow, dimensionless

Copyright 2005 by Taylor & Francis Group, LLC

for Re > 4000

(2.39)

2785_A005.fm Page 387 Friday, April 1, 2005 3:06 PM

APPENDIX E

387

2.51 = −2Log10 f Re f

1

e = −2Log10 3.7 D f

1

for turbulent ﬂow in smooth pipes

(2.40)

for turbulent ﬂow in fully rough pipes

(2.41)

The transmission factor F is related to the friction factor f as follows: F=

2

f=

4 F2

(2.42)

f

Therefore, (2.43)

where f = friction factor F = transmission factor 8. Colebrook equation in terms of transmission factor F e 1.255F F = −4 Log10 + Re 3.7 D

(2.45)

9. Modiﬁed Colebrook-White equation for turbulent ﬂow using friction factor e 2.825 = −2Log10 + f 3.7 D Re f

1

(2.46)

Modiﬁed Colebrook-White equation in terms of the transmission factor e 1.4125F F = −4 Log10 + Re 3.7 D

(USCS and SI units)

(2.47)

10. AGA equation 3.7 D F = 4 Log10 e

(2.48)

11. Bend index BI =

Copyright 2005 by Taylor & Francis Group, LLC

total degrees of all bends in pipe sectionn total length of pipe section

(2.51)

2785_A005.fm Page 388 Friday, April 1, 2005 3:06 PM

388

GAS PIPELINE HYDRAULICS

12. Weymouth equation T P 2 − e s P22 Q = 433.5E b 1 Pb GT f Le Z where Q = E = Pb = Tb = P1 = P2 = G = Tf = Le = Z = D =

0.5

D 2.667

(in USCS units)

(2.52)

volume ﬂow rate, standard ft3/day (SCFD) pipeline efﬁciency, a decimal value less than or equal to 1.0 base pressure, psia base temperature, °R (460 + °F) upstream pressure, psia downstream pressure, psia gas gravity (air = 1.00) average gas ﬂow temperature, °R (460 + °F) equivalent length of pipe segment, mi gas compressibility factor, dimensionless pipe inside diameter, in.

Weymouth transmission factor F = 11.18(D)1/6

(in USCS units)

(2.53)

Weymouth equation T P 2 − e s P22 Q = 3.7435 × 10 E b 1 Pb GT f Le Z

0.5

−3

D 2.667

(in SI units)

(2.54)

where Q = gas ﬂow rate, standard m3/day Tb = base temperature, K (273 + °C) Pb = base pressure, kPa Tf = average gas ﬂow temperature, K (273 + °C) P1 = upstream pressure, kPa P2 = downstream pressure, kPa Le = equivalent length of pipe segment, km Other symbols are as deﬁned previously. Weymouth transmission factor F = 6.521(D)1/6

(in SI units)

(2.53a)

13. Panhandle A equation T Q = 435.87 E b Pb

1.0788

P 2 − es P 2 1 2 0.8539 T f Le Z G

T Q = 4.5965 × 10 E b Pb −3

Copyright 2005 by Taylor & Francis Group, LLC

1.0788

0.5394

D 2.6182

P 2 − es P 2 1 2 0.8539 T f Le Z G

(USCS units)

(2.55)

0.5394

D 2.6182

(SI units)

(2.56)

2785_A005.fm Page 389 Friday, April 1, 2005 3:06 PM

APPENDIX E

389

The equivalent transmission factor for Panhandle A equation QG F = 7.2111E D

0.07305

(USCS)

(2.57)

(SI)

(2.58)

and in SI units, it is QG F = 11.85E D

0.07305

13. Panhandle B equation T Q = 737 E b Pb

1.02

P 2 − es P 2 1 2 0.961 G T L f eZ

T Q = 1.002 × 10 E b Pb −2

1.02

0.51

D 2.53

P 2 − es P 2 1 2 0.961 G T f Le Z

(USCS units)

(2.59)

0.51

D 2.53 (SI units)

(2.60)

The equivalent transmission factor for Panhandle B equation is QG F = 16.7 E D

0.01961

QG F = 19.08 E D

(USCS units)

(2.61)

0.01961

(SI units)

(2.62)

14. IGT equation T P 2 − e s P22 Q = 136.9 E b 01.8 0.2 Pb G T f Le µ

0.555

D 2.667

T P 2 − e s P22 Q = 1.2822 × 10 E b 01.8 0.2 Pb G T f Le µ

(USCS units)

(2.63)

D 2.667

(2.64)

0.555

−3

(SI units)

15. Spitzglass equation The low pressure (less than or equal to 1 psig) version T P1 − P2 Q = 3.839 × 10 E b 3 . 6 Pb GT f Le Z 1 + D + 0.03D 3

(

)

0.5

D 2.5

(USCS units) (2.65)

Copyright 2005 by Taylor & Francis Group, LLC

2785_A005.fm Page 390 Friday, April 1, 2005 3:06 PM

390

GAS PIPELINE HYDRAULICS

The high pressure (more than 1 psig) version T P12 − e s P22 Q = 729.6087 E b Pb GT f Le Z 1 + 3D.6 + 0.03D

(

0.5

D 2.5

)

(USCS units) (2.67)

The low pressure (less than 6.9 kPa) version T P1 − P2 Q = 5.69 × 10 E b 91 . 44 Pb GT f Le Z 1 + D + 0.0012 D −2

(

0.5

D 2.5

)

(SI units) (2.66)

The high pressure (more than 6.9 kPa) version T P12 − e s P22 Q = 1.0815 × 10 E b Pb GT f Le Z 1 + 91D.44 + 0.0012 D −2

(

0.5

)

D 2.5

(SI units) (2.68)

16. The Mueller equation T P12 − e s P22 Q = 85.7368 E b 0.7391 0.2609 T f Le µ Pb G

0.575

D 2.725

T P12 − e s P22 Q = 3.0398 × 10 E b 0.7391 T f Le µ 0.2609 Pb G

(USCS units)

(2.69)

0.575

−2

D 2.725

(SI units) (2.70)

17. Fritzsche formula T P2 − P2 Q = 410.1688 E b 01.8587 2 T f Le Pb G T P 2 − es P 2 Q = 2.827 E b 10.8587 2 T f Le Pb G

0.538

D 2.69

(USCS units)

(2.71)

0.538

D 2.69

(SI units)

(2.72)

CHAPTER 3 1. Total equivalent length—series piping 5

D D Le = L1 + L2 1 + L3 1 D 2 D3

Copyright 2005 by Taylor & Francis Group, LLC

5

(3.6)

2785_A005.fm Page 391 Friday, April 1, 2005 3:06 PM

APPENDIX E

391

2. Equivalent diameter—parallel pipes 1/ 5

1 + Const1 2 De = D1 Const1

(3.17)

where 5

D L Const1 = 1 2 D2 L1

(3.18)

Flow rates Q1 and Q2 are calculated from Q1 =

QConst1 1 + Const1

(3.19)

Q2 =

Q 1 + Const1

(3.20)

and

3. Temperature proﬁle of gas in a pipe segment T2 = Ts + (T1 − Ts )e −θ

θ= where U ∆L ∆A T1 T2 Ts D m

(3.29)

πUD∆ L mCp

(3.28)

= overall heat transfer coefﬁcient, Btu/h/ft2/°F = length of pipe segment = surface area of pipe for heat transfer = p D∆L = gas temperature upstream of pipe segment, °F = gas temperature downstream of pipe segment, °F = average soil temperature surrounding pipe segment, °F = pipe inside diameter, ft = mass ﬂow rate of gas, lb/s

4. Line pack T P Vb = 28.798 b avg ( D 2 L ) Pb ZavgTavg

Copyright 2005 by Taylor & Francis Group, LLC

(USCS units)

(3.34)

2785_A005.fm Page 392 Friday, April 1, 2005 3:06 PM

392

GAS PIPELINE HYDRAULICS

where Vb = line pack in pipe segment, standard ft3 D = pipe inside diameter, in. L = pipe segment length, mi T P Vb = 7.855 × 10 −4 b avg ( D 2 L) Pb ZavgTavg

(SI units)

(3.35)

where Vb = line pack in pipe segment, standard m3 D = pipe inside diameter, mm L = pipe segment length, km CHAPTER 4 1. Compression ratio r=

Pd Ps

(4.1)

2. Isothermal work done Wi = where Wi G T1 P1 P2 Loge

= = = = = =

P 53.28 T Loge 2 G 1 P1

(4.4)

isothermal work done, ft-lb/lb of gas gas gravity, dimensionless suction temperature of gas, °R suction pressure of gas, psia discharge pressure of gas, psia natural logarithm to base e(e = 2.718)

The ratio

( ) is also called the compression ratio. P2 P1

Wi = where Wi = T1 = P1 = P2 =

(USCS units)

P 286.76 T1Loge 2 G P1

isothermal work done, J/kg of gas suction temperature of gas, K suction pressure of gas, kPa absolute discharge pressure of gas, kPa absolute

Other symbols are as deﬁned earlier.

Copyright 2005 by Taylor & Francis Group, LLC

(SI units)

(4.5)

2785_A005.fm Page 393 Friday, April 1, 2005 3:06 PM

APPENDIX E

393

3. Adiabatic work done γ −1 53.28 γ P2 γ − 1 Wa = T G 1 γ − 1 P1

where Wa G T1 g P1 P2

= = = = = =

(USCS units)

adiabatic work done, ft-lb/lb of gas gas gravity, dimensionless suction temperature of gas, °R ratio of speciﬁc heats of gas, dimensionless suction pressure of gas, psia discharge pressure of gas, psia γ −1 286.76 γ P2 γ Wa = T1 − 1 G γ − 1 P1

where Wa T1 P1 P2

= = = =

(4.8)

(SI units)

(4.9)

adiabatic work done, J/kg of gas suction temperature of gas, K suction pressure of gas, kPa absolute discharge pressure of gas, kPa absolute

Other symbols are as deﬁned earlier. 4. Horsepower γ −1 Z1 + Z 2 1 P2 γ γ − 1 HP = 0.0857 QT γ − 1 1 2 ηa P1

where HP g Q T1 P1 P2 Z1 Z2 ha

= compression horsepower = ratio of speciﬁc heats of gas, dimensionless = gas ﬂow rate, MMSCFD = suction temperature of gas, °R = suction pressure of gas, psia = discharge pressure of gas, psia = compressibility of gas at suction conditions, dimensionless = compressibility of gas at discharge conditions, dimensionless = compressor adiabatic (isentropic) efﬁciency, decimal value

Copyright 2005 by Taylor & Francis Group, LLC

(4.15)

2785_A005.fm Page 394 Friday, April 1, 2005 3:06 PM

394

GAS PIPELINE HYDRAULICS

In SI units, the Power equation is as follows: γ −1 Z1 + Z 2 1 P2 γ γ − Power = 4.0639 QT 1 γ − 1 1 2 ηa P1

where Power g Q T1 P1 P2 Z1 Z2 ha

= = = = = = = = =

(4.16)

compression power, kW ratio of speciﬁc heats of gas, dimensionless gas ﬂow rate, Mm3/day (standard) suction temperature of gas, K suction pressure of gas, kPa discharge pressure of gas, kPa compressibility of gas at suction conditions, dimensionless compressibility of gas at discharge conditions, dimensionless compressor adiabatic (isentropic) efﬁciency, decimal value BHP =

HP ηm

(4.17)

5. Compression ratio 1

r = (rt ) n

(4.25)

where r = compression ratio, dimensionless rt = overall compression ratio, dimensionless n = number of compressors in series CHAPTER 6 1. Barlow’s equation Sh = where Sh = P = D = t =

PD 2t

(6.1)

hoop or circumferential stress in pipe material, psi internal pressure, psi pipe outside diameter, in. pipe wall thickness, in.

Axial or longitudinal stress Sa =

Copyright 2005 by Taylor & Francis Group, LLC

PD 4t

(6.2)

2785_A005.fm Page 395 Friday, April 1, 2005 3:06 PM

APPENDIX E

395

2. Internal design pressure P=

2tSEFT D

(6.8)

where P = internal pipe design pressure, psig D = pipe outside diameter, in. t = pipe wall thickness, in. S = speciﬁed minimum yield strength (SMYS) of pipe material, psig E = seam joint factor, 1.0 for seamless and submerged arc welded (SAW) pipes F = design factor, usually 0.72 for cross-country gas pipelines, but can be as low as 0.4, depending on class location and type of construction T = temperature deration factor = 1.00 for temperatures below 250°F 3. Blowdown calculations 1

1

0.0588 P1 3G 2 D 2 LFc T= d2 where T = P1 = G = D = L = d = Fc =

(USCS units)

(6.9)

(SI units)

(6.10)

blowdown time, min initial pressure, psia gas gravity (air = 1.00) pipe inside diameter, in. length of pipe section, mi inside diameter of blowdown pipe, in. choke factor (as follows)

Choke factor list Ideal nozzle = 1.0 Through gate = 1.6 Regular gate = 1.8 Regular lube plug = 2.0 Venturi lube plug = 3.2

In SI units, 1

T=

where P1 = initial pressure, kPa D = pipe inside diameter, mm

Copyright 2005 by Taylor & Francis Group, LLC

1

0.0192 P1 3G 2 D 2 LFc d2

2785_A005.fm Page 396 Friday, April 1, 2005 3:06 PM

396

GAS PIPELINE HYDRAULICS

L = length of pipe section, km d = pipe inside diameter of blowdown, mm Other symbols are as deﬁned before. 4. Pipe weight w = 10.68 × t × (D − t)

(USCS units)

(6.11)

where w = pipe weight, lb/ft D = pipe outside diameter, in. t = pipe wall thickness, in. w = 0.0246 × t × (D − t)

(SI units)

(6.12)

2[( p1 − p2 )/ρ + g( z1 − z2 )] 1 − Cc2 ( Ao / A)2

(9.2)

where w = pipe weight, kg/m D = pipe outside diameter, mm t = pipe wall thickness, mm

CHAPTER 9 1. The discharge through the oriﬁce meter Q = CcCv Ao where Q = Cc = Cv = Ao = A = p1 = p2 = r = z1 = z2 = g =

ﬂow rate, ft3/s contraction coefﬁcient, dimensionless discharge coefﬁcient, dimensionless cross-sectional area of the oriﬁce, in.2 cross-sectional area of pipe containing the oriﬁce, in.2 upstream pressure, psig downstream pressure, psig density of gas, lb/ft3 upstream elevation, ft downstream elevation, ft acceleration due to gravity

2. Sharp-crested oriﬁce 5

A 2 Cc = 0.595 + 0.29 o A

Copyright 2005 by Taylor & Francis Group, LLC

(9.4)

2785_A005.fm Page 397 Friday, April 1, 2005 3:06 PM

APPENDIX E

397

3. The fundamental oriﬁce meter ﬂow equation described in the ANSI 2530/AGA Report No. 3 is as follows: qm =

π C Y d 2 (2 gρ f ∆ P)0.5 4 0.5 4 (1 − β )

(9.6)

or qm = KY

π 2 d (2 gρ f ∆ P)0.5 4 d D

(9.8)

C CD 2 = (1 − β 4 )0.5 ( D 4 − d 4 )0.5

(9.9)

β= K= where qm rf C b d D Y g ∆P K

= = = = = = = = = =

(9.7)

mass ﬂow rate of gas, lb/s density of gas, lb/ft3 discharge coefﬁcient beta ratio, dimensionless oriﬁce diameter, in. meter tube diameter, in. expansion factor, dimensionless acceleration due to gravity, ft/s2 pressure drop across the oriﬁce, psi ﬂow coefﬁcient, dimensionless

4. Buckingham and Bean equation endorsed by the National Bureau of Standards (NBS) and listed in AGA Report No.3 For ﬂange taps: 1 0.5 0.007 0.076 + 0.364 + 0.5 β 4 + 0.4 1.6 − 0.07 + − β D D D D 5

Ke = 0.5993 +

65 0.034 − 0.009 + (0.5 − β )1.5 + 2 + 3 (β − 0.7)2.5 D D where Ke = D = d = b =

ﬂow coefﬁcient for Reynolds number Rd = d(106/15), dimensionless meter tube diameter, in. oriﬁce diameter, in. beta ratio, dimensionless

Copyright 2005 by Taylor & Francis Group, LLC

2.5

(9.12)

2785_A005.fm Page 398 Friday, April 1, 2005 3:06 PM

398

GAS PIPELINE HYDRAULICS

For pipe taps: Ke = 0.5925 +

0.0182 0.06 2 0.225 5 + 0.440 − β + 0.935 + β D D D

+ 1.35β 14 +

1.43 (0.25 − β )2.5 D 0.5

(9.13)

where all symbols are as deﬁned before. 5. Expansion factor For ﬂange taps: Y1 = 1 − (0.41 + 0.35β 4 )

x1 k

(9.20)

For pipe taps: Y1 = 1 − [0.333 + 1.145(β 2 + 0.7β 5 + 12β 13 )]

x1 k

(9.21)

and the pressure ratio x1 is x1 =

Pf 1 − Pf 2 Pf 1

=

hw 27.707 Pf 1

(9.22)

where Y1 = expansion factor based on upstream pressure x1 = ratio of differential pressure to absolute upstream static pressure hw = differential pressure between upstream and downstream taps in in. of water at 60°F Pf1 = static pressure at upstream tap, psia Pf2 = static pressure at downstream tap, psia x1/k = acoustic ratio, dimensionless k = ratio of speciﬁc heats of gas, dimensionless 6. Supercompressibility factor C = Fb Fr Fpb Ftb Ftf Fgr FpvY where the dimensionless factors are Fb = basic oriﬁce factor Fr = Reynolds number factor Fpb = pressure base factor Ftb = temperature base factor Ftf = ﬂowing temperature factor Fgr = gas relative density factor Fpv = supercompressibility factor Y = expansion factor

Copyright 2005 by Taylor & Francis Group, LLC

(9.34)

2785_A005.fm Page 399 Friday, April 1, 2005 3:06 PM

APPENDIX E

399

CHAPTER 10 1. Present value PV = where PV R i n

= = = =

1 R 1− i (1 + i)n

(10.1)

present value, $ series of cash ﬂows, $ interest rate, decimal value number of periods, years

2. Pipe material cost PMC = where PMC L D T C

= = = = =

10.68( D − T )TLC × 5280 2000

(10.2)

pipe material cost, $ length of pipe, mi pipe outside diameter, in. pipe wall thickness, in. pipe material cost, $/ton

In SI units, PMC = 0.0246( D − T )TLC where PMC L D T C

= = = = =

pipe material cost, $ length of pipe, km pipe outside diameter, mm pipe wall thickness, mm pipe material cost, $/metric ton

Copyright 2005 by Taylor & Francis Group, LLC

(10.3)

Our partners will collect data and use cookies for ad personalization and measurement. Learn how we and our ad partner Google, collect and use data. Agree & close