[BR (OIL EXPLORATION) COMPANY OF NEW ZEALAND LiMiTED, P.O. BOX 892 WELLINGTON. NEW ZEALANDGAS CONDITIONING AND PROCESSING VOLUME 4 GAS AND LIQUID SWEETENING By ROBERT N, MADDOX Leonard F. Sheerar Professor Oklahoma State University Edited for Campbell Petroleum Series by John M. Campbell ©Copyright 1982 by Campbell Petroleum Series. This book is sold for personal or library reference use, or for regularly scheduled university classes only; use as a reference or manual for adult training programs is specifically reserved for John M. Campbell and Co. No part of the text or original figures may be reproduced in any form without written permission of the publisher. All translation rights are likewise reserved.GAS CONDITIONING AND PROCESSING Volume 4 GAS AND LIQUID SWEETENING Third Edition No part of the text or original figures may be reproduced in any form without written permission of the publisher, except for bona fide reviews of the book for general publication. Library of Congress Catalog Card No.: 73-91966 First Printing, April 1982 Published by: Campbell Petroleum Series 1215 Crossroads Blvd. Norman, Oklahoma 73069 Printed ané Bound in the US.AChapter TABLE OF CONTENTS INTRODUCTION... - Toxicity of hydrogen sulfide, 6; gas specifications, 7. BASIC PROCESS PRINCIPLES... 1... eee eee eee eee es B Basic dimensions, 13; mass and energy balances, 16; fluid flow, 183 pump power, 31; heat transfer, 33; vapor-liquid equilibrium, 38; absorption and stripping, 41; vessel sizing, 43; gas P-V-T behavior, 45; water content, 51. AMINE PROCESSES... . . Moncethanolamine (MEA), 625 diethanolamine (DEA), 635 triethanolamine (TEA), 63; diglycolamine (DGA), 63; diisopropanolanine (DIPA), 64; methyldiethanolamine (MDEA), 64; solution concentrations, 65; corrosion, 715 solution degradation, 71; foaming, 75; amine reclaiming, 753 filtration, 76; foam inhibitors, 78; corrosion inhibitors, 79; inlet scrubbing, 79; amine losses, 795 filtration, 79; amine~amine HEX, 61} amine regeneration, 81; piping design, 84. AMINE PROCESS CALCULATIONS . 22... eee eee ee ee ee ee 87 Shortcut calculations, 87; monoethanolamine, 945 diethanolamine, 105; factors affecting MEA performance, 1153 stripper design, 118; ethanolamine-~acid gas equilibrium, 120. CARBONATE PROCESSES ©. ee eee ee ee ee ee 129 Reactions, 129; split-stream process, 133; two-stage process, 133; other process modifications, 135; process choice, 136; corrosion, 137; CO, absorption calculations, 138; ¢atacarb process, 143; Benfield’process, 1463 DEA~~hot carbonate, 146; Gianmarco-Vetrocoke proccesses, 1473 seaboard process, 152; vacuum carbonate process, 154; tripotassium phosphate process, 155; sodium phenoline process, 157; alkazid process, 158; hot carbonate--acid gas equilibrium, 159. PHYSICAL ABSORPTION METHODS . . . . see 169 Water absorption, 1693 Fluor solvent process, 1735 equilibrium data for propylene carbonate, 176; Selexol process, 180; Rectisol~Purisol processes, 189; Estasolvan process, 198; other physical solvents, 201; power recovery turbines, 211.Chapter 7. SOLID BED SWEETENING ©... 22... eee eee ee ee ee Iron oxide (sponge) processes, 215; molecular sieves, 221; mercaptan removal, 226; EFCO process, 227; formation of carbonyl sulfide, 235; chemsweet, 237; NCA process, 239; slurrisweet, 239. LIQUID SHEETENING © oe ee eee eee ee eee ML COS removal, 241; hydrogen sulfide removal, 245; amine Process, 246; molecular sieve processes, 247; doctor treatment for mercaptans, 249; caustic wash, 249; regenerated caustic wash, 251; unisol process, 256; mercapsol process, 257; tannin solutizer process, 258; dualayer process, 258; prediction of mercaptan removal, 258; mercapex process, 261; copper sweetening, 261; solid copper treating, 263; Merox process, 263; Bender process, 265. 9, SULFUR PRODUCTION. 2. ke ee ee ee ee ee 287 Liquid oxidation chemistry, 274; Ferrox process, 275; Gluud process, 276; Manchester process, 276; Stretford A.D.A. process, 277; Takahax process, 281; Thylox Process, 281; Perox process, 283; Giammarco-Vetrocoke process, 284; Shell sulfolane process, 288; Freeport sulfur process, 288; Claus process, 288; Amoco process, 294; summary of Claus reactions, 297; selectox process, 306. 10. TAIL GAS CONDITIONING. 2... ee eee ee eee eee 8B ASR sulphoxide process, 320; U. S. Bureau of Mines citrate process, 320; Westvaco process, 321; Sulfreen process, 322; TFP process, 323; Aquaclaus process, 325; cold bed adsorption, 327; Wellman-Lord process, 328; LUCAS process, 3325 Allied $0, process, 333; Pritchard "Cleanair" process, 334; Beavon process, 335; Shell Claus off-gas treating process, 336; Flue ges desulfurization (Hep), 338. APPENDIX 1 -- CONVERSION FACTORS 2... eee ee eee ee ee eee ee 340 APPENDIX 2 -- MEA, DEA and DGA PROPERTIES... 2... 1... ee ee ee 806 INDEX Soi te cee dty Pre eet ee eee eeeeead paeeetne ett a70:LIST OF FIGURES Western world sulfur consumption by end use... . ~~~ aieeeiee U.S. sulfur production trends... . 1. U.S. sulfur supply/demand... 2. : Sulfur recovery from natural gas. . +.) eee ee ee eee te Removal of carbon dioxide only... ee eee ee eee Removal of hydrogen sulfide only... . « eS eee eae cance Simultaneous removal of HS and COp-- ++ ee eee ee Selective removal of HS ee ee ee ee eee ee ee Density of MEA solutions ©. - sll lscvseecerereee Density of DEA solutions . see... wee ee ee pee ee Friction factor as a function of Reynolds Nunber Peeieaare Viscosity of aqueous MEA solutions as function of weight percent MEA Viscosity of aqueous DEA solutions as function of weight percent DEA Nomograph for finding line pressure loss. - + eee eee ee eee Density of aqueous DGA solutions .... +. esse eee eee Viscosity of aqueous DGA solutions .....- SHEESH eH Pump horsepower requirements «s+ + ee ee ee ee eee ee Nomograph for estimating overall heat transfer coefficients . 2... Generalized correlation of the fugacity function in the form of a fugacity coefficient... eee ete tee eee Absorption factor chart . . 2. 6 ee ee ee ee ee ee Compressibility factors for natural gas 27. eee 1 Compressibility factors at low reduced pressures... ... +. += Conpressibility factors for gases near atmospheric pressure»... Pseudo critical temperature adjustment factor, c, °R see ww ee Water content of saturated natural gas (at its water dew point) . Effective water content of saturated hydrogen sulfide in natural gas” systems 2... 2... See eefeeeeettereeisteede ae teat ttetec Effective water content of saturated carbon dioxide in natural gas iiisetupestactenrte aistea tists eeu re ueiiseceistenatiscee eee area rae tee Water content of natural gas mixtures... 0. ee ee eee ee Water content of natural gas mixtures... ee ee ee ee Water content of natural gas mixtures...) ee eee eee Chemical structural formulas for alkanolamines of interest in gas sweetening. ee ee Seer eee Process flow for typical ethanolamine sweetening unit... 2... + Flow sheet of split-stream amine process «1. eee eee eee NEA process flow sheet showing typical temperatures»... 5. se DGA process flow sheet showing typical temperatures... .... : Normality of aqueous amine solutions ... +... s eee pH of MEA solutions at various temperatures... . +. + + pH of DEA solutions at various temperatures... ... - pH of DGA solutions at various temperatures... . Effect of acid gas loading on pH of MEA solutions . . . « ‘ Graded bed activated carbon filter... +--+ eee rene Deep bed activated carbon filter»... 5-2 essa 5 Series installation of filters . 6-0 ee eee ee ee ee Typical temperature distribution in an amine reboiler Temperature profile across the wall of an amine firetubePage Density of MEA solutions as a function of temperature and concentration 88 Nomograph for estimating MEA circulation rate... ss... eee 88 Graphical determination of amine circulation rate... 111. 10.~« (90 Amine contactor size required... .... Bie reise eee eine 3 Amine plant cost (1977 dollars). at 93 4 Equilibrium data for HyS and 15.3% MEA at 40°C with COp present...) 94 4 Equilibrium data for H2S and 15.3% NEA at 60°C with COp present |... 95 4 Equilibrium data for HS and 15.3% NEA at 100°C with COp present... 95 4 Equilibrium data for H2S and 15.3% NEA at 120°C with CO2 present... 96 4,10 Equilibrium data for C02 and 15.3% MEA at 40°C with H2S present... 96 1 Equilibrium data for COz and 15.3% MEA at 60°C with H2S present... . 97 4.12 Equilibrium data for C02 and 15.3% MEA at 100°C with HS present... 97 4.13 Equilibrium data for COz and 15.3% MEA at 120°C with H2S present 11. © 98 4.14 MEA treating limitations ....... 0... Piscreeratieree etic ia areas HH#EHOO: 4.15 Trends showing expected CO2 retention in MEA solutions... 11... 99 4.16 Typical MEA contactor temperature profile. .............. (100 4.17 Nutual solubility data for HS and C02 in 2.0N DEA at 50°C... + 106 4218 Mutual solubility data for HoS and COp in 2.0N DEA at 60°C 11... 106 4:19. Mutual solubility data for HoS and COp in 2-0N DEA at 100% ||) 1). toy 4:20 Mutual solubility data for Hs and COp in 2.0N DEA at 100°C |)... 107 4:21 Mutual solubility of Hos and/COp in 3-5N DEA at 0°... 2.1... 108 4.22 Mutual solubility of HgS and COz in 3.5N DEA at 50°C 2.222122. 108 4:23 Mutual solubility of HS and COp in 3.5N DEA at 100°C |). 109 4:24 Mutual solubility of HoS and COp in 3.5N DEA at 100°C |. 22). og 4.25 Effect of acid gas composition and temperature on the viscosity of MEA sOlutiona/jectaisesaeecienierer iets ae ee eererieteceeess ea 4.26 Effect of acid gas composition and’ temerature on the specific gravity Of MEA solutions... eee eee eee ee ee ieee eet pmereeeec) 4.27. Effect of acid gas composition and temperature on the pH of MEA SOUubsons ites terse tater eee ora ciiietecsitesidees eats tuttet yy 4.28 Kent and Eisenberg comparison of calculated and experimental data for HS - MEA equilibrium... . Pepe eorciernicne + 123 4.29 Kent and Eisenberg comparison of calculated and experimental data for Op = DEA equilibrium se ee ee ee ee ee 123 4.30 Predicted and experimental values of partial pressure of (0p... . 124 4.31 Predicted and experimental values of partial pressure of mall te 125 5.1 Equilibrium pressures of H0 vapor over KzC03 solution... ... .. I 5.2 Schematic flow diagram for a conventional hot carbonate process... . 133 5.3 Schematic flow diagram for a split-stream hot carbonate process... 134 5.4 Schematic flow diagram for two-stage hot carbonate process ...... 134 5.5 Reducing steam consumption with ejectors ..........- 2+ 1S 5.6 Equilibrium partial pressures over solutions of potassium carbonate |. 136 5.7 Nomograph for calculation of COz absorption... 2... eee ee es 139 5.8 Nomograph for calculation of COz absorption. +... 2.222... 10 5:9 Nonograph for calculation of (Op absorption |... 1... ..5552~ Ul 5:10 Nonograph for calculation of COs absorption |. 1... ssl .se 1. (ae 5.11 Selection quide to obtain specified gas purity’. . sere eens 5.12 Effect of contamination on activity of carbonate solutions | 11... 145 5.13 Activity of catalyzed carbonate solutions... +++ +e eee ee ee 145 5.14 Flow sheet for DEA-Hot carbonate process»... 1... te 2. M5 5.15 Comparison of COz absorption in G-V and hot carbonate solutions... . 148 ivFigure Page 5.16 G-V process with steam regeneration... 2. ee ee ee eee ee 150 5.17 G-V process with air regeneration... . . anaes ate 5.18 Basic flow sheet for the Seaboard process». +... ee eee ee 183 5.19 Basic flow sheet for vacuum carbonate process... 2... ++ + 185 5.20 Basic flow sheet for tripotassium phosphate process... . 2... ++ 186 5.21 Equilibrium vapor pressure of HS over 50% and 20% tripotassium phosphate solutions»... ee ee ee tt ee ee . 187 5.22 Schematic flow diagram for Alkazid process... 1... D158 5.23 Equilibrium partial pressures of HS and CO over Alkazid solutions . . 160 5.24 Rate of absorption of HS and CD2 in Alkazid solutions ........ 162 5.25 Partial pressure of CO over 20% equiv. Kp003.---- +--+ e+e e+ 163 5.26 Partial pressure of C02 over 30% equiv. K3C0}.-- +. 1 ee 163 5.27 Partial pressure of C02 over 40% equiv. K2003..... 6. sae 164 3128 Partial pressure of COs over 30% equiv. KsC03 with total loading | |) lea 5.29. Partial pressure of COS over 40x equiv. KoC03 with total loading... 165 5130 Partial pressure of Hos over 30% equiv. KSCO3 with total loading |. . 165 5.31 Partial pressure of HpS over 40% equiv. KsCOy with totai loading | | 166 6:1 Typical water wash flow sheet. ss swe S peepee) 6.2 Calculated performance of water wash unit)... eee eee ee ee 122 6.3 Typical fluor solvent process for synthesis gas treating ....... 174 6.4 Equilibrium constants for carbon dioxide in propylene carbonate... . 179 6.5 Equilibrium constants for hydrogen sulfide in propylene carbonate... 179 8.6 Solubility of C0 in DMTES containing 1.5, 2.5 and S% Hgd . . .. . «180 6.7 Solubility of CO2 in DMPEG containing 1.5%H20 .........+.. 180 8-8 Pikes Peak carbon dioxide renoval plant... Peeeone eee alist 6.9 Selexol process for a gas with a high C02 to HzS ratio. pene 183 6-10 Selexol process for a gas with a low COz to H25 ratio . tee 188 GO greene eae teeeosa tic teesteitatie east EC aa eee eS e 6.12 Gas solubiiities in Seiexol 1112 T IIIT I! see 6.13 Flow sheet for Selexol plant treating Sour gas with high HS... . 186 6.14 Mitchell plant selexol unit... see ee ee eee eee ee ee 188 6.15 Equilibrium constants for carbon dioxide in selexol ss... ee ee 188 6.16 Equilibrium constants for hydrogen sulfide in selexol......... 189 6.17 Basic flow sheet of Purisol process... see eee eee 190 6.18 Vapor-liquid equilibrium diagram for M-Pyrol-water system at atmospheric Presa iseeeditisescre cress gereeise uate ateeeatictCateeteatteietatsareadasteesteeO? 6.19 Solubility of sulfur compounds and COg in M-Pyrol ss... eee ee 193 6.20 Solubility of paraffin hydrocarbons in M-Pyrol «es. ee ee ee. 193 6.21 Vapor-liquid equilibrium values for sulfur compounds and COg in M-Pyrol at atmospheric pressure . 2... . See ceite lier eee eeeeie 19m 6.22 Vapor-Viquid equitibrium values for paraffin hydrocarbons in M-Pyrol at atmospheric pressure... 2... : secaie tiered 6.23 Correlation of equilibrium values for ethane and heavier hydrocarbons in M-Pyrol at atmospheric pressure... eee eee eee ee 195 6.24 Relationship between "Kk" values and solute vapor pressures in M-Pyrol . 195 6.25 Flow sheet of Rectisol process for HgS and C07... .- +. +++ ++ 196 6.26 Equilibrium curves for COz in various washing agents ......... 197 6127 Basic Estasolvan process for renoving H2S . sss «ss cvs ts 199 6.28 Estasolvan process for simultaneous recovery of hydrocarbons and sweetening. 2... eee ee see ieee saree dates ele etiart Pedieren 99Figure Page 6.29 Equilibrium chart for natural gas components in MCA... . 2... ++ 202 6.30 Equilibrium chart for natural gas components in MCA... 2.12] pgp 6.31 Equilibrium chart for natural gas components inMCA... 2... 1.) (202 6.32 Equilibrium chart for natural gas components in MCA . Bee eee eee e202 6.33 Equilibrium chart for natural gas components in MCA... 1.1. (303 6.34 Equilibrium chart for natural gas components in MCA... +.) 203 6.35 Equilibrium chart for natural gas components in MCA... 2... 2. (203 6.36 Equilibrium chart for natural gas components in MCA... 2+... (203 6.37 Equilibrium chart for natural gas components in MCA. >>)... ) (203 6.38 Equilibrium chart for natural gas components in MCA... 2) 11) 2 (204 6.39 Equilibrium chart for natural gas components in MCA... 11... (208 6.40 Equilibrium chart for natural gas components in MCA... 1... 204 6.41 Equilibrium chart for natural gas components in MCA... 11). 204 6.42 Graphical solution of absorption factor equation... +... 2... .~« 206 6.43 Solubility characteristics of physical and chemical solvents |...) 207 6.44 Solubilities of several gases in SepasolvMPE...........~, + 208 6.45 Vapor pressure for several physical solvents»... 211. 208 6.46 Equilibrium water partial pressure over Sepasolv MPE.. 7...) 1) 209 6.47 Sepasolv MPE schematic flow sheet... 1... ee el) 209 6.48 Characteristic curves for constant inlet and outlet pressures of turbine SING BMD recedgece cette die eee et ates eee eee te eee eae 6.49 Application ranges for power recovery turbines 12.1... 10«: 212 6.50 Comparative performance of power recovery turbine and pump... ... 212 7.1 Typical iron oxide process flow sheet... 2... eee sll 217 1.2 Progress of stain front through iron oxide bed»... 2... 2.2. (219 2.3 Oxygen content of regenerating streams ......- Geeetaetee eee 220) 7.4 Temperature profile for iron oxide tower»... sl 220 7.5 Adsorptive capacity of molecular sieves at 25°... 2.211111) be 7.6 Adsorptive capacity of molecular sieves at 25°C... 11 1 lll) 222 7.7 Effect of temperature on adsorptive capacity of molecular sieves at atmospheric pressure... eee ee ee ee ee ee ee eBeD: 7.8 Effect of contact time on adsorptive capacity of molecular sieves a atmospheric pressure... ee ee ee 222 7.9 Effect of pressure on adsorptive capacity of molecular sieves at 25°C . 223 7.10 Schematic representation of the adsorption process .......... 224 7.11 Schematic flow diagram for a simple molecular sieve adsorptive process. 224 7.12 Adsorption of sulfur compounds on 13X molecular sieve during First Vee sie rerrecteiiiare ee tea teectisieereteliie nits Retains tutea Ee itig2g: 7.13 Desorption characteristics of sulfur from 5A molecular sieves .. 1. 2 226 7.14 One system used for removal of mercaptans....... 4. +00... 227 7.15 EFCO closed cycle sweetening process... 2. 1 et 288 7.16 Equilibrium adsorption loading of COz on 4A pellets. 2... 12. 229 7.17 Equilibrium adsorption loading of COz on 5A pellets... .......~ 230 7.18 Equilibrium adsorption loading of COp on 13X pellets... 111. 230 7.19 Equilibrium adsorption loading of Ho$ on 4A pellets... 2.2... .~ 231 7.20 Equilibrium adsorption loading of HpS on 5A pellets... ... 2... 231 7.21 Equilibrium adsorption loading of H2S on 13X pellets. 1.) 1111. 232 7.22 Adsorption equilibrium loading of Hp0 on 4A pellets... 1 1s 232 7.23 Equilibrium adsorption loading of H20 on SA... . 22.7 238 7.24 Equilibrium adsorption loading of H90 on 13X21. 1) lll 233 7.25 Comparison of experimental and calculated breakthrough times for HS on SA molecular sieve... ee ee ee ee ee 284 viFigure Page 7.26 Effluent gas concentration of COS and HyS during a cycle... . .. + 236 7.27 Perfornance of one molecular sieve product in the laboratory at 90°F amie BO seat ee reeeeeaaeeaetn eee tet eeeree eae eee ieee Hea eg) 7.28 Trisiv adsorbent... 2. ee ee ee eae ce eects 238) 7:29 Chemsweet contactor. 2). ee ee 288 7.30 Flow diagram for Slurrisweet unit 2. ee ee ee ee ee 240 &!1 Flow sheet for COS renoval with KOH and methanol | 2.2. +2. 1.2 282 8.2 Flow sheet for Malaprop process... ee ee sees eee eee es 28 813 Tarnish of copper strip as function of HS concentration |. +... + + 245 8.4 Schematic flow sheet for amine treating process ...-.-- ere seer 2NG, 8.5 Schematic flow sheet for molecular sieve process». +--+ +++ +++ 247 8.6 Flow sheet for removal of HyS with dilute caustic»... se... + + 208 8.7 Schematic flow sheet for caustic wash process... 2... + eee 251 8.8 Schematic flow sheet for regenerated caustic wash system... . REST eo2 8.9 Schematic flow sheet for a solutizer process. e+ + ee eee eee + 255 8.10 Schematic flow sheet for methanol (Unisol) process... ++... ++ + 256 B.ll Schematic flow diagram for Mercapsol process. +--+ +++ +--+ + 257 8.12 Schematic flow diagram for Dualayer process... ++ se ee ees + 259 8.13 Stripping coefficient in normal mercaptan ... +--+ +++ ++ +++ 260 8.14 Extraction coefficient in normal mercaptan. ...----- +++ +++ 260 8:15 Schenatic flow sheet for copper sweetening process «=. 6 + ss +: 262 8.16 Solid copper treating process... eee ee ce te ee ees 263 8.17 Schematic flow sheet for Merox process»... +e e+ eee see ees 264 8.18 Schematic flow sheet for Bender process... es ee eee ee ees 265 9:1 Equilibrium composition of various sulfur species in sulfur vapor at different temperatures... ee ee ee ce 268 9.2a Density of liquid sulfur in region of transition temperature... +. + 272 9.25 Density of liguid sulfur... - . - See ene anette 9.3 Viscosity of liguid sulfur... . - pees ete eee et a7e 9.4 Flow sheet for Ferrox process. ee ee eee ee ee ee es 2 9.5 Flow sheet of Manchester process 2. 0+ et ee eee eet es 276 916 Flow sheet for Peabody-Holmes-Stretford process... +++ +++ tes 280 9:7 Flow sheet of Thylox process. ee tee eee ee ee fs 283 9:8 Giammarco-Vetrocoke flow sheet for HzS removal from gases also containing C07 .. +... ++ pee eee pe ceiee eae cee ne 9.9 Giammarco-Vetrocoke flow sheet for HS removal when no C02 is available 285 9:10 Claus “once through" process flow scheme... ++ seer vee eee = 290 9.11 Typical reaction furnace and waste heat boiler... +... +++ +++ 290 9:12 Claus "split stream" process flow scheme... -- +--+ eee ees 291 9:13 Claus flow sheet using in-line burner ss ee ee ee ee ee es 292 9.14 Claus flow sheet using gas-to-gas exchanger se cea eee de ene oie 202 9.15 Claus flow sheet using two steam heaters... 2 12 ee ee + 293 9116 Claus flow sheet using gas-fired indirect heater»... ++ e+ 22s 293 9:17 Sulfur burning Claus flow sheet... . - ee see Aue Eeee Een a2) 9:18 Variation of sulfur vapor composition with temperature at one atmosphere pressure... 1. poate aerees aise s eeeeeetiate eters etetieteo 1209) 9.19 Equilibrium constant for reaction 21 ee ee ee 300 9.20 Vapor pressure of Sp(g) in equilibrium with St) aerate + 300 9:21 Vapor pressure of Sg(g) in equilibrium with S(1) 2... + +++ ++ + 301 9.22 Vapor pressure of Sg(g) in equilibrium with S(1) 6... ee eee eee 301 viiVapor pressure of sulfur in equilibrium with liquid sulfur . Equilibrium between Sp, Sg, Sg vapor... ee ee ee Equilibrium between Sy, Sg, Sg. e+ eee eee Effect of acid gas composition on sulfur yield. 2. 2 12! Selectox process in Claus tail gas cleanup application Water content of typical Claus unit tail gas... . . Flow diagram for once-through Selectox process . . . . Flow diagram for recycle Selectox process .-.-. - Calculated equilibrium conversion efficiencies from Slay 3 uni Calculated tail gas sulfur vapor losses»... - 2-2-2 ee Experimental oxygen consumption in RSRP reactor .... . meee Conceptual flow sheet for RSRP sulfur recovery process . . . Sulfur oxide emissions from typical Claus plants... 2. . Capital cost comparison for various configurations of Claus unit tail gas clean-up processes... ee ee ee ee ee Expected profitability of sulfur produced by different processes . . . . Schematic of ASR Sulphoxide process .. ++ +--+ +e eee eres Generalized citrate process flow sheet... +. +e see The relationship of the Westvaco recovery process to a Claus Flow sheet for Sulfreen process .. 5s ++ er ees Flow sheet for IFP process... ee ee ee eee Aquaclaus process flow diagram. 2... 1. ee eee Flow diagram for Wellman-Lord SO» recovery process . . Flow sheet for Lucas process... ee ee ee ee Flow sheet for cold bed adsorption process... . . eee Process flow for Allied SO2 removal process»... +--+ --+---- General flow sheet for Pritchard cleanair process .... . Beavon sulfur removal process schematic flow diagram... . Basic process flow diagram for SCOT process -.-.--. + vitiTable LIST OF TABLES Page Ld Sulfur Supply Sources Worldwide... - ee ee ee ee ee 4 1.2 Toxicity of Hydrogen Sulfide To Man... eee ee © wee eee 7 2:1 Summary of Common Systems of Units. sss c esc e cee eres 2:2 Derived Units for the Gas Industry 121 tte. 18 2:3 Conversion Factors tor Volume Related Units (at equal conditions of measurement)... -. 2 +s HERES Perera ato eee 2.4 Conversion Factors for Gas Flow Rates (under different standard Conditions of measurenent. in conventional U.S. and wetric systens) . 17 2.5 Constants for Two-k Method... ge eee sae ag0. 2.6 Overall Design Coefficients for Shell-and-Tube Heat Exchangers pre aa 2.7 Typical Overall Coefficients for Aerial Coolers... +++ esses + 37 2.8 Fouling Factors... - BEEP RE Earl tai or rare eee Pe Peete 2.9 Physical Constants of Hydrocarbons Bebe Hatter to RSet etal ar ap eat 3.1 Properties of Absorbents . 2. +. se ee Pee ee wee See 3.2 Some Characteristics of Ethanolamines . . . ee ee ee ee ee ee eure! srg (aranco DeNTPlantS cee Gee aes ee eects 68 @.] Heat Exchange Requirenents for Amine Plants |) 1122222 l i... 89 4.2 Pump Horsepower Requirements for Amine Plants... . - . ae oa 4.3 Regeneration Vessel Sizes. - - - eee ee ee ee 91 414 Amine Unit Operating Expenses |. . + Se ee ea) 4.5 Effect of Hydrogen’ Sulfide and Carbon Dioxide Loading on Equilibrium Partial Pressures Over A 15% Solution of MEA... see eee eee 93 4.6 Enthalpy of Solution of COp in DEA Solutions ...-- +++ seer ee 110 4.7 Enthalpy of Solution of H2S in DEA Solutions ... 6... ee eee pee 4.8 Heats of Reaction of C02 with Diethanolamine .....+- pee ee 4.9 Calculated Heats of Reaetion for HS and COz with Diethanolamine eee 127 $.1 Performance of Benfield Carbonate Golution .- . 1. ee eee ves 182 312 Performance of Catalyst in Commercial Plant A. . sera AG 5.3 Comparison of COg Removal Processes on Typical ‘Ammonia “synthesis Gas and Natural Gas‘Streams . 2... uate taaee eee aE 5.4 Summary of Errors in Predicting Pod, for the CO2/K2C03 System... +. . 167 5.5 Summary of Errors in Predicting Poo and Pigg for the HzS/COp/KyC03 System. ss Bebe tare eee wee Beret tae ae eer ete 6.1 Henry's Law Constants for Gases and Physical solvents Bee eee eee ae 6.2 Experimental Data for the Solubility of C02 in Propylene Carbonate Solvent vv. soo eee TdT 6.3 Experimental Data for the Solubility “oF HeS in Propylene Carbonate Solvent . 2... BER eoE Hae ia see ~- 178 6.4 Henry's Law Constants for HyS and CO, in Propylene Carbonate Solvent + + 178 6.5 Material Balance for Pikes Beak Plant... 0... + seen 182 6.6 Operating Data for Pikes Peak Plant...) 1... cect ee tee 182 €.7 Relative Solubility of Various Gases in Selexol .. 1.2 ssc ess + 186 6.8 Feed Gas for Selexol Unit... -.-. +... asec Igy 6.9 Material Balance for Streams Leaving the Unit . . ee a7 6.10 Comparison of Fuel Gas Requirements and Sales Gas Flows for Three Processes at the Mitchell Plant se eoeeeeane ge nice ay 6.11 Physical and Chemical Properties of N-! Pyrot erat tee eet «191 8.12 Vapor-Liquid Equilibrium Data for M-Pyrol-Water Systens . 52 6. 2... 191 6.13 Loading Capacities of Some Typical Solvents. ....- fer eet Heat ixPhysical Properties of Tri-n-butylphosphate (TBP) . . Estimated Cost of Treating High COS Content Propane . . Estimated Cost of Treating Low COS Content Propane... Estasolvan Plant Operating Data... ++. eee e et lle Properties of Various Solvents .. 2... 00.0000. Solubility of Gases in Selective Solvents... 1122.21! eee Physical Properties of Sepasolv MPE... . . ete a eeeesetetee Plant Material Balance... 2... 1. else ee ueeac etait Equilibrium Constant for Reaction... 1. e+e. ee eee eee COS Formation Data... +... ee ee 7 tects 7 Vapor Stream from Fluid Catalytic Cracking Unit . . . Sulfur Values in Vapor Stream Leaving DGA Unit . . . : Maximum Mercaptan Sulfur for Negative Doctor Test . . : Effect of Disulfide Sulfur on Research Octane Number... | Constants for Use in Equation 8.17 ....... 2.05 Substances That Promote Solubility of Mercaptans in Caustic 7 Removal Efficiency of Mercapex Process .....-... 0.0000 ee Heat Capacity Equation Coefficients for Sulfur in the Condensed Phases Ideal Gas Thermodynamic Properties for Monatomic Sulfur... 2... + Ideal Gas Thermodynamic Properties for Diatomic Sulfur... 1... Ideal Gas Thermodynamic Properties for Cyclohexasulfur + - Ideal Gas Thermodynamic Properties for Cycloheptasulfur . . Ideal Gas Thermodynamic Properties for Cyclooctasulfur Equation Coefficients for Vapor Pressure of Liquid Sulfur : Stretford Plant Operating Conditions ......... 5 : Cost Sunmary for Sweetening with Stretford-Holmes Process . - 7 Comparison of HpS Removal Processes... .... a erettaite ter) : Sunmary of Cost Used in Economic Analysis (Based on’ 1959 Prices) i Comparison of Economics Using Costs from Table 9.11. ...... : Procedure for Heat and Material Balance in Reactor Furnace . . . ‘| 01 Procedure for Equilibrium and Heat Balance Calculation for Claus Reactor Sid Richardson 20 LTPO Recycle Selectox SRU Design Basis... .... Sulfur Recovery Potentials of Various Combinations of Claus Unit and TaitzGas!Cleanlipl Processes tintcitantiscreioaert tte aitsHteeaieeY S02 Removal Processes for Furnace and Utility Stack Gases... 1.1! Cost of Removing S02 by Various Processing Schemes... 2... a Sulfreen Plants Built as of 1977 2... ... ee eee iTnives tent tror-sulfreenPlant eexciteasantetoore tiene atte ag gece oa Three Applications of the IFP Process for Treating Claus Unit Tail Gas Fixed Investment 400 ITPD Recovery Plant... esse ee ee ee Operating Requirements for Recovery Plant... 2... 2. le Utility and Chemical Requirements for Wellman-Lord Process»... 2. Investment and Operating Costs for Wellman-Lord Process on 200 Ton per Daye ClauiedUn tt eereterceiceeet race cettepeesasrrst tracer steeecsrreeteecte Gas Compositions at Lingen Refinery. 2 2 lt Utilities Required per 1000 kg Sulfur Produced 222). D lI! age 200 200 201 201 207 210 235 237 244 204 244 245 250 250 253 254 261 268 269 269 270 270 ani 2m 279 279 286 287 288 304 305 307 314 315 319 324 324 326 328 328 330 331 333 333ACKNOWLEDGMENT This book is possible because numerous persons have shared their knowledge and experience in talks and papers for trade and professional groups. I thank them not only for their contribution to this book but. to knowledge in general. 1 also wish to thank many past university and adult students for their questions, comments and other input which have contributed to my attitudes about this and other subjects. Special thanks are due Fran Porter and Joy Garrison for their edi- torial help and Danny Stowe for drafting services. dune, 1982 Dr. RN. Maddox1 INTRODUCTION This, the fourth volume in the series “Gas Conditioning and Processing,” is devoted to a discussion of the removal of sulfur compounds from gas and liquid streams. The emphasis is on those processes used to condition gas and liquid to meet specifications for commercial sale or to permit safe utilization. Some attention is given to tail gas cleaning since this is an auxiliary function of a conditioning plant to meet environmental regulations. However, this book is not intended as a definitive work In gas cleaning which is a subject in itself. Likewise, sulfur recovery is a natural part of the sweetening process in an integrated plant, particularly when a large amount of fluid is being sweetened. Like tail gas cleaning, this is a subject whose scope could |ustify a separate book. Thus, the chapter on sulfur recovery may be considered a detailed overview of the subject. This book is self-standing in that it provides all the data needed for the calculations discussed. It also contains many correlations of physical and chemical data that hopefully will be of value when used as a desk reference. ‘A sweetening installation contains many mechanical components discussed in Volumes 1 and 2. In preparing this book it has been assumed that the reader is familiar with the characteristics and performance criteria for such components. The emphasis herein is on that part of the installation which removes the contaminants from the gas and the stripping of these contaminants from the liquid or solid employed. Familiarity with Volumes 1 and 2, through training andlor experience, is valuable if one is to use this book most effectively. Natural gas, whether produced from a condensate field or as associated gas from an oil reservoir, usually contains water vapor and frequently contains H.S and/or CO,. Because of its large volume, gas traditionally has been transported overland in high pressure pipelines. These were made possible by the development of high pressure seamless tubing (pipe) and field welding techniques to replace screwed fittings. Overland transportation was enhanced by the development of techniques for removing water to prevent cold weather line “freeze ups” caused by water condensation and subsequent formation of hydrates. Hydrogen sulfide removal is desirable to reduce corrosion. In many cases it is necessary from a health and safety standpoint. The first processes used for removing H,S from natural gas (“sweetening the gas) were those that had been developed for treating coke oven and other manufactured gases, While suitable in some applications, the large volumes, high pressure and high H,S concentrations common in natural gas called for the development of new sweetening processes.‘The amine processes (7.2, 1.3) were developed to meet this need for a high pressure, high volume H,S removal process. For many years the amine processes were just about the only choice available to the gas engineer for treatment of natural gas streams. More processes have been developed, however. Today there are large numbers from which the engineer may choose. There are many factors that need to be considered when selecting a process for a given sweetening application. These include: 1, The types of impurities to be removed from the gas stream. 2, The concentration level of these impurities and the degree of removal desired. 3. The acid gas selectivity required, if any. 4, The volume of gas to be processed and the temperature-pressure conditions at which the gas is available. 5, The feasibility and desirability of sulfur recovery. 6. Relative economics of the suitable processes. Acid gas constituents present in most natural gas streams are hydrogen sulfide and carbon dioxide, Many gas streams, however, particularly those in a refinery or for manufactured gases, may contain mercaptans, carbon disulfide and/or carbonyl sultide. it any of these constituents are present in the gas stream, they provide a boundary condition that may eliminate some sweetening processes. Irreversible reactions degradation of sweetening solution or non-removal of the acid gas constituents may cause many processes to be ineffective or economically unattractive. The level of acid gas concentration in the sour gas is an important consideration for selecting the proper sweetening process. Some processes are applicable for removal of large quantities of acid gas. In many cases, however, processes will not sweeten to pipeline specifications. Other processes have the capacity for removing acid gas constituents to the parts per million range. Some of these processes, however, are applicable only to low concentrations of acid gas constituents in the sour gas to be treated. ‘The selectivity of a sweetening agent is an indication of the degree of removal that can be obtained for one acid gas constituent as opposed to another. There are sweetening processes which display rather marked selectivity for one acid gas constituent. In other cases there is no selectivity demonstrated and all acid gas constituents will be removed. There are processes for which operating conditions can have a marked effect on the selectivity exhibited. Some sweetening agents absorb relatively large amounts of hydrocarbons while others are much more selective for the acid gas constituents, Only rarely will natural gas streams be sweetened at low pressures. However, there are processes which are unsuitable for removing acid gases under low pressure conditions. Other processes are adversely affected by temperatures much above ambient. Some processes lose their economic advantage when large volumes of gas are to be treated, The feasibility and desirability of sulfur recovery can place considerable limitation on the sweetening process selected. in general, most sulfur plants based on feed stock derived from natural gas,streams use the Claus process. The quality of sulfur produced from a Claus unit is quite sensitive to the presence of hydrocarbons, particularly the heavier hydrocarbons, in the acid gas feed. Those processes in which the solvent absorbs substantial quantities of hydrocarbons may require additional processing on the acid gas leaving the unit in order to make it suitable for feed to a Claus unit. Until about 1970 the decision as to whether or not sulfur should be produced in conjunction with the natural gas sweetening operation was a matter of economics. If there was a demand for the sulfur that was produced and if the price of the sulfur was sufficient to make the project economically feasible, then ‘a sulfur unit was installed. If the economics were not attractive, the acid gases were either vented to the atmosphere or, more commonly, incinerated through a flare. With the advent of greater concern for the environment and sulfur emission, the picture on sulfur production has changed. In most of the industrialized countries, conversion to sulfur of the hydrogen sulfide removed from the natural gas stream is a matter of necessity rather than an economic consideration. Concer for safety and health, given impetus by sulfur emission standards, has led to a rapid development of processes for conversion of gaseous sulfur containing compounds to sulfur. The mechanics and chemical reactions vary but most are aimed at producing elemental sulfur (though other chemical compounds sometimes result). These forces have been acting not only on the natural gas industry but also on smelting, power piants and many other industries in which gases vented to the atmosphere contain sulfur compounds. ‘The second edition of this book was written in the early 1970's before the rapid escalation in cost of world energy resources. At that time, because of environmental forces, there were many who forecast a world sulfur oversupply. We stated: “The large production of sulfur from exhaust gases led to a supply-demand problem for sulfur, This is best exemplified by a study (1.4) which illustrates the relationship between consumption (demand), recovery (supply) and sulfur discharge to the atmosphere. In 1970 production and consumption were pretty well in balance. In 1980, based on the assumption that 50% of the gases discharged will be processed for the recovery of sulfur or sulfuric acid, a surplus production of some 20-25 million tons of sulfur per year is predicted, This excess sulfur production has led to the shut-down of a number of sulfur mines (1.8) because they cannot be competitive in a surplus supply situation.” ‘The point that was forgotten, of course, is that most methods of producing sulfur, notably the Frasch process, are energy intensive. As energy prices rose, the price of sulfur increased almost in parallel, The fesult has been a much better balance between overall sulfur supply and demand than was forecast ‘There have been geographically localized sulfur surpluses and shortages, but for the most part world supply and demand have maintained a satisfactory balance. In 1972 Vroom (1.6) published a breakdown of sulfur uses as of that time. His breakdown is shown at the top of the next page. Newton (1.7) showed a similar breakdown for 1978. The categories in his breakdown are slightly different than those of Vroom. However, as Figure 1.1 shows, fertilizer still is the domlnant user of sulfur World wide. Newton also showed the production of sulfur from different sources, and commented that Frasch sulfur is maintaining its historical 25% of production. Its share of the total sulfur supply had increased from 18% in 1965 to 29% at the time of Newton's article.VROOM'S SULFUR USE BREAKDOWN, _Sulfur use Percentage of Total Chemicals 19 Fertilizers 49 Insecticides 2 Carbon Disulfide 3 Pulp and Paper 4 Petroleum 2 Iron and Steel 2 Rayon and Film 3 Pigments 5 Other mise. Uses al TABLE 1.1 Sulfur Supply Sources Worldwide % = z ie g re 3 Tet pins, [=e " 2 Ba: 4 8 Tet : @ cS Phosphate ferizer 82 Figure 1.1 Western world sulfur consumption by end useIn recent years, thirty year trends in sulfur production (1950-1980) were presented by Rangnow and Fasullo (1.8). They show a steady increase of recovered sulfur with some small decrease in Frasch production. Their projections to the year 2000 show a dramatic increase in recovered sulfur production as shown in Figures 1.3 and 1.4. ion matte one 6 oso 1955 1060 1985 197 1875 1980 Yeor Figure 1.2 U.S. sulfur production trends § 2 E § Year Figure 1.3 U.S. sulfur supply/demandos. os oa os. 02 Suir recovery, Iisct os 00. 1a70 1975 1980 1085 79002000 Your Figure 1.4 Sulfur recovery from natural gas TOXICITY OF HYDROGEN SULFIDE Hydrogen sulfide is a coloriess flammable gas which has an offensive odor similar to rotten eggs. It is extremely poisonous ~ essentially as deadly as hydrogen cyanide, Unfortunately the human system adapts readily to H,S odor so that increasing concentrations may not be noticed. High concentrations of H,S result in rapid poisoning and death. Archibald (1.9) has presented some very interesting information concerning steps to be taken in processing sour gas safely. He gives the allowable concentrations for periods of exposure as follows: Parts of H.S per million parts of air Period of exposu 10 Maximum. allowable ‘concentration for pro- longed exposure 70-180 Slight symptons after exposure of several hours 170-300 Maximum concentration that can be inhaled for one hour without serious consequences 400-500 Dangerous after expo- sure of 30 minutes to ‘one hour ‘800-800 Fatal in exposure of 30 minutes or lessThe physiological reaction of the human body to H,S is shown in Table 1.2 (1.9). TABLE 1.2 Toxicity of Hydrogen Sulfide To Man aan ama | sitmanes | sommeottin |__ sean | satan erect Sees eae me cor selbst Ros saa a Te | Fria [a Bag a ca ; Seokingta aa Tramway S| ad 288 ear ak ‘o0g is ‘Diearepiwtfoari’| Toereened ikation of | Disiogm; wenioew: | Dex am Bienes | Srerestan| aamaaires | -€.080 (00) ‘Cousin, calnawe ood | Respir disturbnon: | Serious ere rita ‘wd | - qi et sere acre uo: dest | Sass: death pa eect toe seinen WH dog wl Dae asmerlay ie owen. PUL gee ‘As can be seen, relatively low concentrations of H,S are deadly so extreme care must be taken in ‘operating to avoid leaks, spills and vents. GAS SPECIFICATIONS The general specification in the United States for “sweet” gas is that it may contain no more than ‘one quarter grain of HS per 100 standard cubic feet of gas. The “grain” is a unit of mass more common in pharmaceutical or medicinal use than in engineering use. There are 7,000 grains in one pound mass. So-called “quarter grain gas" contains about 4x10~* mole fraction of H2S. This is equivalent to four parts per million on a gas volume basis. ‘The concentration on a weight basis depends on gas gravity. For 0.65 specific gravity gas, this would be approximately seven parts per million by weight. in the metric system quarter grain gas contains approximately 6 mg of HgS per normal cubic meter of gas. At 1,000 psia the H2S partial pressure in quarter grain gas is 0.2 mm of mercury. What all of this says is that removal of H9S to produce contractually sweetened gas is a difficult thing to achieve. Very slight upsets can throw ‘the gas off specification and result in an unsatisfactory product. Operation of a sweetening unit treating to these levels requires a delicate balance and Continuous attention to what might be considered minor details in other circumstances. Selecting the best sweetening process requires extensive investigation of many processalternatives. The number of variables involved and the large number of possible processes available make an absolute determination difficult. However, some guideiines can be developed that will serve to consolidate some of the variables and, perhaps, make the determination somewhat easier. Tennyson and ‘Schaaf (1.10) presented a series of graphs that are quite helpful in this regard. These are shown in Figures 151.8, Figure 1.5 illustrates the case for removal of CO only from a gas stream. Currently, often will be encountered in synthesis gas purification for ammonia manufacture. However, the increasing use of COz injection for enhanced oil recovery indicates that this case will be of more interest to the gas industry in the future. The processes to be considered are dependent on the concentration of C02 in the inlet gas and on the degree of removal of CO2 required. For low inlet concentrations of GO. (in the range of 10-15 psi), amine solutions are preferred as they can reduce the outlet concentration of 60> to the 0.005 psi partial pressure range. From about 15 psi to 100 psi partial pressure of CO in the inlet the hot carbonate solutions (which can produce outlet partial pressures as low as 0.5-1.0 pel) should be Considered. For inlet CO2 partial pressures greater than 100 psi, physical solvents should be considered, though they generally are limited to outiet partial pressures of 1.0 to 3.0 psi. this case most (0a Ho ter inputs preset) +00) ‘wo 00 0 0 110 ry a el pes of Cyn ed, ps orf oa ear 246 8105 2040 B0 B9100 Pail preset CO in prods pi Figure 1.5 Removal of Carbon Dioxide onlyIn the case of removal of H2S only, represented by Figure 1.6, process selection must take into account the final disposition of the H2S. In most cases today, this will be to produce sulfur, At low inlet HS partial pressures, the Stretford and Vetrocoke processes are suitable. One process not shown by Tennyson and Schaaf, molecular sieve sweetening, must also be considered, At about 3 ps! inlet partial pressures, amine processes enter the HS picture. Physical solvents must be considered about HS Inlet Partial pressures of about 60 psi, though these processes generally encounter difficulty in treating to sweet gas specification. Simultaneous removal of H2S and COz, represented by Figure 1.7 is frequently encountered in the natural gas industry. In this case, process selection is severely dependent on the COp to Hy ratio. Amine, hot carbonate and physical solvents can be used. In many cases, the exact amine process will depend on the C2 to HS ratio. Hot carbonate is suitable for high CO to HS ratios, but not for high H9S to CO2 ratios. Selective removal of H2S in the presence of CO2 is of great interest. The Stretford and Takahax processes are selective and remove only H2S from the gas stream. ADIP offers some selectivity, as do Rectisol and Selexol. The new MDEA (methyldiethanolamine) process also shows selectivity for HS in the presence of COz. Figures 1.6-1.8 are useful as a general guide but should not be used to select or reject a given process. Economics and local circumstances may dictate choice of a process not conforming to the conditions used to generate these general figures. 1.0 ‘tt a] re) 2a0] Pd bee a rth montamiepbeestt 1] te, el pea of fd, pl 7 ‘ i mm ‘ ; . rat Varo Prt hee ee Figure 1.6 Removal of Hydrogen Sulfide onlyth no ther impurities pro va 2 =e “ = : “| ieee, rT 200} on t a i a Patil pres al OF ager to 146 8100 28 40 so aNTO0 Perl pase xd gsi prod, Figure 1.7 Simultaneous removal of HS and CO, 1.400; c a) 400 Acta Sz eal pera fH no, a a OC) ‘Seva in pot, pm Figure 1.8 Selective removal of Hp: 10In the discussions which follow no effort has been made to detail every sweetening process that has ever been developed. Instead, emphasis is placed on those processes which have major current or potential application for the sweetening of natural gases. Attention has been given to some processes which are, or have been, primarily used in manufactured or refinery gas sweetening but whose development and technology (for one reason or another) have interest or potential application in the natural gas industry. Many of the processes discussed have never reached commercial production. They are discussed because they either show promise for application, because of the interesting technical problems they present, or the uniqueness of the concept behind the process developed. Every effort has been made to bring together those data available that will permit the engineer to perform his own calculations and make his own economic and operational evaluation of the various processes. The amount of information available on different solvents and processes varies widely. Where no information is available, the author can only express the hope that the necessary data will soon be developed and released. Unless comparative detailed calculations are made in every instance, there can be no assurance that the optimum sweetening process has been chosen. REFERENCES 1.1) Huntington, A. L Natural Gas and Natural Gasoline. New York: McGraw-Hi Bottoms, R.A. US. Patent 1,783.90" (1990), Chem. and tet. Eng. Hyne, J. 8. OW and Gas Journ No. 1, 1940, p. 97. ‘Aug. 28, 1972, 9.66 Chemical Engineering, May 14, 1973, p. 8, room, A. H. Hydrocarbon Processing, July, 1972, 6.78 Newton, 8. F., Hydrocarbon Processing, Jan., 1978, p. 181 Rangnow, D. G., and P, A. Fasullo, Ol and Gas Journal, Sept. 28, 1981, p. 242. ‘Archibald, R. G., Hydracarbon Processing, March, 1977, p. 219, Tennyson, A. N., and R. P, Schaal, The Oi! and Gas Journal, Jan, 10, 1978, p. 78.2 BASIC PROCESS PRINCIPLES Volumes 1 and 2 of this series discuss the process principles that apply for sweetening processes. But, in order to enhance the use of this volume as a desk reference or for training, certain of these principles are summarized in this chapter. Basic Dimensions In all processes we are required to express quantities in terms of certain fundamental dimensions. ‘These fundamental dimensions are more or less arbitrary assignments of value from which many derived dimensions and quantities can be developed. In most cases the fundamental dimensions are considered to be length (L), time (6), mass (m), force (f) and temperature (7). Table 2.1 summarizes the conversion between these fundamental dimensions for several sets of common units. Much emphasis is now being placed on conversion of all activities to Le Systeme International d'Unites (SI) system of units. Most technical publications are in the process of converting all of their published articles to this system of units. The fundamental units in the SI system are: Meter — the length equal to 1,650,763.73 wavelengths in vacuum of the radiation corresponding to the transition between the levels 2P4o and Sdg of the krypton-86 atom. Second — is the duration of 9,192,631,770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the cesium-133 atom. Kelvin — is the fraction 1/273.16 of the thermodynamic temperature of the triple point of water. Candela — is the luminous intensity, in the perpendicular direction, of a surface of 1/600,000 square meter of a black body at the temperature of freezing platinum under a pressure of 101,325 newtons per square meter. Mole — the amount of substance of a system that contains as many elementary entities as the number of atoms in 0.012 kilogram of carbon-12.Joule — the work done when the point of application of a force of one newton is displaced a distance of one meter in the direction of the force. Newton — that force which, when applied to a body having a mass of one kilogram, gives it an acceleration of one meter per second squared, Watt — the power which gives rise to the production of energy at the rate of one joule per second, Pascal — the pressure or stress of one newton per square meter. Common Systeme of Unite uiemd Engineering] Absolute | Technical | Absolute Metric] Bapinceving ons Engen | Engitsh | engtian cae MKe | Metric | pound pound gram | wctogram | sstogrem Mass | "aes fase | etog 7 degram | sstogran vey | BRS fam) | thle) | tae, ) ee ane ‘er ier | em to ten _ | t Time 1 second | second | second cond second eset “eect ‘ech ‘sce) eee Force | pound | poundal pound Newton | Mlogsam Floree ipa Bree aye i force io ‘be tog | 4 Tynre | Rageine | Satuine i Rein Keim | “Kista ah Ry Cai i Ca 1545.33 |astioxo? [acon [asiseao? [anaoao® | earsaact ar0ttop cera — | ate) — [taynertersd | atlimy | (xstogCem) Pra) RE Jem | Rey cw [fer Jes ote) CKD 22.1739 Lo w | ue 10 900. 665 Valse of Jamey — oe tey | fatugnie ito, Mem) Lf _| eptees | fetvaecs Table 2.1 Summary of Common Systems of Units (Ref. 2.1) uGas Digest (2.13) presented a collection of S! derived units that would be of interest to the gas Industry. The units suggested are shown in Table 2.2. Derived Units for the Gas Industry Standardization of units is still in the embryo stage worldwide. Many non-standard derived units are employed commonly. Consider the bar (1 bar = 100 KPa). There is still no unanimity of opinion on standard conditions although the majority of the petroleum industry is now using 101.925 kPa (14,696 psia) and 15°C (almost 60°F). The application of different units always is frustrating. Fortunately, in the engineering applications, small differences in standards andlor conversion factors present no problem. In this book many of the TABLE 2.2 ree conversion factors have been rounded off for convenience. Appendix 1 shows some convenient conversion factors between units in addition to those in this chapter. 16 int i?MASS AND ENERGY BALANCES Two of the most powerful tools the engineer has in his bag of tricks are the material or mass balance and the energy or heat balance. The material balance can be used to determine the production from a plant, the necessary flow rate of treating solution, the size of pump required, and many other variables. The energy balance can be used to determine the heat input required in a reboiler, the horsepower of a pump, or the heat duty of a heat exchanger. Ih order to use and apply heat and material balances properly, we must understand the thermodynamic definitions of the system and the surroundings. The system is very simply defined as that Portion of the universe in which we are interested and concerned at a given time. The surroundings are all other parts of the universe save the system, though ordinarily we are most concerned with that part of the surroundings adjacent to the system. Mass and energy are considered to be “in” or “out” of the system any time one or the other crosses the arbitrarily defined boundary of the system. Mass Balances For a given defined system we can express the mass balance for a given period of time as: mass change in system in given time period = (mass in during that time period) - (mass out during that time) + (mass gained or lost by atomic transformation) In symbolic form, the mass balance can be expressed:as Me - Mg = 2M) = IM) + Mp (2.1) where: Mp - My EM, = summation of all mass entering the system in the time period summation of al1 mass leaving the system in the time period Np = gain or loss in mass from atomic transformation In the processes that we will consider, Mp (the gain or loss in mass caused by atomic transformation) is ze10. mass change in system in time period My ‘Amass balance can be written based on the total amount of material crossing the boundaries of the, system. In this case it is commonly referred to as an overall material balance. In some instances there is advantage to writing a material balance for one of the components in the streams entering or leaving the system. In this case the balance is referred to as a component material balance. There are even instances when there is good reason for writing a material balance based on one atomic species such as the sulfur in H,S and other compounds entering or leaving the sweetening process. In working with gas one is continually faced with the problem of conversion from gas volume to either liquid volume or equivalent mass rate flow. Shvartsburd (2.14) has presented conversion factors that are helpful in making the conversion from volumetric to mass flows. Table 2.3 shows the factors he suggests for volume conversions and Table 2.4 shows conversions for gas volumes under different temperature and pressure conditions. 16Table 2.3 Conversion Factors for Volume Related Units (at equal conditions of measurement) Conversion factors te 98) 0y— Trem From ‘onventonal —comantional US, metic teconvetional te conventional Tmebiesystem US. system 7 se} gee ae a ca ee Sa wid | i Table 2.4 Conversion Factors for Gas Flow Rates (under different standard conditions of measurement in conventional U. S. and metric systems) ‘Conventional | __ Conventional —1S.astea—} — netespien — seth set | tari seryte 14E86psia 14.780 psig | 760mm He 760mm He Cu A Somenion | set 1 09977 | 02679 oze7s sien, =f 1.0023 1 | oneses —oczese am'/he 160 am He Comaton | dim) ROE 3726 S72 10732 sete 160. ma amin "REPT 700 sasoo | ose FReference baw conversion factors at equal condions are’ F US. to mew) = 0063168: F imei wo US) = 353467, V7Solution We can evaluate the individual terms in the energy balance equation as follows: AP _ 1000 - 23 AF . 1000 - 23, . at x 146 The density of the MEA solution can be determined from Figure 2.1. If the flowing temperature of the amine solution is assumed to be 100°F., we find that the density is 0.999 gm/cc. In 1b/cu ft this will be 0.999 x 62.4 = 62.34 Ib/cu ft. AP then becomes ® sa OU saN ee een 9” 62.34 1, The difference in elevation will be the difference in level between the pump discharge and the top of the absorber, which was specified to be 60 feet. In most cases, the velocity terms are so small as to be negligible. Unless there is a change in line size or the line discharges into a tank, the velocity term is 0. The two remaining terms in Equ. 2.3a, Wp and W, cannot be evaluated at this point. There is insufficient information. The final equation as written then would be: 2257 + 60 +0 = The implications of this are that, according to the definitions given earlier for work, the pump must do work on the amine solution in order to pump it to the higher pressure, This, of course, is perfectly logical and reasonable and does explain the negative sign which is certain to appear on the external work term in Equ. 2.3a as written for this problem. Friction Loss ‘The friction loss term in Equ. 2.3 and 2.3a includes essentially all of the irreversibility effects for the fluid flowing in the line. It can be expressed as: = 2fu7 (2.4) where: friction factor, dimensionless fluid velocity in ft/sec ‘ equivalent length of straight pipe in feet inside diameter of pipe in feet 20iz L109 i) 130 oso EE 1080 1080] toro loro oso oso tos0 Loso| too & 104d 030 § 1030] 1020 $ Lozole 1.010 . 1.010 L000 ff 0990 o.s90fi7 960 0.00 0.970 Peer 0.970) 0960 ET 30 400 60 70 G0 90 100 09600 20 30 40 50 80 70 60 90 100 MONOETHANOLAMINE, PERCENT BY WEIGHT DIETHANOLAMINE, PERCENT BY WEIGHT Figure 2.1 Density of MEA solutions Figure 2.2 Density of DEA solutions (Ref. 2.5 Courtesy Dow Chemical Co.) (Ref. 2.5 Courtesy Dow Chemical Co.)The “friction factor” fhas been proven to be a function of the Reynolds number Dup/1! the roughness of the pipe wall and the type of flow (laminar, turbulent or transition). In nearly all applications of industrial interest the flow will be in the turbulent regime (Reynolds number greater than 4,000). Variation of friction factor, f, with Reynolds number is shown in Fig. 23. The Reynolds number is a dimensionless number which means that all terms used in calculating the Reynolds number must be in consistent units. Determining the friction loss with Fig. 2.3 requires calculation of the Reynolds number, the friction factor, and the friction loss from Equ. 2.4. In terms of readily available numbers, the Reynolds number can be calculated as: 3150) S. (2.5) a tas where: g = liquid flow rate in gallons per minute S.G. = liquid specific gravity referred to water d= inside diameter of pipe, inches v= liquid viscosity in centipoises O= PIPE DIAMETER FEET Ve LNEAR VELOCITY FT/SEC, PeONSITY-CB PER CUFT, pe iscosiTy- Laver see Fa prction FACTOR FRICTION FACTOR - ooo te POE gs ee REYNOLDS NUMBER = 02> Re Figure 2.3. Friction Factor as a function of Reynolds Number (Ref. 2.1) Other versions of Fig. 2.9 are available. Some of these use different forms of Equ. 2.4. This means that fone must be careful to use the graph for determining friction factor that corresponds to the form of Equ. 24 used for calculating friction loss. In some instances the effect of wall roughness Is included in terms of a relative roughness factor. This factor usually relates the “height” of the average roughness to pipe diameter. As a practical matter, the two curves showing smooth tubing and commercial pipe on Fig. 2.3 bracket the range of pipe conditions usually found. For the convenience of those using programmable pocket calculators, Fig. 2.3 has been fitted to equation form by Churchill (2.20). The form of his equation is: 2212 = 2 [¢aney!? + 1/(as8)°/71 (2.6) 16 ] where: qs [2.457 In (7/Re)°"? + 0.27¢/0 B = (37,530/Re)!® © = pipe roughness Equ. 26 fits Fig. 2.3 for all values of Reynolds number. The solution is direct if flow rate through the pipe is fixed, but trial and error must be used if pressure drop is fixed. For most cases of industrial interest, the term (0.27 EID} in the denominator of Equ. 2.6 will be very small. For fully developed turbulent flow, Equ. 2.6 reduces to: 1.559 In (FS) (2.7) VF Equ. 2.4 and Fig. 2.3 can be solved nomographically. When doing this, some simplifying assumption generally is made. Fig. 2.6 is a modification of the nomograph presented by Lewis (2.2). His assumption is that the friction factor is constant at a value of 0.004. Fig. 2.6 can be used as described below for several different friction loss cases of interest to the process engineer. Example Problem 150 gpm of a 25% diethanolamine (DEA) solution is being pumped through a 3 inch inside dianeter pipe. If the flowing temperature of the amine is 120°F., calculate the pressure drop in the line per 100 feet of pipe. Solution For a 25% DEA solution: S.G. = 1.012 (From Fig. 2.2) = 1.3 ep (From Fig. 2.5) The Reynolds number then is = 3150_(150) (1.012 had G.0)-3) 122,600 From Fig. 2.3 using the line for commercial pipe, we read a friction factor of .0051. We must now determine the velocity in the line. 150, 1, 4x 144) = Fam * Go * Foyt ~ 68 ft/sec 23100, 1909 20 988 0 00 a “400 : 200 20] 199 28 i 6 gee 34 =o - z 2 a : . ea Ba Ee 8 2 gs Ss, 7 08 a os ' \ oa of oe oa oz E oa | : ou g 50 60 70 80 0100 120 140 160 180 200 230 ou, TeMeRATUREITE 80 60 70 80 80100 “120 140 160 180 200 — 230 TEMPERATURE, * Figure 2.4 Viscosity of aqueous MEA solutions as function of weight percent MEA (Ref. 2.5 Courtesy Dow Chemical Co.) Figure 2.5 Viscosity of aqueous DEA solutions as function of weight percent DEA (Ref. 2.5 Courtesy Dow Chemical Co.)We can now determine the friction loss as ft - 1b, 2(0.0051) (6.8%) (100) 7 1B, 32.2(3/12) Wee 5.86 ‘The pressure drop per 100 feet of line is calculated from the head loss and the fluid density as follows: = Gah) () _ 5.86 x (1.012 x 62.4) ar 194 ae 2.57 psi/100 ft 1. Example showing use of Figure 2.6 to find the pressure drop in 2 given line for a specified flow rate: A. Locate a working point on the reference line by laying a straight edge from the density scale to the flow scale. B. Move horizontally from the working point to the diagonal line representing the correct line size. ©. Read the pressure drop as indicated by the vertical lines at this point. Use Fig. 2.6 to determine the pressure drop for the 25% DEA solution. The pressure drop for this was previously determined using Fig. 2.3 and Equ. 2.4. Yb/be = gpm x 60 x 9 = 150 x 60 x (8.33 x 1.012) = 75,870 From Figure 2.6: AP = 2.2 2. Example to find the line size giving a desired pressure drop: A, Locate a working point as described in (1) above. B. Move horizontally from the working point to the vertical line showing the desired pressure drop. C. Read the line size on the diagonal scale. Diglycolamine is being pumped at a rate of 100 gpm. The pressure drop in the line is 1.2 psi/100 ft of line. If the diglycolamine solution is 60% by weight DGA and the flowing temperature is 100°F., what is the in- side diameter of the line? Solution From Fig. 2.7, $.G. = 1.047 Ib/he = 100 x 60 x (8.33 x 1,047) = 52,330 P= 62.4 x 1.047 = 65.3 From Fig. 2.6, i.d. = 3.0" 250.01 000,000 600,000 0.02 400000 0.03 O04 0.06 fanten ot 100,000 a 60,000 40900 ea i 30,000 i 20,000 06 3 2 caer gi 10,000 ge = es] 4000 5 = : z 3900 4 Be 3 we a e ie to 3 1900 a ae a 600 5 83s 400 8 30 2 S| 300 40 5 oe 2 200 a 100 100 70 200 50 Figure 2.6 Nomograph for finding line pressure loss (Ref. 2.2 Courtesy Petr. Ref.) 26- Example to correct for a different friction factor: A. Hind 2 working point as in case (1) above. B. Move diagonally to a value equal to 100 times the new friction factor as shown on the pressure drop scale. ©. Move horizontally back to the reference line to find the cor- rected working point. Continue with steps B and C as in cases (1) and (2) above. Suppose for the problem in case 1 we wish to adjust the Friction factor to its correct value of .0051. Solution Using the working point found in case 1 we would then move along the diagonal pipe diameter line to a value of 0.51 (100 x .0051) and then horizontally to read a pressure drop of 2.7 psi/100 ft of pipe. 4, Example to find the pressure drop in a new line size: A. Pind the point on the chart where the old line size intersects the existing pressure drop. B. Move horizontally to the new line size and read the new pressure drop. In the diglycolamine example suppose that we wish to find what the pres- sure drop would be in a 4 inch diameter line. Solution We would move horizontally from the intersection from the 1.2 psi in the 3 inch diameter line to the 4 inch diameter line and read a pressure drop of approximately 0.25 psi per 100 ft of line. 5. Example to find the allowable flow in a given line: A. Use the specified line size and allowable pressure drop to find a point on the chart. B, Move horizontally to the reference line to find a working point. Lay a straight edge from the density scale through the working point and read the allowable flow rate on the flow rate axis. We wish to find the amount of diglycolamine that can be pumped through a 2 inch inside diameter line with an allowable pressure drop of 2.0 psi/ 100 ft of Line. 27Solution Connect the density point (65.3 lb per cubic ft at 100°F.) with the reference point found from the 2 inch i.d. and 2.0 psi line and read from Fig. 2.6, 26,000 1bs/hr gpm x 60 x (8.33 x 1.047) = 26,000 gpm = 49.7 Use of Fig. 2.3 and Equ. 2.4 or Fig. 2.6 requires the total equivalent length of “straight” pipe in the line. n order to determine this equivalent length of straight pipe, the valves and fittings In the line must be converted to their equivalent length of straight pipe. This can be done in a number of ways. However, the “two:K" method Is the one applicable over the widest range of Reynold's numbers and flow conditions. In the two-K procedure the total head lost in the fitting is expressed as an “excess” loss over that which would be experienced in straight pipe. The excess loss is expressed as: aH = KHy (2.8) where: AH = excess head loss in the valve or fitting K = dimensionless multiplication factor for head loss Hg = the number of velocity heads in the flowing gas stream = v7/2g. K in Equ. 2.8 is a dimensionless factor defined as the excess head loss in the pipe fitting expressed in velocity heads. Hooper (2.15) expresses K as: K = Ky Mig + Kyo(l + 1/10) (2.9) where: Ky = K for fitting at N= 1 Koo = K for a large fitting at N., = 00 ID = Internal diameter of the pipe, inches Values of the constants Ky and Kg for use in Equ. 2.8 and 2.9 were presented by Hooper and are shown in Table 2.5. Table 2.5 covers the normal pipe fittings and valves with the exception of three special cases. In these three cases Equ. 2.10 is used to predict the value of K. ie (2.10) K = Ky Mee + Koy 28le 2 30 40 30 60 70 #0 $0 100 pievycot amine Owetent reRceKT Figure 2.7. Density of aqueous DGA solutions (Ref. 2.6 Courtesy Jefferson Chem. Co.) VIReosiTY, centpoine DIGLYCOLAMINE , WEIGHT PERCENT Figure 2.8 Viscosity of aqueous DGA solutions (Ref. 2.6 Courtesy Jefferson Chem. Co.)Equ. 2.10 is used for: 1. Pipe entrances. Ky = 160: Kgg = 0.6 for “normal” entrance and Koo entrance, 2. Pipe exit. Ky = 0: Kog = 1 3. Orifice, There is no constant value for Ky Equ. 2.9 and 2.10 are difficult to apply to orifices. TABLE 2.5 Constants for Two-k Method iting type Ky Ke standard (RID = 1), eerewed 800040 Standard (RID = 1), tlagediveléed 8000.28 Long-radius (R/0 = 1.5) all ‘ype 900 0.20 so 7 Word (80° ara Mivered | 2 Weld (48° ane ‘800 035 3 Wold (30° srsies) 8000.20 4 Woig (225° anges) Bi 5 Weld (18° angles) ‘Standard (R/O = 1, all pee ‘500020 age |Lonevodive (R/0~"1Sh.all types 5000.15 Miered, 1 weld, 45° angle 500 025 Miered, 2 weld, 22% angles 500028 [Standard (A/D = 1) serened 7000050 190" | Standard 1R/D~ 1), fiansed/welded 11000 0.35, Long radius (R/D=7.5),all types 1,000 0.30 Stendad,acrwed «ODOT es | npradn sree B00 0.40 | Stra, ange or welds 800 0.0 ews [220% _| Stainevoe ranch 1.000 1.00 Fon | Serene 2000.10 shou | Flanged or welded 180030 tee [Stabinsyoeoraney 1000.00 at, [Fil ine sie, 9-10 3000110 bat, | Reduced vie, 9 = 08 500016 910g _| Reduced vim. 8-08 1,000 _ 0.25 [ cove, anaara 1500 4.00 aves | 10S anne oF Yaype 000 2.00 | Disragm, cam type tc00 2.00 Buterty 00028 tu 2000” 10.00 cnet. Sing 1.500 1.50 Tints 900 080 Note: Use A/D = 1.5 values for A/D = 5 pips bance, 45°10 180%, ‘Use approorate ine value for flow through cross, 30 1.0 for “borda”Use of Equ. 2.9 and the values from Table 2.5 will be clarified by the following example: For the pressure drop example calculation estimate the pressure drop for a standard weld elbow. K = [800/122,600 + 0.25(1 + 1/3)] = 0.34 Hy = 6.87/32.2 = 1.44 ft-1b,/1b Py AH = 0.36 x 1.44 = 0.49 fet, /1b, PUMP POWER ‘The horsepower of a pump required for pumping a fluid between two points in a line generally is expressed in terms of the pressure difference across the pump. The required pressure difference is determined from proper application for Equ. 2.3 between the desired points in the pipeline. The pump horsepower required is expressed as: ee (2.11) {35,000} (7-48)(E) where: Q, = the liquid flow in gations per minute Wo= the work requirement in horsepower Po-P = pressure gain in pounds per sq ft E = the pump efficiency Equ. 2.11 is a form to be solved easily through nomography techniques. Tan (2.16) has presented a ‘nomograph which allows estimation of the required horsepower input to a pump, and at the same time allows for estimating pump efficiency automatically according to the hydraulic institute standards. The ‘nomogram prepared by Tan is shown in Fig. 2.9. The brake horsepower requirement is determined easily by connecting the pump capacity on the left axis with the differential head on the center axis and extending the straight line to the brake horsepower axis on the right. The pump efficiency may be estimated by: ep = 1- (1.6/(6°M)°?7] (2.12) Ep = Pump efficiency expressed as a fraction © GPM = Gallons per minute of fluid being pumped Example Problem Determine the horsepower required to pump 200 gallons/min of m-pyrole against a total pressure head of 250 psi. Solution = 200_x (250 x 144) 33,000 x 7.48” 29°? BP 31This represents the so~called hydraulic horsepower or the amount of work that must be transmitted to the liquid (in this case the m-pyrole) by the pump. ‘The horsepower of the pump driver must be greater than this because of the efficiency (B) of che pump. 0.0 ata oes a 000 7 co sf. ow : 1400 5 oa ote 080 00 ry om x 0 soca} 076 one a on, co seo -o70 i i H 7 : 2 _ 2 i a i : : 7 7 : i j i i : i 2 4 i cea eee 1 aut i 0 é bat i “te & ae = 900 = 109 1000 S200 108 020 109 040 Ps aaron 0 0 of sco L109 Figure 2.9 Pump Horsepower Requirements 2.9 also may be used. Connecting 200 gallons per minute on the left hand axis with 250 psi differential head gives a brake horsepower of 49. This is different than the number calculated using Equ. 2.11 because it automatically includes the pump efficiency. The pump efficiency from the right hand side of the left axis of Fig. 2.9 Is about 61%. Adjusting the calculated horsepower for this efficiency gives 48 horsepower. This is about as good agreement as could be expected when reading a nomograph, 32HEAT TRANSFER Heat transfer normally is expressed in terms of the overall heat transfer coefficient as: Q = Unt (2.13) where: Q = rate of heat transfer, Btu/hr _ U = overall heat transfer coefficient, Btu/hr, sq ft, °F. A = area through which heat is being transferred in sq ft At = the temperature difference “The temperature difference for heat transfer usually is expressed in terms of the log mean temperature difference Atm. For a concentric pipe-in-pipe exchanger Atm is found by: aty - Ate (2.14) im Talhty/Bty) at, Example Problem Lean amine solution enters an amine cooler at 160°F. and is cooled by water to 110°F. Water enters the cooler at 85°F, and is discharged from the exchanger at 105°F. What is the log mean At for the exchanger? Solution bt, = 160 - 105 = 55° de, = 110 - 85 = 25° 2 55-25 30 at, "in 55725 " In 2.2 ~ 38-1 Lf the area of the exchanger is 600 square feet and the overall heat transfer coefficient is 135 Bru/hr/sq ft °F., how much heat will be transferred from the amine to the water per hour? Q = UAde, = 135 x 600 x 38.1 = 3.09 x 10° Beu/ar For counter-current flow of two fluids in an exchanger composed of two concentric pipes, Equ. 2.13 applies exactly. For other exchangers, such as shell and tube exchangers Atm must be corrected for the lack of true counter-current flow. The correction is made by multiplying Atm by a correction factor F. Charts for estimating F are available in Volume 1. Coefficients In Equ. 2.13 there is an overall heat transfer coefficient U and an area A that are related to each other. In the general case heat will be transferred through three resistances: a liquid film coefficient on one side of the tube wall, the resistance of the tube wall and the resistance of the liquid or fluid film on the other 33side of the tube wall, The overall film coefficient can be based on any of these areas. The amount of heat transferred through each resistance in series is the same, but the areas of the resistances are different. This means that the overall heat transfer coefficients must be different if the rates of heat transfer per unit of time are to be equal. The overall heat transfer coefficient is related to the individual film coefficients as shown in Equ. 2.15. 2.15: 7 a (2.15) ary hy = film coefficient for fluid 1 ip = Film coefficient for fluid 2 thermal conductivity of tube wall A, = Surface area of pipe for film hy urface area of pipe for film hy verage wall area of pipe or tubing thickness of wall of pipe or tubing (eee: Predicting the individual heat transfer coefficients for use in Equ. 2.15 is frequently a difficult and time-consuming task. Also, particularly in the case of shell and tube exchangers, exchanger configuration and geometry have considerable effect on the individual coefficients. Consequently, determination of the individual coefficients is best left to the heat exchange expert who has the physical dimensions of the particular exchanger. Use may be made of Fig. 2.16 which relates individual and overali heat transfer coefficients through Bedebh (2.16) pM overall heat transfer coefficient coefficient for process fluid coefficient for thermal fluid Fig. 2.16 is obviously an approximation to the overall heat transfer coefficient, but its simplicity and speed of use are compensating factors. In the example shown by the dotted line, an organic liquid is being cooled by cooling tower water. By connecting the appropriate points on the thermal and process axes, the overall coefficient is shown to be approximately 90. The value selected in the range of individual coefficients should be at the upper limit of the range when: the condensing system is free of inerts; the thermal conductivity of the liquid is known to be high; the viscosity of the liquid is known to be low. The value should approach the lower limit when the viscosity of the liquid is known to be high. A factor for fouling or scaling of the heat transfer surface frequently is included in the overall heat transfer coefficient, 34(2.17) 1 alee bir 1 KAY Rea The factor F in the denominator of Equ. 2.17 is to allow for the resistance to heat transfer of the scale that forms in many heat transfer services. Values for F are available in many works on heat transfer. Typical values for many services in the gas processing and sweetening area are presented in Table 2.6. The values given are based on plant data, but not data gathered in a systematic investigation. Consequently, the values in Table 2.6 should be looked upon as guidelines rather than firm values of the fouling factor. 8 os S eo ve ‘ of tionnmee wo . Boing oegarics “ a enim omnes S/F e oats SO S ‘Heavy organic WX Moms SX ¥ * ee 6 wate SN, itn 7A * \y Sere we ee ee em | a F ; al gag Swen | OEE ‘rena oe ie Mencs, cg Rar Hot \a condense saat ine Refrigerants ‘Thermal fluid coefficient, h,, Btu/h-ft? °F Cue Figure 2.10 Nomograph for estimating overall heat transfer coefficients. (Ref. 2.10) The fouling factor (F) in Equ. 2.17 is the only thing that differentiates it from Equ. 2.15. This Fis thus an additional resistance in series, Across a given surface there may be fouling on both fluid walls. Thus, Equ. 2.17 could have two values of F in the denominator. Typical values overall heat coefficients for various process services are given by Bell (2.17). These values also are shown in Table 26. 35TABLE 2.6 Overall Design Coefficients for Shell-and~Tube Heat Exchangers Tota fang eaatan u ig fia? neReR tus eruriet ‘Water q 0.0015 280-300 Mate. PGi eta 10 psig Soot is 20 Water as about 1007p | DT a0 ao Water | Gee sot toe ese | ot | Ro ter Upitcrgiie fauas, =| Dols, | age ye Water Meum organi iglds to | FEI Water Hea eraate ign, S095 as Witr ve | oe | | | 1 we | \ 13 [Gas about 10 psig —«am00g is ao Ges, aout 100 ig eos 3 Bis aout 800 pte ‘pos 7110 | Tira qu 001 135199 Medimgsie tts | Danis 20133 | Heavy ergot tuts oom 0 ‘ey eaty reac nas 158 Tits | water 201 300400 Take aoe io0.190 | Mei onsets | ams | roi Reavy erane nc) iiguide | tating wp Cooling Bo very hey organic ‘cot Fase, a i = eto one Medium orgenic liguds 0003) 0 gues Medi organic Meavy organic £ ons gess ecgst ! Heating | 0% tecting | 88 verry ewmnic | aonts eve | Heating | 15 0 Coal ! 58 eae oven Heany evga Fels ny wie Heavy orgie Very eae ora 55 ude tule Gx, about 19 oi Gas, about 10 psig 015 Gis, bout 10°9s¢ | Gs: sout 100 ong 50 os shou 10 ose | Eas abet C000 pig 1B is, abet 100 peg |x a0 is, sbut 100 bait BS Gos about 1,000 psig =o Water ~~ Condensing tight 150-200 i Water condensing medi 001 e080 organ apars fe comport Meter Ccondesing bee ome 75199 ‘rg rapre, ate component 36TABLE 2.7 sypteal overall Goeffictests for Aerie! Coolers are Tobe Approximate Ate condenses cuelfisiese Temperatare Alse Propane, Bocene ” a Sell! Overnese 5 iS Stese 050 pet a % Rapnene 8 2 Nah Regeoerator ‘vested so ° vatural Garolsoe 0 2 Freon ” % » 0 > o ° » 0 3B 3 » 5 8 TABLE 2.8 Fouling Factors Trea foting rsistences ae vetoed io the srtace on which hay occu, Inthe absence of pec dit or surg ope vaistances Ine ust maybe Qulced hy he vases Leuleea below Upto 20° 210s00"F Temperature of Water eras oer a0] oe Son wate ‘00008 ] 0000s o001 frocten water ‘oon | coor 000s Coating tower ane anita spray pond on | cco: once ‘ty or wa water uch a8 Groat Lake) oo! [eco nana (rout Lakee ‘oot feo oon ier Vater Maddy or aty oc | 000 ooo Hard (vee ranagal 003 }oms 008 Engin lack ‘ooo [aso — ooot ities ons | coos oams Teste tlie woewaer an” | eons — oaor Daler Blowdown aoc | oo oon “Finn a calm # and & ae bend on «Temperate of he hating modi 01240400" onthe ‘cooling medio kxown a sea, hove rings shoud be mole ecco sw, ndatalorgarichent vanaar meale oom Festa ooos Stale ide raw anc bottom produt v<.ores-1- 2001, Engnetube a 2008 Matenneattranstraate 20008 Gavailcsetlons.-. e2 MEAEDEA soletont mz Lean ‘e002 Dea aTEG otutona Seog Rieko 00: Retegerat que ‘ooot Sage and Vana Travotolorgalchont vanaer media 2001 ovement products... 0001 Retngeratervarers (ll beeing 001 Manufactured gan 004 Exnaoststeum ol betng) "0001 Comorests a ‘cz Satie overneadproaucte ‘omot oor ‘Steam (ronal beeing) asses 0s aot Engine exnevst gee ont 01Exchanger types other than shell and tube ate becoming increasingly popular in the natural gas Industry. New exchangers such as Rod: Battle, Plate, and extended fin aerial coolers are frequently used. Predicting heat transfer coefficients for these exchangers is more complicated than for shell and tube exchangers. However Table 2.7 does show typical values of heat transfer coefficients or aerial coolers in typical natural gas services. VAPOR. LIQUID EQUILIBRIUM There are many ways in which equilibrium between vapor and liquid can be expressed ‘mathematically. Perhaps the most widely used expression is shown in Equ. 2.18. Kx; (2.18) mole fraction of one component in the vapor phase mole Fraction of that same component in the liquid phase equilibrium constant The equilibrium constant is not really a constant in Equ. 2.18. Itis a function of temperature, pressure and composition. However, charts and graphs are available for most of the hydrocarbons and natural gas constituents. For many sweetening agents no experimental data are available and resort must be made to estimation techniques. One of these techniques makes use of the fugacity of the component. Using fugacities, Equ. 2.18 can be written as (2.19) where: f{ = fugacity of component in the pure vapor state v f§. = fugacity of component in the pure liquid state iL Fugacities can be determined from Fig. 2.11 as functions of reduced temperature and reduced pressure. Equ. 2.18 can be used to predict the composition of the vapor in equilibrium with a liquid of given composition or vice versa. It cannot, however, be used to solve the most frequently encountered process problem — at a given temperature and pressure, what proportion of a specified stream will be vapor and what proportion liquid? Based on Equ. 2.18, we can derive an expression for determining the amount of liquid or vapor present.(2.20) (2.21) where: £, = moles of one component in the liquid formed f} = moles of one component in the feed v total moles of vapor L_ = total moles of liquid Equ. 2.20 permits estimation of the amount of liquid formed when a stream of specified composition is flashed at a given temperature and pressure. The corresponding equation for the vapor formed is: KR (2.22) (2.23) where v, 4 7 moles of one component in the vapor formed Equ. 2.20 and 2.22 cannot be solved directly — the solution must be trial and error, We must know the value of V and L or the ratio of the two in order to solve either of the two equations. However, in the solution of either equation we can calculate the total tiquid or vapor formed. The procedure for solution is to assume a value for the total amount of liquid or total amount of vapor formed. Using this as a basis, the liquid or vapor formed is calculated by either Equ. 2.20 or 2.22, When the assumed value and the calculated value are the same, the proper solution has been found. There are rules which, if followed, will make the solution of Equ. 2.20 and 2.22 simple and rapid. 1. Always use the form of the equation that will permit you to calculate the stream that you expect to be present in the smallest amount. For example, if you expect only 10% liquid and 80% vapor, you would use Equ, 2.20 and calculate the amount of liquid formed in the flash. 2. Once a calculation has been made, the following rule always applies: If the calculated value differs from the assumed value, the next assumption should be even further in the direction of difference, Use of these two rules will help cut down the calculation time involved in solving either Equ. 2.20 or 2.22, Example Problem A stream of the composition shown below is to be flashed to form liquid and vapor under conditions such that the equilibrium constants for the different components are as shown in the table. Calculate the amount of liguid and vapor formed in the flash. Because we expect less vapor to be formed than liquid we will use Equ. 2.22. 39FuggerTy goErrcieNT REQUCED PRESSURE, P, Figure 2.11 Generalized correlation of the fugacity function in the form of a fugacity coefficient (Ref. 2.1) 40Solution mol fr Assume Comp, £) K Ke V = 0.014 0 ~30% 61 ~207 70029 60, -016 5.28 +0845 +0011 oy +024 50.5 1.212 -0100 Solvent 656, 0 ° 1.000 +014 Since the assumed value for the moles of vapor formed and the calculated value agree, the calculation is correct. As an explanatory note, the solvents used in gas sweetening normally are so non-volatile that they have essentially a zero K value at normal temperatures and pressures- ABSORPTION AND STRIPPING In the case of absorption and stripping operations where physical absorption is involved, we can predict solution circulation rates required through the use of the simplified Kremser and Brown type absorption calculation. Y -Y yy (2.24) ntl ° where: Y, = moles of a component in the entering rich gas stream per mole of entering rich gas Y, = moles of a conponent in the dry gas leaving the absorber per mole of entering rich gas = moles of a component in the gas stream that would be in equilibrium with the lean solvent per mole of entering rich gas n= number of theoretical plates in the colum A= average absorption factor = Lo/(KV,41) L, = moles of entering lean solvent Yel = total moles of entering rich gas If the lean solvent is essentially free of a given component, Yo wili equal 0 for that component, Equ, 2.24 has been solved and plotted in graphical form for convenience, This graph is shown in Fig. 2.12. In acalculational sense a stripper is simply an upsidedown absorber. Trays are numbered from the bottom up with the top plate designated as m. Equ. 225 shows the proper form of Equ. 2.24 for use in stripping. 4Kae 7% ml nth (2.25) Mma % SMT where: S$ = stripping factor = KV/L Xj = moles of component in stripped lean 011 leaving bottom of the stripper per mole of rich oi] entering the stripper X41 7 Moles of @ component in the rich oil entering the mt" stripper per mole of rich ofl entering the stripper Laer = Moles of rich oil entering the stripper + Y, = moles of component in stripping medium per mole of stripping medium entering the stripper V, = moles of stripping medium entering Xq = moles of component in liquid in equilibrium with stripping medium per mole of entering rich oil Ly = moles of lean oil leaving the stripper Example Problem We desire to recover 65% of a component froma gas stream. With the solvent used and at the temperature and pressure of the absorber the equilibrium constant (K) for the component is 1.2. How much solvent must be circulated over the absorber? Solution We will use Equ. 2.24. If we assume that the solvent is completely free °F the component to be absorbed, Y, will be zero and Yaa ¥, gat tae amt In this form the left hand side of the equation represents the fraction of the component absorbed or, in our case, 0.65. If the absorber sonreins che equivalent of four theoretical trays we determine from Fig. 2.12 that the absorption factor A must be equal to 0.715. wy, nt] * 0.858 mols solvent /nol gas 42Figure 2.12 Absorption factor chart Equ. 2.24 and 2.25 apply to physical absorption and stripping operations. In those cases where a chemical reaction is involved in the sweetening process these two equations may or may not give reasonable approximations to the efficiency of the operation. Predicting rates of chemical reactions is a difficult and involved process. For those cases where chemical reaction is involved the usual procedure Is to use either pilot plant data or data on an existing plant as a guide for developing a workable, usable process. VESSEL SIZING ‘The sizing of any vapor-liquid separation device, such as an absorber or a separator, is dependent on several factors. The primary driving force for the separation is the difference in density between the liquid and the gas. Equations describing the separation that should occur are relatively easy to formulate from the principles of fiuid mechanics and fluid dynamics. There are many practical variables, however, that are difficult to describe, such as particle size distribution, presence of entrained solids, etc. For this reason, nearly all techniques for vapor-liquid separation design make use of safety factors based on ‘experience and past performance of vessels that have been designed. A simple but accurate procedure is to use Equ. 2.26 (2.1).W = CL(o, = ogo]? (2.26) = gas mass velocity pounds/hr/sq ft constant found from the tabulation below where: Wl 2, = liquid density in pounds/cubic ft 94 = Gas density in pounds/cubic Ft Service Bay Spacing ze” | 2a" | 30” | Absorbers - oil 700 | 800 | 380 | Absorbers - glycol = | 500 | 550 Absorbers ~ amine ~ | 350 | 39s Fractionators and Stills | 440 | 540 | 600 Vertical separators - 600 (8 ft length min) Horizontal separators ~(1440)(L/20)9+56 (where L = £t between inlet and outlet nozzles) If the horizontal separator has a length greater than 20 ft. and the LID ratio is §:1 or less, a value of “C" of 1440 is suitable for calculations. Example Problem What would be the maximum allowable gas mass velocity in an amine absorber? Under conditions in the absorber the density of the gas is 3.5 1b/cu ft and the density of the amine solution is 63 Ib/cu ft. The absorber will have 24 inch tray spacing. Solution From the table we see that the value of the constant C for amine absorbers with 24 inch tray spacing is 350. W = 3501(63 - 3.5)]9°> = 350(208.3)°5 = 5050 1p/e? - ne Equ. 2.26 provides only for the necessary area required for gas (vapor) flow in the column. Column area required for liquid flow is normally provided in terms of “residence time" in the downcomer. Residence times will range from as low as 3 seconds for light hydrocarbon liquids with low foaming characteristics to as high as 10 or 11 seconds for badly foaming liquids such as ethanolamine. Final decision on the design value for liquid residence time in the downcomer is a matter for the experienced design engineer with good engineering judgment. The required diameter of a gas-liquid separator can be estimated from Equ. 2.26 or the velocity variation shown below. As noted in Volume 1, they are the same equation algebraically so should give the ‘same results provided consistent values of k"" and “C” are used. 440.5 y)/Py] (2.27) KElog ~ where: u = allowable gas velocity, actual cubic feet per second K = proportionality constant (normally 0.17-0.45) Unies tne separation takes place in a hydrate forming or freezing range, a mist extractor is desirable in most circumstances. With a foaming liquid (like the amines) this mist extractor should be not less than 12 inches above the liquid level in the separator or the top tray of a tower. ‘As in the case of fractionators and absorbers allowance must be made in the separator for liquid hold up. The liquid hold up time must be based on instantaneous rather than average rates of flow. If water-condensate separation is required, a maximum allowable settling velocity of 10 inches per minute normally is used (2.19). Depending on relative gas-liquid flow rates, this liquid settling velocity may represent the limiting factor on separator diameter. GAS P-V-T BEHAVIOR “There are several ways in which sour natural gases differ in behavior from sweet natural gases. One cof these is in terms of the PVT behavior of the gas. The simplest and most direct method for predicting the PVT behavior of natural gases is through the use of the compressibility factor. The compressibility tactor is defined through Equ. 2.28. Py = 2nRT (2.28) P = absolute pressure in psi Vv = volume in cu ft n= number of pound moles of gas R = universal gas constant = 10.73 ft-Ib_/1b mole ~ T z where: absolute temperature °R (OR = °F + 460) compressibility factor, dimensionless The compressibility factor is ordinarily plotted as a function of reduced temperature and reduced pressure according to the theory of corresponding states. A typical plot is shown in Fig. 2.13. The pseudo reduced temperature and pseudo reduced pressure are based on pseudo or molal average critical temperatures and pressures calculated according to Equ. 2.29 and 2.30. ae BP, (2.29) Tye = ERT, (2.30) 45Por = a T Tor The where: P,. = pseudo critical pressure psia X= mole fraction in gas mixture P = system pressure psia Tye = Pseudo critical temperature, °R T, = component critical temperature, °% T = operating system temperature, °R Pe = component critical pressure psia Pp = reduced pressure based on pseudo critical temperature, Pr dimensionless Toe = system reduced temperature, based on pseudo critical temperature dimensionless subscript = any component in a multicomponent system Duer the reduced temperature and reduced pressure have been determined according to Equ. 2.29 to 7.22. the compressibility factor for the gas mixture is read from Fig. 2.13. After the compressibility factor has been determined, the volume of gas can be calculated from Equ. 2.28, Example Problem A natural gas of the composition shown below is available at 100°P. and 1000 psia. What is the molar volume and the density of the gas under these conditions? ave Mol : _ Mol com fn _"e ce pe pe we C, 85 3U3.G 67.8 16.0 (567.6 F519 Tho C, 12 -550,1 707.8 30.1 0 3.6 Cy 03 666.1 616.3 44.1 1.3 18.5 _ 560 = 4900 _ Te 7 U7 pee eeeey niet From Fig. 2.13, Using Equ. 2.18 1000 x V = 0,853 x 10.73 x 560 v= 48.5 e 5.13 3.61 1b/£t3 46 5.13 £09/4b molFigure 2.13 Compressibility factors for natural gas Pseuto-reduced pressure, Pr (wat. 2.4) Freuosreaced Temperatures Tr Conpresstoflity Faetor Conoressbility Factor Pseudo-reduced pressure, lity factor works very well for sweet natural ‘The procedure outlined for determining the compress! ges mixtures. Unfortunately, it does not work well for sour natural gases. As the acid gas content of the natural gas increases, the deviation of the natural gas from the compressibility factor behavior shown in Fig, 2.13 increases. A number of efforts have been made to account for this deviation from sweet gas behavior. Perhaps the simplest, and among the most successful, |s that presented by Wichert and Aziz (2.3). They recommend the use of a pseudo eritical temperature adjustment factor which is a function of total acid gas composition and the HpS to CO2 ratio in the acid gas. A plot of the pseudo critical temperature adjustment factor as a function of COz and H2S composition is shown in Fig. 2.14. 47Figure 2.13b 014 for gests near ateaphare pressure Figure 2.13a ees. 2.0) Figure 2.14 Pseudo critical temperature adjustment factor, e, °R (Ref. 2.3 Courtesy Hydr. Proc.) 48BI Egg Ey 7" Stel ReBeRena| ae agaga| gta zeae sqnaaesr| agE23| sia ASSN) EER ES EEE) S88 90294) 5) suaacaaded 2 2 g 3 ee 8 | Sena aa oe ze PERSABae3 Hata aada| He 35s] NaS 28842 ths aa EB | URERARARR | ARARNHADE] SUOSECG | Sona va|FoAITCTTS 49The procedure recommended by Wichert and Aziz is as follows: 1. Calculate the pseudo critical temperature and pressure of the gas mixture by Equ. 2.29 and 2.30. 2, From the known concentrations of CO and HS in the mixture, determine the pseudo critical temperature adjustment factor from Fig. 2.14. 3. Calculate the adjusted pseudo critical temperature as: = The (2.33) where: Tj. = the adjusted pseudo critical temperature, R © = pseudo critical temperature adjustment factor as determined from gas composition and Fig. 2.14 4. Calculate the adjusted pseudo critical pressure according to: pros (2.34) Pe pe * BU - Ble adjusted pseudo critical pressure psia where: Ps. B mole fraction of HAS in the sour gas 5 Calculate the pseudo reduced pressure and the pseudo reduced temperature according to Eau. 2.31 and 2.32, but using the adjusted pseudo critical values calculated in steps 3 and 4. 8. Obtain the compressibility factor from Fig. 2.13 using the pseudo reduced Pressure and temperature values calculated in step 5. Wichert and Aziz tested their procedure over a wide range of temperatures, pressures and acid gas sour esitions. They found it to do an excellent job of reproducing experimental PVT measurements for Sour gas mixtures. Wichert and Aziz also fitted Fig, 2.14 in equation form. Based on limited evaluation by this writer, the eavation shown below gives excellent representation of the data plotted in Fig. 2.14 © = 120(ae9 ~ al-6) 4 1581/2 _ p4y where: A= mole fraction (CO, + H,S) 8 = mole fraction H,S' The procedure for correcting a sweet gas correlation like Fig. 2.13 for sour components is useful for anual calculations. in computer calculations use of other P-V-T equations of state may be more desirable, as outlined in Volume 3. 50Example Problem What would be the correction factors to be applied to the critical temperature and critical pressure of a natural gas that contains 15% €O, and 20% HS? Solution From Fig. 2.14, © = 31°F. ‘The correction factor for the pseudo critical pressure requires knowl- edge of both the pseudo critical temperature and the pseudo critical pressure before the calculation can be made. However, having deter~ mined ¢ from Fig. 2.14 makes the calculation by Equ. 2-33 and 2.34 simple. WATER CONTENT Prediction of the water content of sour gases is more difficult than for sweet gases. Both H2S and COz are very soluble in water and possess polar type molecular behavior (like water). ‘The most precise approach for predicting water content of sour systems utilizes one of the P-V-T equations of state, Both Volumes 2 and 3 discuss some of the proven alternative methods utilizing thi type of equation. ‘The other alternative is to use a sweet gas correlation and correct it for the sour components. ‘The equation for approximating water content is: Woe Wag + Yih + phy (2.35) where: Wj. = water content of gag in Ibs water per NMscf/D measured at 14.7 psia and 60°F. (Fig. 2.15) 1 effective water content contribution of H,S to the gas mixture (Fig. 2.16) effective water content contribution of co, to the gas mixture (Fig. 2.17) y = mole fr. of all components in gas except HyS and CO, ¥y = mole fr. of HS in gas Yq = mole fr. of CO, in gas 51Example: Calculate the water content of a natural gas at 100°. and 1000 psia containing 9.5% H,$ and 2.1% CO). From Fig. 2.15, 2.16, and 2.17, Wi, = 61, Wy = 150 and W, = 68 W = (0,884) (61) + (0.095) (150) + (0.021) (68) 69.6 lbs water /MMscf/D Both Fig. 2.16 and 2.17 are based on data for pure components, and mixtures containing up to 30% H2S andlor COg. The water content shown is an effective value to make Equ. 2.35 fit actual water content data. All data are for systems where water is the only liquid phase; there is no liquid HS or Op phase present. In Fig. 2.16 the solid lines are based on pure HS data, the dashed lines on mixture data. The solid lines end at the dew point conditions for pure HyS. Above 220° F. the dew point pressure is above 5000 Psia. Two lines are thus shown, a solid one reported for pure HS and a dashed line for expected mixture behavior. Below 300 psia the water content of pure saturated HzS and COp is equivalent to natural gas. As Pressure increases the water content of sour gases increases above sweet gas. From our experience, Equ. 2.36 will give the maximum value of the water content expected for the gas. The actual water content likely will be lower than that predicted, An alternative method for estimating the water content of sour natural gases has been presented (2.19). The method is based on charts which were calculated using the Soave-Redlich-Kwong equation of state with specially prepared and fitted binary interaction coetticients. The method recommended in Equ. 2.35 and the chart method recommended by Robinson et al. give essentially identical results below Pressures of about 1000 psia. At higher pressures the Robinson et al. chart method will give superior results, The charts prepared by Robinson et al. are shown in Fig. 2.18, 2.19 and 2.20. Remember: the water content shown in all the figures and equations is for a saturated gas at its water dewpoint. 521000 00 609 400 300 200 100 80 60 WATER CONTENT OF SWEET NATURAL CAS, LB PER Mscf (at 60°F and 14.7psia) 740 0 “0 80 120 160 200 240 WATER DEW PoINr, °F Figure 2.15 Water content of saturated natural gas {at its water dew point) (Ref. 2.1) 53g. vs & 8 : 3 7 Z ° 7 z 3 3 : é g 3 2 Water Gontent, lbs per MMscf (@ 14 230 op 1005 80°p 200 300 400 S00 1000-2000 3000 Pressure, psia Pressure, psia Figure 2.17 Effective water content of saturated carbon dioxide in natural gas mixtures Effective water content of saturated hydrogen sulfide in natural gas systems19, Beer ge = eee te eta cnas, N/M trees Matec, Ma ta cnr at er ps mrs eps capcinfy & BEER EE Figure 2.18 Water content of natural gas mixturesWer cote, Me Water cron, 4M aul ts aad an HEE oe ee Wear coe, Hc Figure 2.20 Water content of natural gas mixtures se ee Figure 2.19 Water content of natural gas mixtures210 an 212 219 218 238 2.16 2a7 218 219 220 REFERENCES ‘Campbell, John M., Gas Conditioning and Processing. Vol. 1 and 2 Lewis, John ., Petrateum Refiner, 40, March 1981, p. 169. Wienert, & and Aziz, K., Hydrocarbon Processing, 51, May 1972. p. 119. Engineering Data Book. Natural Gas Procossors Suppliers Association, Tulsa, Okla. th Ed., 1972. Gas Conditioning Fact Sook. Dow Chemical Co., Houston, Texas, 1989, Gas Treating Data Book, Jetterson Cnemical Co., Houston, Texas, 1969, ‘ud, Ey Applied Process Design for Chemicel and Pelrochemical Plants, Vol. 1, Wouston, Texas Gult Pushing Co. Campball, J. M., Ges Conditioning Processing, Vol. 2. Lugmla €.€, Applied Process Design or Chemie! and Petrochemical Pans, Hovston, Texas: Gulf Publishing Cov 198, P Frank, 0, Chemical Engineering, May 13, 1974, p. 126. "Ges Treating Chemicals,” Union Carbide. Rohsenow, W. M. and Hartnett, J __ Gas Digest, July 1977, p. 38. Shvartsburd, V, The Ol! and Gas Journal, Oct. 8, 1978, p. 99. ‘Handbook of Heat Transfer,” McGraw Hil, 1972. Hooper, Wi. B., Chemica! Engineering, Aug. 24, 1981, p. 96. ‘Tan, &.H,, The Oil and Gas Journal, Nov. 2, 1981, p. 140. Bell, K. J, The Oll and Gas Journal, Dec. 4, 1878, p. 59. Gerunda, A., Chemical Engineering, May 4, 1981, p. 81 Robinson, J. M., Moore, R. G., Heidemann, RA. and Wichert, E., Gas Conditioning Conlerence, Norman, Okla, March 05, Poke" fsisa “the OW and Gas voumal, Fed, 6, 1978, p. 78; and Energy ProcessingiCanade, Jan-Feb., 1973, p40) Churchill, 8. W., Chemical Engineering, Nov. 7, 1977, p. 913 AMINE PROCESSES ‘The alkanolamines are the most generally accepted and widely used of the many available solvents for removal of HaS and COp from natural gas streams. Credit for use of the alkanolamines must go to Bottoms (3.1) who was granted a 1930 patent covering their use for sweetening natural gas. Because of thelr reactivity and availability at low cost, notably monoethanolamine and diethanolamine, have achieved a position of prominence in the gas sweetening industry. ‘The alkanolamines to be considered in the natural gas sweetening inciude monoethanolamine, diethanolamine, triethanolamine, diglycolamine (88 hydroxiaminoethy! ether), diisopropanolamine, and methyldiethanolamine. All of these materials may be classed as “chemically reactive.” This is what decounts for their popularity for use in sweetening natural gas. There are differences in the chemical feactions that the various ethanoloamines undergo with cerlain acid gas constituents. These reaction similarities and differences will be covered in detail in Chapter 4, The prefixes “mono”, “di and “tri” refer to degree of radical substitution on the amino nitrogen. Monoethanolamine, for example, has the structure: Heed H (listers 4 Wwo—c—c—N DEE a HOW H ‘The OH group is the part of the structure that causes the compounds to be “olamines” rather than simple amines. The nitrogen is the so called “amino” nitrogen. Study of the structure shows that the first member of the amine homologous series would be ammonia (NH3). Ammonia can be used for acid gas removal from a natural gas stream but its volatility and other problems cause operating diffculties. The chemical structure of the remaining amines are shown in Figure 3.1. Physical properties for the ethanolamine are shown in Table 3.1. The ethanolamines are clear, colorless liquids that have a slightly pungent order. Ali except triethanolamine are considered to be chemically stable materials because they can be heated to their boiling point without decomposition, Triethanolamine decomposes below its normal boiling point of 680° F.FIGURE 3.1 CHEMICAL STRUCTURAL FORMULAS FOR ALKANOLANINES OF INTEREST IN GAS SWEETENING Diisopropanolamine 8 t Ho-¢~ a monn ~0- B-o-= T Bom Methyldiethanolamine Diglycolamine © 60B Hotline potne @ 760 nn fg." 338.9 eeering saint, *F 0.9 eosity @ 20°C, anise 1.018 speetfis gravity, 20%C/20° 017s syecttic nent @ 60°F, Bru/IA/* 0.608 8 68°F ‘meraat conductsetey & 68°F Sear Pee paitee ory Saat of vaportaation, BTU/T® 255 € 760 on He 3 : vianoete, oF 21 8 68" Aefcactive Indes, ny @ 66°F 2.4899 ro 0? me 1.085 9.09 0 60'F 1.0819(20"0/20%e) 0.400 250 8 oF fax 908 ot. sonusond siz TABLE 3.1 Gee HIF xocoacomponsey 9.07 8 oat 20 € 760 aay i @ 6st as 988 es solution} O78 Engineering Dita Book C3). 4. Ansoine constants for HEA and OFA from Do (3-3)s 4 sntotne conatonte for TEA, 004, DIPA snl MOEA ftetad by 5. caloulated ty wethed of Udeesen (3-2). Woh soe ose @ 60°F 82 @ 60 20st oan 219.16 €760 ants soe oor 1 NORA Cexceat for Antotce consten:s) fon Kobt and Aiosentetd ee) i o.999 @ note 0.989 @ astcr20°¢ 0.69 #20 185.0 (me Bp nase @ 147 9.9698 Soo. ae souss one BeeThe Antoine constants for estimating the vapor pressure for the various ethanolamines are shown in Table 3.1, The Antoine equation is: B log P = A-F pe (3.1) where: P = vapor pressure, mmHg T = temperature, °C. A,B, and ¢ = constants, obtained from Table 3.1 ‘The vapor pressures of ethanolamine-water mixtures can be satisfactorily estimated using ideal solution behavior, ethanolamine vapor pressures from Table 3.1 and the vapor pressure of water. The Antoine equation constants for water are: A 7.96681 B = 168.210 C= 228.000 Table 3.2 shows some of the key properties of the alkanolamines. The vapor pressure values, which are at typical contactor conditions, show that vaporization losses of monoethanolamine (MEA) will be much greater than for any of the other ethanolamines. Also on the basis of relative acid gas capacity MEA shows two distinct advantages. However, as will be shown in Chapter 4, there are many other considerations than must be taken into account when selecting a sweetening agent. Monoethanolamine (MEA) MEA is the strongest base of the different amines and so reacts most rapidly with the acid gases. MEA will remove both hydrogen sulfide and carbon dioxide and generally is considered to be non- selective between these (wo acid gases, With the lowest molecular weight of the common amines, it has the largest carrying capacity for acid gases on a unit weight or volume basis. This generally means less solution circulation to remove a given amount of acid gases. in addition, MEA is chemically stable, which minimizes solution degradation; it can be separated easily from the acid gas constituents by steam stripping. MEA reacts irreversibly with carbonyl sulfide and carbon disulfide. This results in solution loss and In the buildup of reacted solids in the MEA solution. MEA has a higher vapor pressure than the other amines. This can result in significant solution losses through vaporization. The problem usually can be overcome by a simple water wash of the sweetened gas stream. The rate of reaction of COp in MEA is slower than that of H2S. The process is not considered selective, however, because carbon dioxide is readily absorbed and essentlally will be removed completely when treating natural gas for H2S removal to pipeline specifications. MEA will easily reduce acid gas concentrations to pipeline specifications (generally less than 0,25 grains per 100 cu fl). By proper design and operation, the acid gas content can be reduced as low as 0.05 ‘rains per 100 cu ft. 62TABLE 3.2 SOME CHARACTERISTICS OF ETHANOLAMINES MEA _DEA. ‘TRA DCA. DIPA MDEA Mol. Wee 61.08 105.14 148.19 105.14 133.19 119.17 ve @ 100°F om Fg. 1.05 0.058 0.0063 0.160 0.010 0.0061 Rel. Capacity, % 100 58 aL 58 46 5h Diethanolamine (DEA) ‘The aqueous DEA process is similar in principle and operation to the MEA process. One primary difference is the reaction of DEA with carbonyl sulfide and carbon disulfide. The reactions of DEA with carbonyl sulfide and carbon disulfide are slower than those of MEA, and result in different products. Consequently there is minimum DEA loss caused by reactions with these sulfur compounds. As a result, It Is especially favored for sweetening refinery and manufactured gas streams where these contaminants are prevalent. DEA is non-selective and will remove both HS and COp, Difficulty is sometimes encountered with the DEA process in reducing hydrogen sulfide concentrations to pipeline speciications. Wendt and Dailey (3.2), however, describe the SNPA modification of the DEA process and report its capability for removing HS to the level of approximately 0.1 grains per 100 std cu ft, DEA is much less volatile than MEA, resulting in significantly lower losses of amine solution by vaporization. After regeneration, DEA solutions typically have much lower concentrations of residual acid gas constituents than do MEA solutions. As will be seen in Chapter 4, this stems from the basic equilbrium relationships between diethanolamine and the acid gases. It does provide a basis for the advantages DEA demonstrate over MEA solution. Triethanolamine (TEA) Although aqueous triethanolamine (TEA) was the first commercially applied amine sweetening process, it has been largely displaced by elther MEA, DEA or DGA. The tertiary amines appear to be less reactive with HS and COp. Because of traditional problems in treating natural gas to pipe line specification very little TEA is used commercially for natural gas sweetening. Diglycolamine (0GA) 88" Hydroxietholamine [22-amino-ethoxy}ethanol] is a more recent addition to the list of ethanolamines used for natural gas sweetening. The process is proprietary and the material has been trade named “Diglycolamine.” It is a primary amine and has all of the potential advantages of high reactivity, low equilibrium partial pressures, etc., that are characteristic of the primary ethanolamines. Table 3.3 from the paper by Huval and Van De Venne (3.6) shows that DGA can satisfactorily meet pipe line HS specifications at rather low pressures. Even though DGA Is primary ethanolamine it maintains the same advantage of low residual acid gas concentrations in the regenerated solution as does DEA.TABLE 3.3 ARAMCO DGA PLANTS FEED PRODUCT RAS TANURA 130 MescFD 15-20 PSIG 125°F 913% HAS 400-500 PPM HS 1.5-2.5% CO, 1000-2500 PPM co, UDHAILIYAH 75 MMSCED 220 PSIG 125°F 1-28 HS <2 PPM HAS 10-112 co, < 150 PPM CO, MASTER GAS SYSTEM > 4000 “ascKD (BERRI, SHEDGUM, 115-170 PSIG UTHMANTYAH) 120-140°F RESIDUE GAS ‘ETHANE 8-37 8S 1-2 PPM HS < 2 PPM HAS 8-147 Co, < 100 PPM CO, 20-80 PPM CO, Diisopropanolamine (DIPA) Diisopropanolamine is the reactive ingredient in the suifinol process patented by Shell. It is also used in the ADIP process which is designed for removal of carbonyl sulfide. Diisopropanolamine will treat to pipeline specifications for sweet gas. Because data for the sulfinol process are proprietary to Shell its use Is normally restricted to plants where the process/design calculations have been carried out by Sheil engineers. Methyldiethanolamine (MDEA) MDEEA is another newcomer to the group of ethanolamines used for natural gas sweetening. It has received a great deal of attention recently because of its capability for “selective” reaction with HS in the presence of COp. It was first mentioned as a sweetening agent in the early 50's (3.7, 3.8) but has had recent attention because of potential energy saving. According to Pearce and Brownlie (3.9) MDEA Is. capable of treating new gas to sweet gas specifications for H2S. 64Solution Concentrations The primary differences in process schemes for MEA and DEA sweetening are in solution concentrations. MEA is ordinarily used in a 10 to 20 per cent by weight aqueous solution. By far the most ‘common concentration would be approximately 15 per cent by weight MEA. DEA is also used in the 10 to 20 per cent by weight concentration. The SNPA (Societe Nationale des Petroles d’Aquitaine) modification of the DEA process uses a higher concentration of DEA. The process unit flow scheme is conventional, but the DEA solution concentration will range upward to 30 or more weight per cent DEA. This higher concentration of DEA (because of the higher molecular weight of DEA) gives about the same molar ratio of amine and water as does a 15 to 20 weight per cent MEA solution. DIPA, DGA, and MDEA are also used in higher concentrations. Typical concentration ranges for DIPA and MDEA are 30% to 50% by weight in the aqueous solution. DGA concentrations range from approximately 40% to 70% by weight. Use of amines in aqueous solution, of course, saturates the sweet gas with water vapor, regardless ‘of whether the entering sour gas is wet or dry. For most amine processes, this means that a dehydration step necessarily follows sweetening. One process which overcomes this shortcoming Is the use of MEA or DEA in combination with ethylene or diethylene glycol. Combination of amine and glycol witl usually do an excellent job in removing acid gas constituents, but generally does not dehydrate as well as a conventional glycol installation. Severe corrosion problems are frequently encountered when using the glycolamine combination so that this process is not currently in widespread use. ‘The general process flow for an amine sweetening plant is shown in Fig. 3.2 (3.10). The process flow scheme varies little, regardless of the aqueous amine solution used as the sweetening agent. The primary pieces of equipment of concern are the absorber column and stripper column, together with the associated piping, heat exchange, and separation equipment. The sour gas containing H2S andior CO» should always enter the plant through a scrubber to remove any free liquids andlor entrained solids. The sour gas then enters the bottom of the contactor and flows upward through the column in intimate countercurrent contact with the aqueous amine solution. ‘Sweetened gas leaves the top of the contactor and flows to a dehydration unit before being considered ready for sale. Lean amine solution from the bottom of the stripper column still is pumped through an amine-amine heat exchanger and then through a water or air-cooled exchanger before being introduced to the top tray of the contactor. The amine moves downward through the contactor counter-current to the sour gas, and removes acid gas constituents from the gas stream. Rich amine solution flows trom the bottom of the contactor through the amine-amine heat exchanger and then to the top of the stripper column. The amine-amine heat exchanger serves as a heat conservation device and lowers total heat requirements for the process. A part of the acid gases will be flashed from the heated rich solution on the top tray of the stripper. The remainder of the rich solution flows downward through the stripper in countercurrent contact with vapor generated in the reboiler. The reboiler vapor (primarily steam) strips the acid gases from the rich solution. The acid gases and the steam leave the top of the stripper and pass overhead through a condenser, where the major portion of the steam is condensed and cooled. The acid gases are separated in a separator and sent to the flare or to processing. The condensed steam is returned to the top of the stripper as reflux Rich amine solution leaves the bottom of the contactor at an elevated temperature due to the heat of reaction released when acid gases react with the amine. For this reason heat exchange on the lean amine solution in the amine-amine exchanger does not cool it sufficiently for many processes. The amine cooler serves to lower the lean amine temperature to the 100° F. range. Higher temperatures for the lean amine solution will result in excessive amine losses through vaporization and also lower acid gas carrying capacity in the solution because of temperature effects. 65ered ce A = {ee ) ourver_ senveeen 99 Do ELASH_ tan q cannon sure Figure 3.2 Process flow for typical ethanolamine sweetening unitA process modification attributed to Shoeld (3.11) and described by Kohl and Riesenteld (3.12) Is shown in Fig. 3.3 This modification uses two amine solutions fed at different points to the contactor — @ ‘semi-lean solution introduced at a mid-way point and a lean solution introduced in the conventional manner on the top tray. This process is particularly advantageous for treating sour gases with ahigh acid gas content — above approximately 30 per cent. The liquid withdrawn from the mid-point of the stripper is not completely stripped and is circulated to the middle part of the contactor, where the acid gas ‘concentration is highest. The lean solution introduced at the top contacts the gas with the lowest acid gas concentration and reduces it to the specified level. —> ACID GAS GAS LEAN RICH SOLN. SOLN. SOUR GAS ‘STEAM PLIT STREAM AMINE PROCESS Figure 3.3 Flow sheet of split-stream amine process This process moditication offers the possibility of significant savings in steam costs, particularly for high acid gas content streams. Ithas the disadvantage that the stripper column needs to contain more trays and both the contactor and the stripper are more complex. The two lean amine streams require separate pumps and heat exchangers together with a more complicated control system. However, in many cases the overall plant investment will be fess than would a conventional process like Figure 3.2 utilizing all of the amines with complete regeneration.Figure 3.4 shows an MEA flow sheet with typical temperatures and solution concentrations. The MEA solution used is 20% by weight. It enters the contactor at 90° F and leaves at 136° F. The sour gas being treated contains slightly more than 11% HgS. By heat exchange the rich MEA is heated to 200° F before being introduced to the regenerator. The lean amine leaving the regenerator is at 240° F and by amine heat exchange combined with water cooling it is cooled to 90° F for introduction to the contactor, Though the temperatures shown in Figure 3.4 are for MEA they would change littie for DEA because the vapor pressure of water is the controlling parameter in determining regenerator reboiler temperatures. 7 fz Figure 3.4 MEA process flow sheet showing typical temperatures The flow sheet for diglycolamine, though basically similar to the MEA and DEA process flow sheet, differs considerably in some details. A typical DGA flow sheet is shown in Figure 3.5 (3.6). Because of the higher concentration of DGA a flash drum on the rich DGA stream leaving the bottom of the contactor is generally considered to be essential in order to prevent excessive hydrocarbons in the acid gas going to the sulfur plant. In the DGA system this flash drum normally contains a small contacting section where a portion of the DGA circulation is used to retain HpS/COz in the amine. Regenerator temperatures tend to run considerably higher being in the range of 260° to 280° F as opposed to the 230° to 240° typical of MEA and DEA. In addition the reclaimer runs almost 100° hotter because of the higher boiling temperature of DGA. High solution loadings lead to a high discharge temperature of the rich solution from the bottom of the contactor. The process shown in Figure 3.6 uses a side cooler on the contactor. This side cooler will typically lower bottom tray temperatures by 10° to 15° F. Regardless of the amine used a contactor intercooler should be considered in high loading ethanolamine units. Fig. 3.6 is a convenient correlation for converting the weight percent of amine in water solution to the equivalent normality. Many solubility test results are reported in terms of the normality of the amine solution. Diglycolamine and diethanolamine fall on the same line because they have essentially the same molecular weights. SUMMARY OF AMINE PROCESSES There actually are five different types of basic amine sweetening processes to be considered for use: — The aqueous monoethanolamine process, which has been widely used for sweetening of natural 928 streams. 68 eee eee ae eeACIO GAS TO CLAUS PLANT SWEET CAS. TO CONPRESSION WAKE-UP waree Pe. | | | Fate” eey [== ‘WASH BACK. PUKP. CONTACTOR RICH AMINE, SIDE COOLER FEED GAS. ia cneuaion @® CORROSION MONITORING POINTS Figure 3.5 DGA process flow shezt showing typical temperaturesFigure 3.6 _| ‘Normality of aqueous amine solutions (Courtesy Jefferson Chemical Co.) auine , werent pencent obese serves ones wis NORMALITY — The aqueous diethanolamine process, which traditionally has been used primarily for refinery or ‘manufactured gas sweetening, but is now finding applications in natural gas sweetening. — Glycotamine processes which use either monosthanolamine or diethanolamine in combination with a glycol to simultaneously sweeten and dehydrate the gas stream. — Aqueous triethanolamine systems, which in general have been displaced by mono or diethanolamine systems. — The methyldiethanotamine system (a tertiary amine like triethanolamine) which, in reality, is not commercially competitive with the other alkanolamine processes. Of the systems listed above, MEA and DEA have found the most general application in the ‘sweetening of natural gas streams. GENERAL OPERATING PROBLEMS. Amine treating processes tend to be troubled with the same operating difficulties regardless of the type amine used. These operating problems can be categorized broadly as corrosion, solution degradation and foaming. These are not necessarily independent problems. The same basic condition may cause difficulty in all three areas. For this reason separating them for discussion is difficult, and is only done as a matter of convenience. 70Corrosion Looking at characteristics of the ethanolamine solutions gives the Impression that corrosion would not be a problem. Asa matter of fact many amines are used as corrosion inhibitors. The amines in water solution are basic as shown in Fig. 3.7 for MEA (3.5), Fig. 3.6 for DEA (3.6), and Fig. 3.9 for DGA (3.14). Extrapolation of the lines in Fig. 3.7 and 3.8 would indicate a basic solution for both MEA and DEA even at conditions in the regenerator reboiler. However, as shown in Fig. 3.10 (3.15), the situation changes drastically when the acid gases enter the picture. Though the individual effects of HzS and CO2 appear to differ, the common ingredient is a sharp lowering of the pH with the possibility at high loading that an acidic solution can be reached at reboiler temperatures. Polderman et al. (3.16) report tests showing that a 20 per cent MEA solution in the absence of acid gases is less corrosive toward steel than is water. They also report that it can be converted (through its carbonate salt) to corrosive products at temperatures expected to be encountered in gas treating plants. As pointed out by Polderman et al. and also by Ballard (3.17) the presence of acid gas constituents causes corrosion directly. This type of corrosion is all but impossible to eliminate as removal of acid gases is the name of the game. It can be minimized or controlled, or at least limited, by proper design and operation. Corrosion will be most severe at places where the highest concentrations of acid gases encounter the highest temperatures. These points will include the amine-amine heat exchanger, the stripping column and the reboiler. Stress corrosion is prevalent in amine systems. This generally is associated with residual stresses which result from localized heating during vessel construction such as welds in absorbers, strippers and piping. Stress relieving all major equipment and piping will help to alleviate stress corrosion. Solution Degradation Amine solutions will slowly oxidize when exposed to air or oxygen. The products of these oxidation reactions are generally considered to cause corrosion problems. The oxidation can be minimized by use ‘of an inert gas blanket on amine storage containers and surge drums. MEA of course reacts with carbonyl sulfide (COS) to form an insoluble salt. This will result ina mol- for-mol ioss of MEA unless the solution is reclaimed. The reaction can be reversed by placing the MEA in a basic solution and heating. DGA also reacts with COS and the reaction can be reversed by heating. ‘There is a large body of literature available on degradation of ethanolamine solutions. While some of the degradation products do have structures that will remove H2S andlor CO they generally are not so effective as the original ethanolamine and, for this reason, represent a loss in solution activity. MEA and CO can react under reboiler conditions to form an oxazolidone which in turn can react to form N-(2-Hydroxyethy!)-ethylenediamine which is more basic than MEA but whose sulfide salts are very difficult to regenerate. As reported by Polderman and Steele (3.18) DEA forms N,N'-bis(2-Hydroxyethyl) piperazine (HEP). Hakka et al, (3.19) showed that degraded DEA contains N,N’,N’tris(2-Hydroxyethyl)- ethylenediamine] (THEED) or 3,6-diaza-1, 8-dihydroxy-3-(2-Hydroxyethyl)-octane. Both HEP and THEED are basic and will react with HS and/or CO but they are not so reactive as DEA and result ina loss of solution activity,PERCENT BY WEIGHT TECHNICAL GRADE MONOETHANOLAMINE 8 50 60 70 80 90 100 110 120 130 140 150 TEMPERATURE, “F. Figure 3.7 pH of MEA solutions at various temperatures 72PERCENT BY WEIGHT TECHNICAL GRADE DIETHANOLAMINE 12.0) 16 4 2 11.0) pH 10.8 10.6, 10.4 10.2! 10.0) 9.8) 9.6! 70 80 90 100 HO 120 130 140 150 TEMPERATURE, °F. Figure 3.8 pH of DEA solutions at various temperaturesDIGLYCOLAMINE® , WEIGHT PERCENT Figure 3.9 pH of DGA solutions at various temperatures 74Figure 3.10 Effect of acid gas loading on pH of MEA solutions Foaming Foaming in an amine sweetening process can result in a number of different problems. Plant gas through-put may be severely reduced; sweetening efficiency may be decreased to the point that pipeline specifications cannot be met; amine losses may be significantly increased. Some of the materials that can cause foam problems in amine units include: — suspended solids — condensed hydrocarbons — amine degradation products = almost any foreign material such as corrosion inhibitors, valve grease, or even impurities in makeup water ‘The use of an amine reclaimer in combination with proper solution filtration will help to solve many foaming problems. Amine Reclai ing The amine reclaimer is usually a batch distillation system, which Is operated semi-continuously. Its purpose is to distill the water and amine from the fouled solution, leaving behind the entrained solids; dissolved salts and degradation products that cause foaming and corrosion problems. Except in unusual cases the reciaimer will operate on 3 per cent or less of the solution circulation rate. Ballard (op. cit.) presents an excellent detailed discussion of the proper operation of an amine reclaimer. He lists the following specific operating suggestions for a reclaimer: “1, An easy-to-open manway should be located on the reclaimer shell so the solids can be simply washed out at the end of the cycle. The drain line should be large enough to pass solids, 752. The tube bundles should be raised 6 inches or more from the bottom of the reciaimer shell to provide space for sludge accumulation below the tubes, and to give better solution flow around the tubes. 3. A packed column should be placed on top of the reciaimer to eliminate foam and entrainment trom the overhead vapor stream. The installation of a glass site port in the vapor line will helo keep a check on this carry over. An analytical evaluation will help pinpoint suspected entrainment. 4. The tubes should be widely spaced for easy cleaning. 5. Make certain the steam supply is not superheated. 6. Use a chart recorder to monitor the reclaimer temperature throughout the cycle. Inspect the recorder routinely to make sure it is indicating true temperature readings, 7. Control the amine feed to the reclaimer by a level controller on the kettle to maintain a liquid level at least 6 inches above the tube. install a rotameter in the feed line to indicate when the feed falls off and steam stripping should be started. 8. Allow sufficient vapor space above the liquid layer in the reclaimer to prevent liquid carryover in the overhead vapor line. 9. A chemical analysis of the reciaimer bottoms and the amine solution identifies the solution Contaminants and determines the degree and rate of solution contamination.” The reclaimer is an integral part of a successful amine sweetening process. It normally operates on a sidestream of the lean amine solution leaving the bottom of the stripping column. Reclaimers are not ordinarily used on DEA systems. The DEA and its degradation products tend to boil at very nearly the same temperature. For this reason they are difficult to Separate in a simple flash distillation in the way that most reclaimers operate. In addition, the boiling point of DEA is higher than normally available process steam temperatures. Filtration Proper solution filtration is extremely important. Some solids, particularly iron sulfide, are very difficult to remove in separators and surge vessels. Any entrained iron sulfide particles entering with the sour gas will be picked up by the amine solution in the absorber. Recommendations that filtration remove particles in the one micron range are frequent. Attempting this with a full stream filter is difficult, ‘and expensive. As a result, a combination of filters is frequently used. A full stream filter on the outlet, from the absorber will remove particles down to the 10-25 micron range. A sidestream activated charcoal filter will then be used on 10 or 20 per cent of the amine solution circulation to remove smaller particles. In addition, the activated charcoal will remove the hydrocarbons, some of the degradation products, and feaction products to improve amine solution quality. Scheirman (3.20) has given an excellent review of solution filtration practices in DEA plants. ‘According’ to Hakka et al. (3.19) the primary decomposition products in a DEA plant are N,N’ - bis, (hydroxy ethyl) piperazine (HEP) and N,N,N’ tris (2-hydroxy ethyl) ethylenediamine (THEED). Scheirman Indicates that the addition of soda ash or caustic soda can aid in regeneration of some of the compounds ‘by adjusting the pH of the solution. He recommends a cartridgetype filter followed by a fixed bed carbon 76filter for DEA service. The fixed bed carbon filter will have a design flow rate of 2 gpm per sq ft of vessel cross section to 15 gpm per sq ft. The recommended minimum bed depth is 10 ft. If a precoat filter is Used, either diatomaceous earth or a mixture of powdered asbestos and powdered carbon may be used as a filter, However, precoat filters can provide a nuisance factor in operations. * Note: gpm as used herein is an abbreviation for U.S. gallons per minute. Perry (3.21) has given a very valuable discussion of the benefits of activated carbon filtration for maintaining clear, clean amine solution. He discusses graded bed carbon filters as shown in Fig. 3.11 and deep bed filters as shown in Fig. 3.12. The graded bed filter has a minimum life and requires frequent backwashing and steam regeneration. In deep bed filtration there is a much greater depth of activated carbon giving longer bed life between regenerations. SOLUTION > —>¢ > mureare 2A gyn ft" FUTRATION occURS 18 Tom. 2K T OF BeO 10X30 Mes ACT WATED CARBON ta" 0 supPORT meCIA © Yate He" fe Figure 3.11 Graded bed activated carbon filter 7SOLUTION y Figure 3.12 Deep bed activated carbon filter X 4xlo MESH 16-19 feu.Fe ACTIVATED CARBON suprorT MEDIA. Foam Inhibitors In some cases, reclaiming and filtration of the amine solution will not cure the foaming problem. In these cases, use of a foam inhibitor may be necessary. If so, an inhibitor should never be used in the Plant without first being tested on a laboratory scale; one that works in one situation may not in another. Important to remember, also, is that the foam inhibitor does not solve the basic problem. It merely Provides a means for controlling foam. Search for the cause of the foaming should be continued until the Cause is located and eliminated. If a foam inhibitor is used, it is necessary to control its concentration very carefully. Too much foam. inhibitor can promote foam. 8Corrosion Inhibitors Corrosion inhibition systems have been developed for ethanolamine units, particularly those for MEA where corrosion appears to be most severe. Kelley (3.22) and Butwell et al. (3.23) have described the UCAR amine guard system developed by Union Carbide Corporation. It is intended for systems removing COg only. Solutions in concentrations as high as 50% by weight and acid gas loadings as high as 0.5 moles per mole have been used successfully without excessive corrosion. Dow Chemical Company also ‘fas a CO2 corrosion system. Dow also has a corrosion inhibitor that can be used on systems with both H8 and C02, GENERAL CONSIDERATIONS Over the years of operation of amine sweetening units a substantial amount of information has been collected and reported for improvernent of operation and design (3.6, 3.14, 3.18). These studies have led to guidelines which, even though empirical, if followed and properly implemented in the design stages of a sweetening unit, will serve to greatly reduce, if not eliminate, most operating problems, Inlet Scrubbing ‘The Importance of proper and efficient inlet scrubbing of the sour gas is difficult to over-emphasize, Many problems in amine plant operations such as foaming, corrosion and reboiler tube burn-out can be traced to the presence of excessive amounts of foreign material in the amine solution. Liquid hydrocarbons and entrained solids frequently can enter the plant with the sour gas stream. In addition, such foreign materials as corrosion inhibitors, drilling muds and well acidizers can also enter with the sour gas stream, particularly on lease-type sweetening units, The inlet separation equipment should be sized and designed with these factors in mind. The fact that these materials do not enter at a steady rate but slug or surge into the plant should also be kept in mind. Instantaneous flow rates may be extremely high and the inlet separating system must be designed to handle that. Most problems in the contactor section of the plant can be traced to foaming or entrainment. The most frequent sources of these troubles are entrained solids or entrained hydrocarbons, Amine Losses ‘Amine losses can be very expensive. For this reason a separator on the sweet gas stream leaving the contactor is usually advisable, A reduction of MEA losses by only 0.1 pounds of MEA per MMSCF will 237 mm Hg, 760 The partial pressure of CO, is 0.02 x 915 x 12% = 948 mm He The ratio of acid gas partial pressures is Be = 0.25 Before the equilibrium composition of the MEA leaving the bottom of the contactor can be determined the temperature of the rich amine stream must be known, As a first assumption, assume that the rich amine leaves the contactor at 140°F, (60°C.). ‘The equilibrium composition of the HS in the amine can then be read from the known value of Ry (0.25) and the HyS partial pressure of 237 millimeters of mercury. From Fig. 4.7, read the equilibrium concentration of H)$ to be 0.096 moles of 1,8 per mole of MEA. From Fig. 4.11, using the 237 @illineters of HS partial pressure in the sour gas and the equilibrium concentration of 0.096 moles of Hj per nole of MEA, read an Ry value of 0.17, mols H,S/mole MEA 4," wos 00;7mole WEA = 0.565 Mols C0,/nole MEA = 2 ‘The equilibrium acid gas concentration of the MEA solution at the estimated temperature leaving the absorber of 140°F. will be: 0.096 mol Hs 0.565 mol co, 102Using @ 70% approach to the equilibrium concentration: 0.096 x 0.7 = 0.0672 mol H)S/mol MEA 0.565 x 0.7 = 0.3955 mol CO, /mol MEA This is the concentration of the H,S and CO in the rich MEA solution leaving the bottom of the contacto. The net pickup in the contactor will be the difference between the concentration leaving the bottom of the contactor and the unstripped acid gas remaining in the lean amine solution as it enters the top of the contactor. 0.0672 - 0.0025 0.0647 mol H,$ pickup/mol MEA 0.3955 = 0.1275 0.2680 mol C0, pickup/nol MEA ‘The total amount of acid gas to be removed can be determined from the total gas flow rate and the percentage of acid gas in the sour gas stream. 50 mnsefd = 34,700 scfm Since the gas contains 2% of CO,, there will be 694 scfm 0.02 x 34,700 = 694 scfm CO, Feo cet mol 1.826 mol CO,/min. 0.5% H,S = 174 scfm H,S = 0.458 mol H,S/minute For the sake of simplicity in material balance calculations, assune that all acid gas carried in with the sour gas stream will be removed by the amine solution. This is not quite correct because there will be 0.25 grains of HS remaining per 100 scf of gas. However, this is only approx imately 4.0 x 10-6 mol-fraction. There will be about the same amount or slightly more CO, but for material balance purposes consider these con- centrations negligible. 0.458 BGee7 ~ 7-079 mol MEA/min needed for HS 1.826 ‘i GBeip > 6-813 mol MEA/min for CO, 7.079 Abide = 347 gpm required circulation rate of MEA solution ‘The design circulation rate of 15% MEA solution would be 350 gpm.The calculated acid gas pickup is 0.0323 moles acid gas per mole of MEA, and very close to the maximum recommended. A rule-of-thumb for acid gas pickup is not more than’0.35 moles pickup per mole of MEA, or, on another basis, no more than 2.5 standard cu ft of total acid gas pickup per gallon of MEA solution. The reader should not infer that an acid gas concentration in the sour gas of 2.5% will always give proper amine solution loading. The ratio of H2S to COp in the acid gas is important. As an illustration, in the case at hand, with an HS partlal pressure of 237 mm of mercury, the equilibrium loading of HoS in the rich amine solution would be approximately 0.72 moles of HS per mole of MEA (from Fig. 4.7). This shows very clearly that Cg in the gas has an effect on HoS. The reverse is true also. The effect of one acid gas constituent on the other is not simple, but can be predicted through the use of the Jones data as plotted by McCoy and Maddox in Fig. 4.6 to 4.13. The next step would be to check the temperature of the rich amine solution leaving the bottom of the contactor. This can be done by making an over~ all heat balance on the absorber to determine the rich amine temperature. There are 694 cu ft/min 380-cu fe/mol 7 }+826 mol CO, /min 1,826 mol CO,/min x 44 Ib/mol = 80.3 1b CO, /min Kohl and Riesenfeld(4.7) give a heat of reaction between CO9 and MEA of 825 Btu per pound, and for HzS and MEA a heat of reaction of 820 Btu per pound. 80.3 1b CO, /min x 825 Beu/Ib = 6.63 x 10° Btu/min Likewise, there are 15.6 Ib of H,S reacting per minute. 15.6 lb H)S/min x 820 Btu/Ib = 1.28 x 10% Beu/ain The total heat release from the reaction of the sour gas and the amine is (6.63 + 1.28) x 10° = 7.91 x 104 Beu/min The heat pickup by the sweet gas can be determined using the enthalpy charts in the NGPSA Engineering Data Book(4.10). Taking the sweet gas to be pure methane (molecular weight = 16) and estimating the temperature of the sweet gas leaving the contactor to be the same as the entering lean amine (based on Fig. 4.16), the heat pickup by the sweet gas in passing through the contactor will be enthalpy at 110°F, 258 Bu/1b enthalpy at 90°F, 247 Beu/1b 11 Beu/1b pickup 108The mass of sweet gas per minute will be (C4,700 G0 ¥ +975) x 16 = 1,425 1b/min The total heat pickup by the gas will be 1.425 x 107 x 11 = 1.56 x 10° Beu/min ‘The temperature rise of the amine solution will be (7.91 x 10% - 1.56 x 104) Bee min °, 23.9°R ib. gat Bey 7 (8-3 Faq) (350 aya) (0-915 Thee This gives a strong amine solution temperature of 134°P. This is close enough to the assumed value of 140°F, that there is no need to repeat the lengthy foregoing calculations. Contact Performance Fixing the number of trays, either theoretical or actual, in the absorber is not a simple task. As a matter of fact, many authors feel that computation of the number of theoretical trays required is an exercise in futility (4.14). The calculation of rates and efficiencies in a simple absorption process is complicated enough. In the case of absorption followed by chemical reaction, the calculations become exceedingly difficult. Limited amounts of experimental work have been reported on the dynamic absorption (reaction) of pure HyS and COp in water solutions of ethanolamines. However, these data are insufficient to provide a ‘sound basis for rational plant/process design. Essentially no data are available on the simultaneous dynamic absorption/reaction of HS and Cp into any of the ethanolamines in water solution. DIETHANOLAMINE Solubility of CO2 and H2S In DEA Lee et al. (4.11, 4.12) have taken mutual solubility data for COp and HyS in DEA solutions. The solution strengths were 3.5 normal and 2.0 normal and the temperatures were 50 and 100° C. The 2.0 normal solution is 20% DEA by weight and the 3.5 normal solution is 35% DEA by weight. The temperatures correspond approximately to temperatures at the bottom of the contactor and the bottom of the regenerator. The Lee et al. data are shown in Fig. 4.17 to 4.24, For temperatures other than those for which data are taken, Lee et al. recommend interpolating by making use of the fact that log vapor pressure theoretically Is a straight line function of reciprocal absolute temperature. To illustrate this point, assume that the equilibrium loading of a 2.0N DEA solution is 0.6 moles of GO2 per mole of DEA at 70° C. No HaS is present Lee et al. also report heat of solution values for COz and H2S in DEA. The values for CO2 are shown in Table 4.6. The heat of solution of CO in. DEA is dependent on both the strength of the DEA solution and the amount of the DEA that has reacted with the Oz. The heats of solution for HS in DEA are shown in Table 4.7. The HS heat values are dependent only on the amount of DEA that has reacted with HpS and are largely independent of the DEA solution strength. 105MOLES. H)S/MOLE DEA IN LiQuIO [iz iijolosiss or los 16s] bal loa, 900 02 0a a6 oa 76 12 a MOLE RATIO IN LIQUID, CO /DEA Figure 4.17 Mutual solubility data for HyS and Cp in 2.0N DEA at 50°. (Fig. 4.17-4.24 are from Ref. 4.11 and 4.12, Courtesy F.0. Otto) 6 NOLES’ COa/MOLE DEA IN LIQUID) x 1p-99/98-o7}oolfos-tbatosttoa+ tor 5 of PARTIAL PRESSURE H.S, ps eho 02 O4 O68 aa 10) 12 ia MOLE RATIO IN LIQUID, HzS/DEA Figure 4.18 Mutual solubility data for HS and COp in 2.0N DEA at 60°C 106Moles Hps/MOLE DEAN LiguID || 12390908 07 oe 05|04 [a3 ot |e PARTIAL PRESSURE COg ,psia e oo 02 4 06 08 10 12 14 MOLE RATIO IN LIQUID, COz/DEA Figure 4.19 Mutual solubility data for H2S and CO2 in 2.0N DEA at 100°C. 1000 pee oc} e Jey sO] gs parr aera moe pss Be61 HOLES, CO,/MOLE DEAN LidUiB Soh PARTIAL PRESSURE HS, psia 00 08 64 co 38 16 13 16 MOLE RATIO IN LIQUID, Hz S/DEA Figure 4.20 Mutual solubility data for HgS and COp in 2.0N DEA at 100°C. 107MOLES. H2S/MOLE OFA IN LIQUID. piped oe o7ieos ols [Oana 38 oct o EEE CECE EEEEH 351 DEA SOLUTION aT 50°C o 4 PARTIAL PRESSURE CO2 , psia 003) 0055 02 08 36 08 13) Te v4 MOLE RATIO IN LIQUID, CO2/DEA Figure 4.21 Mutual solubility of H2S and COz in 3.5N DEA at 50°C. 1000, 2 [coe/ vou! iN Liou % ag 8 a5 PARTIAL PRESSURE HS, psia e 06 0s , 000 Oz 04 0s 08 1.0 12 14 MOLE RATIO tN LIQUID, H,S/DEA Figure 4.22 Mutual solubility of HzS and COz in 3.5N DEA at 50°C. 108Woues Cb;/mOLE OeA IN LIOU|| soo Ft iatiio os ostor+od-jos-tod-{o3-toa 300) oly 2 Zoo) coll Es 2 S B & = 2 = E & z 06 oat bo 82 0a ‘06 3e 10 12 14 MOLE RATIO IN LIQUID, HpS/0EA Figure 4.23 Mutual solubility of HpS and COp in 3.5N DEA at 100°C. INGLE He S/HOLE, BRA TN LTOUTE $ oe i2110030867 08 63 [0 2 oil [oo 600 g S 8 w = Ss 8 g ee | 2 + a = = & on L ‘08 02 04 08 68 10 12 14 MOLE RATIO IN LIQUID, COg/DEA Figure 4.24 Mutual solubility of HS and C02 in 3.5N DEA at 100°. 103What would be the equilibrium partial pressure of CO, over this solu- tion? From Fig. 4.17 at 50°C the equilibrium partia? pressure of CO over this solution would be 7.0 psia. From Fig. 4.19 the equilibriud (0, partial pressure is 73 psia at 100°C. With 0.6 mol CO, /mol DEA at 50°C, Foo, ~ 7 pete at 100°C, "co, = 75 psia at 70°C 1/323 = 1/343 73) _ 2 0.845 + .448 = 1.293 00, = 19.6 psia TABLE 4.6 ENTHALPY OF SOLUTION OF CO, in DEA SOLUTIONS -bH,, Btu/Tb, co. s in 2 erate epee eee errr 0.5 562% S16 431e 364% 308% 265% 229% 2.0 655 580 463 378 312 265 229 3.5 Thy 636 490 390 316 I 266 229 5.0 821 688 513 399 319 267 229 6.5 885 719 529 405 321 267 229 8.0 939 Thh Sate ‘407 322 «| «267 229 * Calculated at 50°C 110TABLE 4.7 ENTHALPY OF SOLUTION OF H)S IN DEA SOLUT!ONS a, Mole Ratio in Liquid OH, HOS/DEA Beu/Tb,, HS 0.2 604 oO. 550 0.6 506 0.8 403 1.0 208 1.2 158 1h 139 Example Problem For purposes of explaining the procedure to be used in involving Fig. 4.17 to 4.24 in the solution of a sour gas treating problem, consider the following solution. 50 MMSCFD per day of gas containing 0.5% HS and 2.0% CO, are available. The gas enters the sweetening unit af 900 psig and“90°F. Lean DEA from the regeneration system is introduced to the contactor at 122°F. What amine circulation rate will be required to reduce the concentration of the H,$ to 1/4 grain per 100 standard cubic feet? Solution The first step in the solution 4s to determine the residual loadings of HS and CO, in the regenerated DEA solution. Unfortunately, infor- matign on DEAhas not been quantified in the form of Fig. 4.14 and 4.15 as has been done for MEA. However, for general purposes, the residual concentration of acid gases and DEA can be assumed as 0.01 mols of H)S per mol of DEA and 0.03 mols of CO, per mol of DEA. Assume that the concentration of DEA in the solution circulating is 2.0 NW or approximately 20% by weight. This means that Fig. 4.17, 4.18, 4.19 and 4.20 will be used in the solution.The first step is the determination of the acid gas constituent partial pressures in the sour gas to be treated: ppH$ = 0.005 x 915 = 4.6 pai ppCO, = 0.02 x 915 = 18.3 psi 2 The next step {nvolves the determination of the heats of reaction of the individual acid gas constituents in the DEA solution. Tables 4.6 and 4.7 show these heats of solution (reaction) to be composition de- pendent. For the purposes of this problem assume that the heat of solution for H,8 is 604 Btu per pound and for CO, is 580 Btu per pound. (These’heats of reaction assumptions wilf be subject to later checking and correction.) Based on these heats of reaction, the total heat of reaction per mol of sour gas entering the contactor will be: AH = (0,005 x 34 x 604) + (0.02 x 44 x 580) = 613 Beu DEA contactors follow the same principle as MEA contactors as outlined in Fig. 4.16. Consequently, the sweetened gas leaving the contactor essentially will be at the teuperature of the entering DEA solution or 122°F, The heat taken up by the gas in being heated from 90 to 122°F will be: £(0.65 x 26.9) x 0.65 x (122 - 90)1(0.975) = 613 Btu The heat capacity of 2.0 N DEA solution 4s 0.93 Btu/lb°F. This must, however, be converted over to an equivalent heat capacity per mol of DEA circulated. Btu. (_11b sotn. _\ / 105 1b DEA cy 7 0099 3s (eines) (I? Define a variable, R, as being the mols of DEA circulated per mol of sour gas. Using the variable R (the minimum DEA circulation required to remove the acid gas from the sour gas) a heat balance around the contactor is ) = 476 Btu/lb mol DEAF 232. 0.487 R(476Ae] + 381 = 613 R= en 7 ae The next thing necessary is to assume a temperature for the rich DEA solution leaving the contactor. If the proper temperature is assumed, the heat balance will balance. If the temperature is improper, the partial pressures of the acid gas constituents will indicate the direc~ tion in which the next temperature assumption should be made, 112Assume a rich DEA solution outlet temperature of 133°F. Under these conditions the minimin DEA circulation ist 0.487 133 ~ 122 The individual acid gas constituent molar concentrations in the DEA solution will be: R = 0.0443 mols of DEA per mol of sour gas co, = 0.03 + = 0.483 mol CO, /mol DEA 0.2 0.0443 = 0.123 mol H,8/mol DEA Under these conditions the acid gas partial pressures are: Partial Pressure Hompenent Tae [21258 co, 5.8 75.0 aS 1.6 9.0 In order to determine the vapor pressures of the components at 133°F, interpolate according to the logarithmic vapor pressure rule. eee Sez 593 25. tn ppco, = In 5.8 + [12 —412 | in 35 ppco, = 8.3 psi 382" 372 gee k 382" 393 3.0 ie : Jn ppHyS = In 1.6 + JAP —212 | an Te ppli,S = 2.0 psi 382 ~ 87 The vapor pressures for both H,S and GO, are lower than those in the sour gas. Consequently, a higher tempefature must be assumed. Assume that the outlet temperature of the DEA from the contactor is 135°F. Under these conditions the partial pressures of the acid gas, constituents over the DEA solution are: 120.0 = In 14.0 + dn aa 9 aire 14.0 ora In ppco, ppCo, = 19.9 pst ia 582 2 a0 3In pplt,S = In 3.8 + a z 7 oot,8 = 67 pel a From the results of the calculations at 133 and 135°F, the limiting acid gas constituent appears to be carbon dioxide. This can be checked by interpolating to determine the temperature of the DEA solution leaving the contactor. for 00, Q2) = 134.8 (2) = 134.95° = 0.487. 7 7 Ruin 7 d7-g 7 +0380 mol DEA/mol Sour Gas This shows carbon dioxide to be the limiting acid gas constituent. At this point the heats of reaction of H,S and CO, in the DEA solution must be checked. For the purposes of determining”the DEA circulation rate, the original assumptions are sufficiently accurate. Under con~ ditions of minimum DEA circulation, the acid gas concentrations will be as follows: HS = 0.01 + = 0.141 mols of HS per mol of DEA 02 co, 0.03 + 0: 07535 ~ 0-556 mols of CO, per mol of DEA The foregoing calculation has determined the minimum DEA circulation rate that can be used to renove the H,8 and CO, from the sour gas. The actual value of DEA circulation will depend upda the factor of safety that one wishes to eaploy. Commonly used factors of safety include the range 1.15 - 1.25 times the minimum DEA circulation rate determined as above. 14‘SOME FACTORS AFFECTING MEA PERFORMANCE There is considerable disagreement as to the mechanism of the absorption of acid gases in amine solution. Some people feel that there is a direct reaction between the acid gases and the amine in the aqueous amine solution. Others believe that acid gas constituents must be absorbed in the amine solution before they can react. The two mechanisms are considerably different and would call for different interpretations of data. Another complicating factor is the effect of dissolved acid gases on the physical properties of the amine solution. Dingman (4.13) has presented information on the effect of temperature and acid gas composition on the viscosity, specific gravity and pH of MEA solutions. Fig. 4.25, 4.26 and 4.27 are taken from his article and show the effects of temperature and acid gas concentration on these variables. Workers generally agree that there is a reciprocal relationship between mass transfer coefficient ‘and viscosity. Thus, the higher viscosities tend to have lower mass transfer coefficients. Increases in specific gravity or density with respect to acid gas loading are not overly large. However, they can have an effect, both on mass transfer and on horsepower required to make up line solution pressure drop. I eoamce counoel wea | \ 15.32 WOLE Coy 08 HOLE MSI/WDLC WEA i pte Sa f | Figure 4.25 oe ee Effect of acid gas composition and temperature on the viscos- | ity of MEA solutions Viscosity, Centistokes | Pe sow eee 0 ea Temperature °FSPECIFIC GRAVITY 130 150 TEMPERATURE °F Figure 4.26 Effect of acid gas composition and temperature on the specific gravity of MEA solutions 116pH 13.5 13.0 12.5 15 wt. z Bale 12.0 Mea. i 11.5 4 11.0 0. = Moig 10.5 1? MOSS, S, ol OS Mole M 10.0 Store on MEF E4 oa Mo S22 Hoy 9.5 le Co SH syne “Mol, Sle Meg 9.0 |? MEA 8.51.06 Mole £0,7Mo} 8.0 © MEA 7.5 50 70 90 110 130 TEMPERATURE °F Figure 4.27 Effect of acid gas composition and temperature on the pH of MEA solutions u7 150The effect of acid gas loading on MEA solution pH is interesting. This may explain part of the reason for corrosion problems in MEA plants. Two trends are definitely clear from Fig. 4.27 — increases in acid {gas loading lower the pH of the solution as does an increase in temperature. Both of these factors may be at work in some areas of severe corrosion such as heat exchangers and the top of the stripping column. The prediction of tray efficiencies and/or mass transfer rates is not exact. In general, 20 actual trays (or an equivalent height of packed column) will be provided in the absorber. In most instances, these 20 trays will provide sufficient contact for treating of sour gas to pipeline specifications, providing that the solution Is properly regenerated and that gas loadings in the rich amine solution are not excessive STRIPPER DESIGN The primary problem in designing the stripper is estimating the heat duty on the reboiler. This heat oad includes: 1. The sensible heat required to raise the temperature of the feed amine solution to that of the regenerated solution leaving the reboiler. 2. The heat of reaction required for dissolution of the acid gases from the amine. 3. The stripper column condenser duty which is essentially the heat of vaporization of the water leaving the stripping section of the column that is condensed and returned to the stripper as reflux. Properly estimating the condition and enthalpy of the feed to the stripper is a very complicated problem. As the feed to the stripper passes through the amine-to-amine heat exchanger, its temperature will rise. As the temperature increases, a portion of the absorbed acid gases will be vaporized. The portion of the acid gases that vaporizes will require 2 heat of dissolution. Estimating the amount and composition of acid gases regenerated is a difficult, trial-and-error procedure. A simple practical approach is to assume that no vaporization occurs in the amine-to-amine heat exchanger and that the feed solution simply takes up the heat released by the lean amine solution; that all of this heat goes to raise the temperature of the feed to the stripper. This increase in feed temperature does tend to decrease the amount of heat required in the reboiler. However, since no acid gases are assumed stripped from the solution in the amine-to-amine exchanger, this tends to increase the reboiler duty with the two opposite effects essentially counteracting each other. In the general case, however, the amine-to-amine heat exchanger will not transfer the maximum amount of heat. As a rule-of-thumb, the hot, rich amine solution leaving the amine exchanger should be at about 180° to 200° F. for proper teed temperature to the stripper. in general, this will require less heat than is available for transfer from the hot lean solution leaving the bottom of the stripper. Also, in most cases, the two amine streams will exchange heat on essentially an equal temperature difference basis — 2 10° lowering in the lean amine solution will result in a 10° rise in the rich amine stream. As a result, the engineer is usually sate in arbitrarily fixing the temperature of the feed stream to the stripper column and 19 a8 follows to calculate the stripper, based primarily on heat balance considerations. 1sExample Problem For the purposes of this example problem, assume that the rich amine feed to the stripper is entering at 190°F. and that the reboiler is operating at 240°F. The original estimate of acid gas content in the Jean amine was based on a steam rate to the reboiler of 1 Ib per U.S. gallon. Assume the steam enters the reboiler saturated at 250°F. (latent heat equal 945 Btu/Ib). A heat balance around the stripper is: Ban + = Hag + % + Bay where: H,, = heat entering with rich amine solution Btu/unit of time Hyg 7 heat leaving with acid gas from top of stripper, Btu/unit of time Hy = heat leaving stripper in lean amine solution Beu/unit of time Q, = reboiler duty, Btu/unit of tine QG = condenser duty, Btu/unit of time The problem now is to properly evaluate each of these terms. There is sufficient information to evaluate the left-hand side of the equation and two of the three terms on the right-hand side of the heat balance equation, with the condenser loading being unknown. With the feed entering the column at 190°F. there are: 350 gal/min of 15% MEA solution containing 0.476 moles of H,S and 2.735 moles of CO, The lean solution leaving the reboiler at 240°F. will be 350 gal/min of 15% NEA containing 12 grains of H,S and 800 grains of CO, per gallon, equivalent to a total of 0.0176 moles of HS and 0.909 mole of CO, per minute. Acid gas overhead will be 0.458 moles of HS and 1,826 moles of CO, per minute. With the acid gas overhead will bé sufficient water vapor for reflux to the stripper. The amount of this water vapor will be deter- mined from the condenser loading as shown below. The term (i, - Hy) can be evaluated as the difference in heat required for the leati“aniné constituents to be heated from the feed temperature of 190°F. to the lean amine temperature of 240°F. This sensible heat requirement is: 2947 1b/min x 0.98 Btu/Ib-°F. x (240 - 190)°F. = 144,380 Btu/ainThe acid gas portion of the rich amine solution is changed from a reacted gas in the liquid phase to acid gases in the overhead. ‘The amount of heat required for this conversion is essentially the heat of reaction of the acid gages with the amine. This has previously been calculated as 8.09 x 10° Beu/min. ‘The heat released from the steam in the reboiler is: 350 gal/min x 945 Btu/1b = 330,750 Bru/min This is based on the assumption that the heat release in the reboiler is the heat of condensation of the saturated steam entering the reboiler. Rearranging the heat balance equation gives: 8 = Fay ~ Bad - Hae + Substituting Q = -(144,380 + 80,900) + 330,750 = 105,470 Btu/min The amount of water condensed in the condenser can be calculated by assuming that the condenser operates at 20 psia, where water had a heat of vaporization of 970 Btu/1b. This gives: 105,470 Beufmin __ Te ib/ecie + 570 BravIb = 6-04 moles water condensed/min Recommendations for the amount of water returned to the column as reflux vary from 2-to-1 to 3-to-I on the basis of the total acid gas going overhead in the stripper. 6,04 moles of water/min _ Tie moles OF acid gas/ein ~ 2°58 moles water/nole acid gas ‘This falls within the recommended range and should give satisfactory stripping operation. Modelling Ethanolamine — Acid Gas Equilibrium Over the years many efforts have been made to systematize the equilibrium between ethanolamines and acid gases in such a way that information could be made useful for processidesign calculations. One approach which has met with reasonable success is to describe the reactions taking place between the ethanolamines and acid gas constituents. With appropriate pseudo equilibrium constants written for each reaction the equilibrium composition of all species in solution can be calculated. Through use of appropriate relationships to equilibrate vapor (gas phase) and liquid composition the entire ethanolamine — water — natural gas — acid gas system can be modeled at equilibrium. Kent and Eisenberg (4.15, 4.16) proposed a set of equations as follows to mode! the COg/H9S reactions with ethanolamines: 120BRIN =e HY + RR NH (4.4) RR'NCOO” + HO === RR'NH + HCO; (4.5) HO + CO, == HY + HCOS (4.6) 10 == + OH (4.7) Heo, === H+ (4.8) HS Se HY + HST (4.9) to eee (4.10) The equilibrium partial pressures of COp and HgS in the gas phase were related to the free concentrations of COz and HS in solution by the Henry's law relationship: p(co,) = H"(CO,)c0,] (4.11) (Hs) =H (H,S)fH,S] (4.12) Values of the pseudo equilibrium constants for Equ. 4.4 to 4.10 and values of the Henry's law constants in Equ. 4.11 and 4.12 are available in the literature. However, when literature values for the equation con. stants were used Kent and Eisenberg reported failure to satisfactorily match experimental data for CO — HS — ethanolamine equilibrium. Their next step was to accept literature values for the equilibrium constants for Equ. 4.6 to 4.10 and for the Henry's law constants, Values for the equilibrium constants for Equ. 4.4 and 4.8 were fitted to experimental pure component data for C02 — ethanolamine and HS — ‘ethanolamine equilibrium. This gave a much better reproduction of the experimental data as shown In Fig. 4.28 and 4.29 and there is good agreement between calculated curves and the experimental data points, Batt, et al. (4.17) used nuclear magnetic resonance (NMR) scans to study the reactions of MEA, DEA and TEA with Hg and COz. This work was later extended by Rahman (4.18). Conclusions concerning the reactions of ethanolamines and acid gases are: 1. Reaction with H2S proceeds to form the bisulfide of the ethanolamine. in ionic form the reaction would be written as a simple protonation of the amine aaa RNH, + HyS === RNHS + HS (4,13) The reaction is exactly the same for primary, secondary and tertiary amines.