cu AIS Hydrate Engineering E. Dendy Sloan Jr. Director, Center for Research on Hydrates Colorado School of Mines Edited by TUTO MEXICANO PEDR™ J. Ben Bloys IBLICTROA Principal Research Engineer Rear con No, IROC AGILE Arco E&P Technology Ine. ree GL Voi First Printing Henry L. Doherty Memorial Fund of AIME Society of Petroleum Engineers Inc. Richardson, Texas 2000Ti PE LBRO _-—— (Ci® Table of Contents tne Chap Te FAG Hon of Prevention Is Worth a Mile of Cure seed Precis ae Libme sp a225 1 agp . 1 ug Depressurization Hazards 1 13. Hazards of Applying Heat To Remove Hydrate Plugs : 2 {2b9s0432K 14 Hazards of Explling Plugs With High Upstream Pressure... 2 13. Genera Hydrate Pg Sale Considers 2 pe 16 Conclusions 3 BIBLIOTET A Chapter 2—Preventon by Design: How To Ensure That Hydrates Won't Form »++-00++ 2.1 Introduction 22. Hydrate-Formation Conditions . . 2.3. ‘Thermodynamic Prevention Methods 24 Where Do Hydrates Form? 2'5 One-Minute Estimate of Hydrate-Formation Conditions (Accurate to 50%) =... 9 2.6 Ten-Minute Estimation of Hydrate Formation/Inhibition (Accurate to 25%) ... 10 27 Most Accurate Calculation of Hydrate Formation and Inhibition... 16 2.8 Hydrate Formation During Expansion Through Valves or Restrictions 20 29. Hydrate Control Through Chemical Inhibition and Heat Management ........ 24 2.10 Hydrate Prevention During Drilling and Testing, 31 2.11 Design Guidelines for Hydrate Prevention cee 36 Chapter 3—Ilydrate-Pag Remediation Introduction 32. How Do Hydrate Biockages Occur? 3.3. Techniques To Detect and Locate Hydrates 34 Techniques To Remove a Hydrate Blockage feos 3.5 Avoiding Hydrates on Flowline Shut-In or Startup .... 56 3.6 Recommendations and Future Development Areas 37 Chapter 4—Eeonomies 4.1 Introduction 4.2. Economics of Hydrate Safety 4.3. Economics of Hydrate Prevention 4.4 Economies of Hydrate Remediation .. Appendix A—Gas Hydrate Structures and Properties and How They Form «+++ 6 AL Introduction : 6 A.2_ Hydrate Crystal Structures 6B ‘A3_ Properties Derive From Crystal Structures ce B ‘Aa Formation Kinetics Relate to Hydrate Crystal Structures beeen Appendix B—User’s Guide for HYDOFF and XPAND Programs B.A Word of Caution B2_ Executive Summary B.3 HYDOFF B4 XPAND Appendix C— Additional Case Studies of Hydrate Blockage and Remediation .. Appendix D—Rules-of-Thumb Summary .. Author Index ..... Subject Index .....-Chapter 1 Safety First: A Gallon of Prevention Is Worth a Mile of Cure 1.1 Introduction This chapter presents ease histories of some of the dangers associated with hydrate plugs. Many instances of line rupre, Sometimes accompanied by los of life, are attributed to the forms tion of hydrate plugs. Three characteristics of hydrates contribute 10 safety problems, 1. Hydrate density is similar to ice, and upstream pressure can pro- pela dislodged hydrate plug at high velecity In 1997 DeepStar Wyo- ‘ming field tests, plug lengths ranged from 25 to 200 ft with velocities between 60 and 270 ftsee. As Fig. 1.1 shows, such velocities and ‘masses provide sufficient momentum to cause two types of file pipeline restriction (rite), obstruction (lange or valve), ora sharp change in direction (bend, elbow, ote). Firs, hydrate impact ean fracture pipe, and, second, extreme gas compresion can cause pipe rupture downstream of the hydrate path 2. Hydrates can form either single or multiple plugs, with no ‘method to predict which will occur. High differential pressures ean be trapped between plugs even when the discharge ends of plugs are depressurized, 3. Hydrates contain as much as 180 volumes of gas at standard temperature and pressure (STP) per volume of hydrate, When hy. rate plugs ae dissociated by beating, any confinement causes rapid _gas-pressure increases. However, heating isnot an option within bu ried pipe because ofthe difficulty of locating the plug and the eco- nomics of heating an inaccessible pipeline. Field engineers discuss the “hail-on-a-tn-roof” sounds that small hydrate particles make when striking a pipe wall, Such small, mo- bile particles can accumulate into large masses that occupy aconsid- rable volume, often filling tens or hundreds of feet ofthe pipeline Tenth. Atempts to blow the plug out ofthe line by increasing up stream pressure (see Appendix D, Rule-of-Thumb 19) result in ‘addtional hydrate formation and perhaps pipeline rupture. ‘When a plugis depressurized with high differential pressure, the dislodged plug can be a very dangerous projectile, as Sees. 12.1 through 1.23 indicate." 1.2 Single-Sided Plug Depressurization Haza 1.2.1 Case Study 1.11991 Chevron Incident. A foreman and an ‘operator were attempting to clear a hydrate plug ina sour-gas flow “Kertano Cost, ALE: Hytn aNtlGineraaoMe 62) SAFETY FIRST: A GALLON OF PREVENTION IS WORTH A MILE OF CURE. line, They had bled down the pressure inthe distant end from the Wellhead and were standing near the ine when the line failed, prob- bly from the impactofa moving hydrate mass, A large piece of pipe struck the foreman. Although an air ambulance was deployed. the foreman was declared dead on artval atthe hospital. No pre-exist- ing pipe defects were found, See Case Study 3.8 fora similar inci- dent involving the loss of life. 1.2.2 Case Study 1.2—1991 Gulf Oil Incident. On 10 January 1991. the Rimbey gas plant was in the startup mode. A hydrate or ice piug formed in the overhead line Irom the amine contactor. The line had been depressured tothe flare system, downstream of the plug. The ambient temperature, which had been ~ 30°F, rose rapid ly owing towarming winds around midnight, At 2m, the overhead line came apart, killing the chief operator. In addition, the plants: fered approximately U.S. $6 million in damage Pre-existing cracks in the pipeline contributed to ths failure, While these cracks didnot impair the piping's pressure-contaning. ability under steady-state conditions, they reduced the piping strength under the transient (impact) conditions when the plug broke free 1.23 Case Study 1.3—1991 Mobil Incident, At 11:30 am. on 2 January 1991. two operators attempted to remove a blockage in a sour-gas flowline, which had been plugged for about 3 days. The downstream side ofthe plug had been depressured completly. The upstream portion of the line, originally at 1.100 psig, was depres sured completely toatruck within a 5-minute period, At 12:15 pam the flowtine failed and gas began flowing from somewhere around ‘the casing. An employes of wel-controlifirefighting company iso- lated the leak at 3:18 p.m The eruption of s hydrate plug at a Schedule 40, 3-in, serewed- pipe nipple caused the failure, Because both ends of the hydrate plug were depressured, two end plugs may have existed, with intermedi ate plogs (Fig. 12a. 1.24 Common Safety-Hazard Conditions. In Case Studies 1.1 through 1.3, common equipment circumstances existed, The sys tems exhibited the following characteristics, 1. They were out of service immediately before the incident 2, They did not have hydrate or Freeze protection, 3. They were pressurized while out of service. 44 They were being restarted. hs= a “Ato pog moves down Novis Aydt pg moves ‘own owlie a ery Highekcie ‘Wine pe bn ine pe Sane ‘one doh pest, XK Cos Vale Ife voc sigh enough te rome of pag am cae pressures Tare enough o pt the Nose Chose Valve Fig. 1.1—Safety hazards for moving hydrate plugs [from an internal report, Hydrate Handling Guidelines: Safety and Loss Control ‘Manus, Chevron Canada Resources (28 July 1992). Fig. 1.2—Safety hazards of high pressures trapped by hydrates [trom an internal report, Hydrate-Handling Guidelines: Safety ‘and Loss Control Manual, Chevron Canada Resources (23 July ‘992)) 5. They had high differential pressures across plug for short periods, The Chevron Canada Resources’ Hydrate Handling Guidelines* suggest thatthe danger of line failure owing to hydrate plug(s) is _more prevalent when (1) long lengths of pressurized gas are trapped “upstream, (2) low downstream pressures provide les cushion be- thveen a plug and restrition, and (3) resrictionsbends exist down stream ofthe plug, 13. Case Study 14—Statoil North Sea Incident. In the mid-1980's, hydrate plug occurred topside on platform ina Stat “Yntaeoning Gade: Sayan 088 Core au ia pe, Cheon nda carne (2182, 2 oil oil ied in the Norwegian sector ofthe North Ses. The line sec tion was valved of, and heat was applied to remove the plug. After some heating time, the work crew went to lunch intending to com plete the task on their return. When they returned, the crew found thatthe section of line had exploded during their absence Heat had been applied tothe midpoint of hydrate plug, and the plug-end portions contained very high pressures until the ine rup tured. Fig. L.2b shows a schematic of such a situation, Chap. 2 shows that pressure increases exponentially with ineressing temper ature when hydrates are dissociated. 1.4 Hazards of Expolling Plugs With High Upstream Pressure LAL Case Study 1.5—EIf Aquitaine North Sea Incident. Inthe 1970's, aplug occurred ona floating platform riser in the North Sea Blocking valves were closed and the pipeline was disconnected downstream of the plug. The discharge end of the pipeline was timed overboard with the intent of using high upstream pressure toextnide the plug from the lin. When the plug was expelled into the ocean, the force was so great that observers say the platform rose 8 in. in the ocean. 1.5 General Hydrate Plug Safety Consider The Canadian Assn, of Petroleum Producers’ (CAPP) hydrate guidelines? suggest three safety concerns in dealing with hydrate blockages. 1. Always assume multiple hydrate plugs; there may be pressure ‘between the plugs. 2, Attempting fo move ice (hydrate) plugs ean rupture pipes and vessels '3 While heatinga plug normally isnot an option fora buried pipe- tine, any heating shoul be done fom the end of a plug rather than the middle The thi recommendation could be expanded in consideration of buried line: (1) heating a submerged plug is not recommended be: ‘ase ofthe inability to determine the end ofthe plug as well spro- Vide for gas expansion on plug heating and (2) depressuring a plug gradually from both ends is recommended, HYDRATE ENGINEERING1.6 Conclusions ‘These case studies warn that hydrates can be hazardous to personnel and to equipment. Yet hydrate plugs can be dissociated safely with the procedures in Chap. 3, Remediation, ‘The prefered procedure, from both safety and economic consider ations to prevent the formation of hydrate plugs through design ad ‘operating practices. While the continuous use of many gallos of inhib toes may be costly, such expenses are overshadowed easily when plugs form and production stops. As tbe case stuies show it isnot uncom= ‘mon for several hundred yards of hydrate plugs o form, preventing production fora matter of weeks or months during remediation. SAFETY FIRST: AG) 1. King. Reval: CAPP Guidelines for the Prevention and Safe Handling ofHdrates, Canadian Assn, of Petleun Per, 1994), ‘SI Metric Conversion Factors fx3088" = E-01 =m °F CR-3V18 =C in, x25” E+00=em psi x6894 757 E+00=kPa LLON OF PREVENTION IS WORTH A MILE OF CURE 3Chapter 2 Prevention by Design: How To Ensure That Hydrates Won't 2.1 Introduction ‘The purpose of this chapter is (1) to indicate common sites of hy {rate formation, (2) to indicate design methods to provide hydrate protection, and (3) to provide designs that make remediation easier i'a hydrate plug occurs Hyarates form in hydrocarbon processes in two fundamental ways: slow cooling ofa fluid as ina pipeline (see Example 2.2) or rapid cooling eaused by depressurizaton across valves or through a turboexpander (see Example 2.3) Sec. 2.2 deseries the condi- tions under which hydrates form, and Sec. 23 gles sx prevention methods. Sec. 24 provides typical examples of hydrate formation ina well, a flowline, a platform, and an onshore process. Sees. 2.5, through 2.7 deseribe the design for hydrate thermodynamic inhi tion with low cooling of a pipeline. Sec. 2.8 provides design prac- tices for hydrate prevention with rapid cooling across a restriction like a valve. See. 29 gives procedures for preventing hydrates through inhibition and heat management, Sec. 2.10 states deiling practice 10 preveat hydrate formation. Finally, Sec. 2.11 summa- Fizes general design guidelines for hydrate prevention in hydrocar- bon processes 2.2 Hydrate-Formation Conditions Three conditions promote hydrate formation in processes. 1. Free water and natural gas components must be present. Gas ‘molecules ranging in size from methane to butane are typical hy- trate component, including COs, No, and HS, The water in hy- rates can come from free water produced from the reservoir or from water condensed by cooling the hydrocarbon uid. The pipe Tine residence time is usually insufficient for a blockage of hydrates to form from water remaining vaporized in the gas of water dis solved in a hydrocarbon liquid. 2. Low temperatures normally are needed for hydrate formation: however, even though hydrates are 85 mol% water, the system tem- perature does not ned to be below 32°F for these ice-like solids to ‘2ecur. Onshore pipelines usually experience hydrate conditions dur- ing winter months. Offshore, however, below approximately 3.000 f. fof water depth, the ocean-bottom (mudline) temperature is mark: ably uniform at 38 0 40°F and pipelined gas readily cools to this em- perature within afew mils ofthe wellhead, Hydeates can form easly temperatures higher than 70°F at high pressures as well 3. High pressures commonly promote hydrate formation. At 38°R, common natural gases form hydrates at pressures as low as 4 Form 100 psig: at 1,500 psig, common gases form hydrates st 66°F. Be- cause pipelines typically operate at higher pressures, hydrate pre vention is a primary consideration for flow assurance. 2.3 Thermodynamic Prevention Methods The three conditions necessary for hydrate formation lead to four classic thermodynamic prevention methods 1. Water removal provides the best protection. Separation e- ‘moves free water, Water condensation from the gas phase is pee- vented by drying the gas, ether with tithylene glycol to obtain wa ter content of less than ? IbmvMMscf or with molecular sieves t0 obtain lower water content, Water removal may not be feasible be- tween the wellhead and the gathering point (or platform), soother prevention schemes, such as inhibitor injetion (see Method 4), ‘ust be used. 2. Maintaining high temperatures keeps the system inthe hydrate- free region (sce Sec. 2.9.4). High reservoir uid temperature may be retained through insulation and pipe bundling, or heat may be added through hot Muids or electsical heating, although this latter ‘option is not economical in many cases 3. The system pressure may be decrease below the hydrate-forma- tion pressure. Todo so, one can design system-pressure drops at high- temperature points (e.g, botiomhole chokes). The resulting lower density, however, decreases pipeline-transporution efficiency. “4 Hydrate prevention is achieved mos frequently by injecting an Inhibitor, such as methanol (MeOH) or monoethylene glycol (MEG), which act a anifteezes, to decrease the hydrate-formation temperature below the operating temperature “Two new means of hydrate inhibition have been added to the the ‘modynamic inhibitor st and ae being brought ino common practice. 5. Kinetic inhibitors are low-molecular-weight polymers and small molecules dissolved in a carrier solvent and injected ant the ‘water phase in pipelines, These inhibitors bond to the hydrate su. face and prevent significant erystal growth for a period longer than the fre-water residence time ina pipeline. Condensate may or may not be present for this prevention method tobe effective, Water and small Hydrate crystals are removed ata platform or onshore 6. Aniagglomerants ae dispersant that cause the water phase 0 be suspended as small droplets inthe oi orcondensate. When the us pended water droplets convert to hydrates, pipeline lows are main- tained without blockage. Because this prevention method eles on HYDRATE ENGINEERINGa Hon Gomi 10; emulsified water hydrates, a condensed hydrocarbon is required. The emulsion is broken, and water is removed onshore of at a pltform “These methods are used individually or jointly for prevention. This chapter provides ways to use these six methods to prevent hy- rates inthe system design. 2.4 Where Do Hydrates Form? Fig. 2.1 shows a simplified offshore process between the subsea well inlet and the platform-expont discharge where virtually all hy- drate problems occur. The igure shows hydrate blockages in sus ceptible portions of te system: a) the well (b) the pipeline. or) the platform; Examples 21, .2,and 2.3, respectively, provide brief ‘descriptions of each, High reservoir temperatures prevent hydrate formation prior to the well when the system is flowing. After the platform, expor lines have dry gas and oiVondensate with insu cient water to form hydrates. InFig. 2.1, two unusual aspects ofthe system should be noted: (1) the water depth i shown as 6,000 fi butt may range to 10,000 ft, and (2) the distance between the well and the platform may range to:more than 60 miles. Such depths and distances provide cooling for the pipeline iuidsto low temperatures and high pressures, which are well within the bydrate-stbility region. Ata typical ocean tem- perature of 39°F, 400 ft of water depth provides pressures required for hydrate formation "The system temperature and pressure atthe point of hydrate formation must be within the hydrate-stabiity egion, a determined by the methods in Secs. 25 through 2.7. The system temperanure and pressure enters into the hydrate-formation region either through & normal cooling process (Example 2.2) or through a Joule-Thom- son process (Sec. 2.8) Fig. 2.2 shows atypical plot ofthe water temperature inthe Gulf of Mexico asa funetion of water depth, The plot shows that a high temperature of 70°F (or more) oocurs for the first 250 ft of depth However, when the depth exceeds 3,000 ft, the botiomwatertemper- ature is very uniform t approximately 40°F no matter how high the temperature is atthe ae/vater surface, ThisTemarkably uniform wa- fer temperature at depths greater than 3,000 ft occurs in almost all the Barth's oceans (caused by water-densit inversion) excepta few that have cold subsea currents “The ovean acts asa heatsink for any gas or oil produced so that ‘without insulation or other heat-control methods, any flowline id ‘cools o within a few degrees of 40°F within fev miles ofthe well "The rate of cooling with lengths function ofthe inital ese. voir temperature, the flow rate, the pipeline diameter, and other fuid-flow and heatransfer factors, As Sec. 2.5 explains, however the ocean-bottom temperature of 40°F is low enough to cause hy: drates to Form at any typical pipeline pressure Example 2.1: Hydrate Formation in a Well. Fig.2.3 showsatypi- cal subsea well in which fluids are produced through the wing valve and choke to the pipeline. A pressure indication just beyond the ‘choke is essential for determination of hydrate formation in the con- 100 Ocean Depth, ft 10,000. 2 9 4 50 6 7 80 Temperature ('F) Fig. 22—Water temperature vs, depth, Gulf of Mexico.? necting lowline. Approximately 300 to 500 ft below the mudline i the downhole safey valve (DSV), which is used as the initial emer: gency barrier between the reservoir and the production system. At the top ofthe well ae swab valves, which provide an entryway into ‘the wel for lubricating hydrate-dissociation tools (e.g. inhibitor in jection, heaters, coiled tubing) 10 reach any hydrate blockage. Hydrate formation in wells is an abnormal occurrence, arising daring drilling testing, o shutivstarcup ofthe well. While this ample provides a brief overview, See. 2.10 details methods fr pre- vention of hydrate formation in well dling andesting. Addressing well blockages may'be done with techniques discussed in Chap. 3, Remediation. Case Studies 3.3 and 3.8 provide two additional hy- drate remediation experiences in a wel avalath and Barker? provide a comprehensive set of conditions for dealing with hydrates in deepwater driling and testing, includ: ing two case studies of problems (summarized in Appendix C Case Studies C.23 and C.24) and four eae studies of successful hydrate ‘management. Typically, MeOH-injection capability is provided in the well at wo places: athe subsea tree and downhole several thou sand fest below the seafloor. The injection location and amount of ‘MeOH injected are determined with the procedure in Sec. 29.1 on MeOH injection. In offshore well drilling, a water-based drilling fuid frequently is used that can form hydrates and plug blowoutpreventers(BOP’s), kil ines, and other such equipment, when a gas bubble (or “kick” comes into the driling apparatus. This represents a potentially dan: gerous situation for well control. Hydrate formation during dling is an area of active research with several jont-industry projects. Barker indicates the following rules of thumb used by Exxon in considering hydrate formation with driling Auids * Drilling hydrate problems occur frequently, but have been ree ‘ognized only in recent years. PREVENTION BY DESIGN: HOW TO ENSURE THAT HYDRATES WON'T FORM s|< Swab Valve Christmas in Tee \ Wing Valve FX Crossover Valve << Master Valve Wellhead — Mausline Downhole Downhole Safety Valve Completion 9518in Fig. 23—Typical subsea well + When hydrates form solids, they remove water from the mud, leaving a solid barite plug +A well should not be designed tht operates only ouside the hy erperalre whore Tyealed | t | Formation Waiee Rate 7” 325 {BD 20 T Condensed Wale —87 [nip Ha ett Waerfo Treat 33 [hia | Wethanslinecin Ras] TOAS wa at Methanol Rateivect — <2 [alia |GALERLATION WORKSHEET (ure MeOH at 77) | Water hot S52 [ipmmitact [MEG recbon Rate 11900 [Water in cod 63 _librvMttsct | (pure MEG) [WATER conBENSEO—| “e188 [IpmNiset | [ MEG Ratetiuast —1 604 [Yai Water CONDENSED | “785 Temp inthe koe, ‘57 TB Ha Nata waie or abe) [53 T Ho cae omomaiae ogee | CETTE (Freere egress regard | “AIOE t we methanol [me |e ‘Summary of Results [nseced in water phase T 1 oa MEG = [needed in wal pasa —| { apr oa oat Hpmiisel 1 [soso ata t fine t t arora ga 7 pa MEG i gas 0 —Tlpm/MMsor aioli conawraste —| 37S oa 1 IMEGawo conderssis [225 [ipmyp — rare pT, 7 [water phase MEG to protoa Tso esters t aS haeransta Te iMESTo pee Ta festa se [wer pasa aetna geagio = —| 79 — Taam MEG in gs o—Tparmisst ashen 10 coodensas | —97 5 I MEG wis condaceate [225 nmi [TOTAL Metvanot Rate | #84 —[ pep fOTAL WEG Rate [1,766 ipmp. aatiaat nection Rate —| 0ST gay ——|- [WEG Hector ae | TS wa —] [pure MeOH at 7) [pore MEG) t Hitool ate S| ea. aga | i i t | Fig 2:12—Example 28 calculated through HYDALC. PREVENTION BY DESIGN: HOW TO ENSURE THAT HYDRATES WON'T FORM‘TABLE 2.7 COMPARISON OF HAND WITH COMPUTER, ‘CALCULATIONS FOR SECONO ESTIMATION METHOD Hand Mathoo | sult With Rules ol HYDCALC Calculates Quantity ‘Thumb Result ‘Water condensed, utiMsct 591 6198 MeOH in water, bmvMMsct 2287 230.7 ‘MeO in gas, omMMsct a 247 ‘MeO in condensate, br/MMsct 17 17 | Total MeOH injection, lomvnaMsct 2674 276.25 Total MeOH injection, galNnsct 403 422 2.7.1 Hydrate Formation and Inhibitor Amounts in Water Phase. HYDOFF is a PC-compatible computer program provided ‘on the disk included with this Book. The program enables the wser to determine hydrate-formation conditions and the amount of inhib ‘tor needed inthe Tree-water phase. Asa minimum, a386-IBM com puter with 2 megabytes of RAM is required. The program may be executed from either the Windows or the DOS environment, "To use the program, first load both HYDOFREXE and FEED,DAT from the accompanying 3.5in. disk onto a hard drive. ‘Appendix B provides user's manual with several examples ofthe tse of HYDOFF. Example 2.9 illustrates the simplest (and perhaps the most beneficial) use of HYDOFF. ity was calculated in Step 1 of Example 2.7 to be 0.784, Fig, 2.12 display all input data and results "The amount of MeOH injected is 42.2 gal/MMsef, and the amount of glycol injected is 59.4 gaMIMsef. For ease of use, the engineer may tum to HYDCALC to pesform the second approximation calculation. The following section pro- vides aceursey and limitations of both HYDCALC and the hand- caleultion methods, which are vita 1o their use. 2.646 Accuracy, Limitations, and Extensions for Second Estim: tion Method. Table 2.7 conains a comparison of the previous results| ‘with the hand-caleulation method and the HYDCALC method. While the hand calculation and the computer progeam provide ‘ony slightly different results, both include inaccuracies. For exam- ple, while iis possible to obtain more significant figures with HYD- CALC than with he chart inthe hand method, HYDCALC inaceu- racies are those of the ehurs on which HYDCALC is based ‘By use of HYDCALC, it was estimated that 7 wie MeOH was require inthe water phase 1o inhibit hydrate formation inthe pipe Tine, while measurements by Robinson and Ng? show that only 20, wi MeOH was requited for inhibition at the same gas composi tion, temperature, and pressure of Examples 2.7 and 2.8, “The major inaccuracies in the second estimation method are in the ¢gs-gravity hydrate-formation conditions, which are only accurate to=7"F of to +500 pia. Inthe Hammerschmidt equation, the i= hibitor temperature depression AT is accurate to 45. With such inaccuracies, the amount of MeOH or glycol injection could be in terror by 100% ormore. The principal vive of the secondestimation ‘method i ease of calculation rather than aceuracy. "A second limitation is that the method was generated for gases without H:S, which represents the case for many gases in the Gulf ‘of Mexico. Bailie and Wichert!® proposed a modification ofthe fas-gravity method for sour gases. 2.7 Most Accurate Calculation of Hydrate Formation ‘and Inhibition the HYDCALC results indicate that hydrate formation will occur ‘without inhibition, the engineer should elect wo do further, more a= ‘curate calculations. The most accurate method for hydrate-forma- tion conditions together with the amount of MeOH needed in the ‘water phase, is available asthe Final estimation technique ina com puter program, HYDOFF. This book provides a usee’s manual and fan example in Appendix B, The method details are too lengthy to include here: the engineer interested in program details is referred toChap. Sin Ref. 8 ‘See. 27 provides examples forthe most accurate methods for the following caleulations: * Calculation of hydrate formation and inhibition in water (See. 274), * Conversion of MeOH to MEG concentration in water phase (See. 212) * Calculation of solubility of MeOH and MEG in the gas (See, 273) * Calculation of solubility of MeOH and MEG in condensate (See. 2.7.4), 6 Example 2.9: Use of HYDOFF To Obtain Hydrate-Formation and -Prevention Conditions, Find the hydrat-formation pressure (ofthe following natural gas a 38°F and the amount of MeOH in the Water phase to inhibit hydrates t 38°F and 1,000 psa. The position is methane= 71,60 mol%, ethane=473 mol%, pro pane = 1.98 mol%,r-butane = 0.79 mol%.n-pentane =0.79 mol, tatbon dioxide = 14.19 mal%, and nitrogen = 5.96 mol% ‘Solution. The gas inthis example has the same composition asthe gs in Examples 2.7 and 28, s0 the results provide a comparison ‘with hand and HYDCALC computer calculations ofthe gas-gravity ‘method (Sec. 2.6.1) and the Hammerschmist equation See. 2.6.2) For convenience with multiple calculations, the reader may wish tweditthe program FEED DAT toreflect the gas composition ofthe problem, Modification ofthe FEED.DAT program is done in a flat, text editor, such 3s Notepad in Windows, by changing the compost tion ofeach component o that of the example gas and saving there sult, However, it is not necessary to use FEEDDAT, the gas com position may be entered as part ofthe HYDOFF program. Inthe following solution, each input from the user is underlined 1. From Windows or in the proper directory, click on, or type HYDOFF; press En 2, Alter eading te il screen, press Ente. 53, Atthe “Unis” screen, press (to choose * an psi then Enter. 44 At the FEED.DAT question seen, press Y and Enter iF you ‘wish to use the data in FEED DAT, of N and Enterif you wishto en- terthe gas composition in HYDOFF by band. The remainder ofthis examples written assuming thatthe user will enter the gas composi tion ia HYDOFF rather than use FEED DAT. The use of FEED.DAT. is simpler and should be considered for multiple calculations with the same ga, 5, Te next seen asks forthe numberof components present (ex- cluding water). Type 7 and Enter 6, The next screen requests Tist ofthe gas components prese ‘coded by numbers shown on the screen. Type 1,2, 3.5.7.8, and 9 Cin that order separating the entries by commas) and then Enter. 7. The next series of sereens request the input ofthe moe trac- tions ofeach component. Methane 0.7160 Enter Propane 0.0194 Enter ‘Butane 0.0079 Enter, [iteogen 0.0596 Enter Carbon Dioxide 0.1419 Enter ‘mPentane 0.0079 Enter. 8. At the “Main” screen, type 1 then Enter 9. Atthe “Options” screen. type I then Enter 10, Atthe sereen asking for the required Temperature, type 38 and Ener ~ ~T, Read the hydrae-formation pressure of 2306 psi (meaning ‘hydrates form at any pressure greater than 230 psia at 38°F for this Bas) 12. When asked for another calculation, type N for“No” then Enter 13, At the “Options” sereen type 2, then Enter. TA, At “Inhibitor” sereen, type 1, then Enier 15. tthe sereen asking forthe required temperature, type 38 and Enter HYDRATE ENGINEERINGTemperature, *F 90 10s 130 Temperature, °F 2 3 : 3 5 é gi | S ow gk | 8 : Zz: “ee | i : “poy 77 cereale comme comm |g sot] nmorshemecuomeyemane time tana sa 7. ae ae , Spur wey eer er cs we = Fig. 2.13-MoOH lost to condensate.® 16, At the seroen to enter the “WEIGHT PERCENT of Metha- nol," type 22 17. Read the resulting hydrate condition of 2.48% McOH, 38°F, and 972.7 psia Itmay require some trial and error with the use ofthe program be fore the correct amount of MeOH is entered that inhibits the system at the temperature and pressure ofthe example. One starting est ‘mate forthe trial-and-error process would be the amount of MeOH predicted by the Hammerschmidt equation (27 wt‘) in Example 2.7. Ngand Robinson! measured 20 w1% of MeOHLin the water ‘aired to inhibit hydrates at 38°F and 1,000 psa, A comparison of the measured value with the calculated value (22 wt) inthis exam ple and through the Hammerschmidt equation provides an indica tion ofboth the absolute and relative calculation accuracy, HYDOFF also can be used to predict the uninhibited hydrate formation temperature at 1,000 psi at $8.5°F through a similar tral land err process, as compared with 65°F determined by the gs ‘gravity method. No measurement are available for the uninhibited formation conditions of the gas in this example. In using HYDOFE, if components heavier than ndecane (Cio Hz2)are present, they should be lumped with n-decane because they are all nonhydate formers. 2.7.2 Replacing MeOH with MEG in the Water Phase. The con ‘centration of inhibiting MEG inthe water phase can be determined from MeOH concentration with a simple coreation of inhibitors wie MEG = ~ 1.209 + 2.34 (tS MeOH) = 00 (tS MeOH)? +0.0008(w1% MeOH) rn) To use Ea, 24, first determine the amount of MeOH required by use of HYDOFF 3s in Example 2.9. Insert the amount of MeOH in Eq, 2-4 to determine the amount of MEG needed in water o inhibit hydrates. Eg. 2.4 should be used for he free-water phase only. Ex ample 2.10 provides summary calculation of all he procedures in ‘See. 27. 2.7.3 Solubility of MeOH and MEG in the Gas. Fig. 2.10 fi ‘of recent measurements by Ne and Chen?® for Kiyjoy defined as the ratio of MeOH mole fraction in gas relative to water ()noW! {you in 50). Once the mole fraction of MeOH in waters deter ‘mined with HYDOFR, it may be multiplied By Kya 10 0bsin the Fig. 2:14-MEG lost to condensate* mole fraction of MeOH in the gas. As can be determined by Fig. 2.10, the solubility in the Water i ony slighty affected by pressure Over the range from 1,000 to 3,000 psia at offshore temperatures, Fora conservative estimate, the 3,000 psa line is recommended: Koy, = exp{8.706 ~ 5738 [1/TER)]} 2s) Fig. 2.11 providesan estimation of MEG dissolvedin gas at 1.000, ‘sig from the data of Polderman As indicated in the Figure, the amount of MEG in the vapors very small; Ng and Chen?® measure «negligible MEG concentration inthe vapor asa comparison. Ex ample 2.10 (See. 27) provides a summary calculation of al the procedures in Sex, 2. 2.74 Solubility of MeOH and MEG inthe Condensate Fig 2.13 isafitof measurements by Npand Chen? for Ky defined asthe MeOH mole ration n condensate elatve to Sale (= coin NChconia #20) Ones the mole fone MeOH in wae ode termined itay be mlipliedby Ky taba the moe frction Of MeOH inthe condense In Fig 15 ines re pressured pendent andthe olen lineshouldnt apply because ie abeence sac compounds piel Condensate The ifort sola Of MeOH In condensates of methane, propane, and w heptane recommended oun = exp(5.90 ~ $4045 x (1/TCRI} 26 214 shows similar measurements by Ng and Chen?® that specily the solubility for MEGGin the condensate By sre of Kg d= fined as the MEG mole fraction in condensate relative "hater ‘Ms i HOSE 20) NOt tha the Key, ves ae to = ders of magnitude lower than Kiyo values No pressure depen- dence is observed and the ine for MEG solubility tn methane ro pre, and heptane (or mesbsleycloheane) Is revommened Because toluene snot in condensate: Kuga = exp[420 ~ 7.2664 « (1/TCR)] en Example 2.10 provides a summary calculation of all the proce dures in See. 27 2.1.5 Best Caleulation Technique for MeOH or MEG Injection. ‘The following example is identical o that of Examples 7 and28, with the exception that both MeOH and MEG injection are calcu: Jated for comparison of each inhibitor withthe less accurate meth- ods of See. 26 PREVENTION BY DESIGN: HOW TO ENSURE THAT HYDRATES WON'T FORM . & ”Example 2.10: Most Accurate Inhibitor Injection Calculation. [A pipeline wth the gas composition described next has inlet condi tions of 195°F and 1,050 psi, The flowing gas is cooled bythe sur oundings toa temperature of 38°F. The gas also experiences. pres Sure drop 10 950 psia. Gas exits the pipeline at a rate of 3. MMsef/D, The pipeline produces condensate at arate of 25 BID. ‘with an average density of 300 Tbm/bbl and an average molecular Weight of 90 lbnvIbmole, Produced sal-fee water enters the pipe Tine ata rave of 0.25 BYD. ‘Natural gus composition: methane = 71.60 mol, ethane =4.73, rmolM, propane 1.94 mol%, n-butane =0.79 mol%, pen: fane 029 mol, carbon dioxide= 14.19 molS, and nitro gen =5.96 mol“ Find the rate ofboth MeOH and MEG injection needed to prevent hydrate formation inthe pipeline ‘Solution, Basis: The solution basis is 1 MMscf/D. Step 1. Calculate the Concentration of MeOH and MEG in the Water Phase: In Example 29 the MeOH concentration was calculated to be 22 wi ofthe Free-water phase at 38°F and 1,000 psa for this gs. By Use ‘of Eq. 24, the MEG concentration was caleulated at 33.6 wat inthe ‘water phase. Siep 2. Calculate the Mass of Liquid Hy O/MMscf of Natural Gas. I. Caleulte mass of condensed #0. Use the water content chart (Fig. 2.9), to calculate the water inthe vapor/MMsct. The inlet gas (at 1,050 psia and 195°F) water contents read 600 Ibm/MMscf The outlet gas (950 psia and 38°F) water contents read a8 9 Yon MMGsef. The mass of liquid water resalting from condensation is 600 thm _9 thm. _ $94 bm MMe! ~ MMscf ~ -MMscf 2. Caleulate mass of produced H;0 flowing into the line. Convert the produced water of 0.25 B/D tothe basis of Ibm/MMsef > 42 gal) (9.34 thm)/__ 1D 12s w/o no (“a )(“RP*) loa omi,0 MMset 3. Total mass of watee/MMscf gas. Add the condensed and pro duced water: 591 Ibm | 27.4 Ibm _ 6184 thm MMscf* “MMscf "MM Step 3. Calculate the Rate of MeOH and MEG Injection. MeOH ‘and MEG can exist in three phases: water, gas, and condensate. The total masses of MeOH and MEG injected per MMsef are calculated 2 follows: 1. Calculate amount of MeOH and MEG in the water phase 3.22.0 wt’ MeOH is required inhibit the fee-water phase, and the mass of water/MMsct was calculated at 618.4 Thm, The mass of [MeOH in pounds in the free-water phase per MMscl is ols Aig 74.4 lm MeOH/MMset in the water 2 ws x 100% Solving for met phase btn Step 133.6045 MEGis required inhibit the fe. water phaser and the mass of wateMMscf was calulted at 6184 lbm in Step 3, The mass of MEG in pounds inthe fee-watr phase pet MMsct is irae + 818 HO Solving for mga =313.1 Ibm MEGIMMsetin the water phase. 2. Caleaate mount of MeOH and MEG Tost othe gas. 4, MeOIH lst o gas. The mole fraction MeOH inthe fee-vater phase is X 100%. 3.6 wie 0437, “The distribution constant of MeOH in the gas s calculated at 38°F (497.7°R) by Eq, 25, relative tothe MeOH in the water Kyou = enp[5.706 ~ 5,738 X [1/497.7R]} = 0.00296, aes) where °R="F + 459,69, ‘The mole fraction of MeOH in the vapor is yyeow = Koy -AMe04 OF YEOH = 0.00296 x 0.137 = 0,0004055. “The daily gas rate is 8.432 Ibm mol (=3.2% 108 sot(379.5 st bm mol), where an efi at 14.7 psia and 60°F). so thatthe MeOH Tost to the gas is 3.42 Tom mol (=0.0004055 8432) or 109.4 TD, Because the calculation bass is 1 MMsef/D, the amount of McOH lost is 34.2 Tom/MIMsef (= 109.4 Thmv3.2 MMset. 'b, MEG ost 10 gas. InFig. 2.1], use the SO iS MEG line todeter rine that the MEG lost to the gos i 0.006 Tom/MMsef at 38°F and 1,000 psig; such an amounts negligible. Ng and Chen?” measured ‘negligible concentration of MEG in the gas phase at conditions similar to those ofthis problem. °, Calculate amount of MeOH and MEG lost tothe condensate. {4 MeOH lost tothe condensate. The distribution of MeOH in the condensate is calculated with Klay, = 09890 = 54045» [1/4972°R} = oon, 26) where "RF 1459.69 ‘Se faction MeOH in condense is Mou 1 Rg ot Fe = 0.007020 137 = 0.009617 Tinea ene 350 th mot (25 BD 300 Iombors I bm sei80 Tom 1D? Mit terete, the eee of MeOH i conve is 002%, fm molt ‘00961? x 241 OG0BE1) or Ob) {EMEC tw to eondensate,The mole aon MEG tthe wate phase Ica ours, “The distribution of MEG between the aqueoss liquid and conden- sate is given by Keg * exp{4.20 ~ 7.2664 > (1/497.7°R] 3.08 x 10° en “The mole fraction MEG in condensate is ze = Kies MEG culated as 3.8% 10° (= 3.04 x 10°5 0.128). The condensate rat is 360 Ibm moV/MMsef (=25 B/D% 300 Ibmbbl> 1 Ibm mol90 bmx 1D/3.2 MMsef) so tht the amount of MEG in condensate is 199% 10-5 Tom mol/MMsci[ ~0,0000038 x 26/1 ~ 0.000038) or 0.0061 Ibm/MMsef), 4. Calculate the total amount of MeQH/MMsef and MEG/MMsef with Table 2.8. [ TABLE 2.8—COMPARISON OF MeOH AND MEG REQUIRED | In wat bavi 5131 Inga, omMMact 006 ineondansate,brtoct 0061 Teil brit sia. {Total gant ses HYDRATE ENGINEERINGa iF ae] Temperature, F Fig. 2.15—Dog lake fieldhydrate curves. (Personal commu- rlgation with JL. Tedd, 18 June 1997.) oom) aa asf =| , oJ | [wan fre wet | iE =| Fig. 2.17—Dag Lake fleld with burial, (Personal communication With JL Tod, 18 June 1987). The example illustrates that, for this gas condition, the injection amounts of MeOH and MEG are comparable. The more precise cal ‘ulation shown here, however, represent a considerable savings in the amount of MeOH injected (31.5 vs. 422 galMMscf in the second estimation method shown in Example 2.8) Case Study 2.1: Prevention of Hydrates in Dog Lake Field Pipe- line. As summary ofthe thermodynamic hydrate-prevention meth- ods, consider he tps taken probbithydatesin the Dog Lake eld export pipeline in Louisiana.” During the winter months, hydrates formed in the line, which traverses and and shallow wate (a marsh) Hydrae-formation conditions, shown in Fig. 2.18, are calculated by HYDOFF with , 10, and 20 wi% MeOH in the water phase. The Dog Lake gas composition is 92.1 mol% methane, 3.68 mol% ethane, 1.732 mol propane, 0.452 mol -butane, 0.452 mol% butane, 0.177 mols i-pentane, 0.114 mol% n-pentane, 0.112 mole hexane, 0.051 mol% heptane, 0.029 mol octane, 0.517 mol ni- ‘wogen, and 0.574 mol carbon dioxide “The pipeline pressure and temperature, calculated with PIPE: PHASE, were superimposed onthe hydrate-formation curve shown in Fig, 2.16. Gasleaves the wellhead at 1,000 psiaand 85°F, far from hydrate forming conditions. As the gas moves down te pipeline, it begin to cool toward ambient temperatures. Once the temperature reaches approximately 63°F, hydrates form, so MeOH must be add fd, The figure shows pipeline conditions andthe hydrate formation ‘curves for various concentrations of MeOH, indicating that 25 wi MeOH in water is needed to inhibit hydrates. ‘Todd et al, indicate that despite large quantities of MeOH in- {ected for hydrate prevention, 110 hydrate incidents occurred inthe Tine during winter of 1995-96 at a cost of U.S, $325,000. Combins- tions of four alternative hydrate-prevention methods were consid Temperature, Fig. 2.16—Dog Lake field—original conditions. (Personal com- ‘munication with JL. Tod, 18 June 1997). foo] af RIS ——] wo} of ff i 400 + i rf IL few enone “IZ oS eee eS Fig. 2.18—Dog Lake field with burial and heating. (Sep.=sepera- tof) Personal communication with J.L. Todd, 18 June 1997.) a 10 ; get / Ey Fig. 2.19—Dog Lake fleld with burial, heating, (Personal communication with JL. Todd, 18 June 1997.) ‘ered: (1) burying the pipeline, 2) heating the gas at dhe wellhead, {G) insulating the pipeline, and (4) adding MeOH. The details of euch prevention measure are considered next 1. Burying the pipeline. Portions ofthe Dog Lake pipeline were built over stretch of marsh. The exposure to winter ambient tem- peratures caused rapid reductions in the gas temperature. Burying the pipeline would protect it from low environmental temperatures because ofthe higher earth temperatures. Fig. 2.17 displays the tem- perature increase inthe pipeline alter exposed areas were buried rel- ative to the exposed pipeline in Fig 2.16. With pipeline burial the ‘ed for MeOH in the water phase was reduced from 25 to less than 19 we. 2, Wellhead heat addition, Catalytic in-line heaters could be installed atthe wellhead to increase the gas temperature to 125°F. Fig. 2.18 shows the pipeline temperature increase caused by the PREVENTION HY DESIGN: HOW TO ENSURE THAT HYDRATES WON'T FORM »4000 3000 2000 1500 1000 200 600 00 Pressure, psia 400 00 200 150 | 100 : - 30 40 50 60 70 80 99 100 1 Tomperature, °F 0 120 130 Fig. 2.20—Expansions into hydrate-ormation region." combined prevention methods of burial and welthead heating. Use ofthese two methods permitted the MeOH concentration tobe r= duced to approximately 14 wi% to prevent hydrate formation inthe line. It should be noted that heating may inrease the amount of cor- rosion inthe line. 3. Insulation, Insulation of exposed areas near the wellhead and batiery maintains higher pipeline temperatures, thereby reducing the amount of MeOH needed for hydrate inhibition. Fig. 2.19 displays the temperature increase inthe buried and heated pipeline when ex posed pipes were insulated, The pipeline is now outside the hydrate formation region, and MeOH addition is no longer needed “4. MeOH addition, Continved MeOH injection could be done at ancost of upproximately US. $1.50 to $2.00/ga. Thecostof MeOH. ‘oanofishore platform was U.S. $2,00/gal during the 1996-97 win- ter Because MeOH recovery may not be economical, MeOH isnor- ‘mally considered an operating cos. ‘This ease study illustrates how combinations of pipeline burial, insulation, heating, and MeOH injection can be used to prevent hy- rates, The selection ofthe hydrate-prevention scheme(s) is then a matter of competing capital and operating costs, as considered in Chap. 4 on economics. 2.8 Hydrate Formation During Expansion Through Valves or Restrictions ‘When water-wet gas expands rapidly through a valve, orifice, or other restriction, hydrates form because of rapid gas cooling caused by Joule-Thomsoa expansion. Hydrate formation with rapid expan son from a wet line commonly occurs in fuel gas or instrument gos Tines, as Example 2.13 indicates. Hydrate formation with high pes sure drops can occur in wel testing, startup, and gas-lift operations, even when the initial temperatures high, ifthe pressure drop is very large, This section provides methods 19 determine when hydrates will form on rapid expansion, A rough estimation method (See 2.8.1 s followed by a more accurate but resource intensive method (Gee. 2.8.2), concluding with prevention techniques in See. 2.8.3. Figs. 2 und 2.5 are schematics ofthe pressure and temperature of pipeline production stream during normal flow with entry into 0 ‘the hydrate-formation region. If the gas expands more rapidly, the ‘normal pipeline cooling curve of Figs. 2.4 and 2.5 takes on a much steeper slope, but the hydrate formation line remains the same, Fig. 2.20 shows to rapid expansion curves fora 0.6 gravity gas. Inter Sections of the gas-expansion curves with the hydrate-formation Tine Timits the expansion discharge pressures from two different high initial pressure/temperature conditions. In Fig. 220, the curves determine the restriction downstream pressure at which hydrate blockages form fora given upstream pres- sure and temperature. Gas A expands from?,000 psiaand 110°F un- ‘iitstrikes the hydrate-formation curve at 780 pia and S7°, so 780 sia represents the limit to hydrate-ree expansion. Gas B expands from 1,800 psia and 120°F to intersect the hydrate-ormation curve ata limiting pressure of 290 psia and 42°F, In expansion processes, the upstream temperature and pressure are known, the discharge temperature is almost never known, and the discharge pressure is normally set by a downstream vessel or pressure drop. Cooling curves, suchas the two in Fig. 220, were determined for constant enthalpy (or Joule-Thomson) expansions, obtained from ‘he first aw of thermodynamics for asystem flowing at steady-state, neglecting kinetic and potential energy changes: AH=0-1, . 28) where AA'=enthalpy difference across the restriction (down- stream-upstream), Q'= heat added, and W, = shafl work obtained at the restriction. Offshore restrictions (eg, valves or orifices) have no shaft work and, because rapid flow approximates adiabatic opera: ‘on, both W, and Qare zero, resulting constant enthalpy (AH™=0) ‘operation on expansion Because of the constant-enthalpy requirement, rapid gas expan- sion results in cooling except at very high pressures, where heating feccurs on expansion resulting from a compressibility derease with temperature. The upstream pressute at which the system changes from heating to cooling on expansion is called the Joule-Thomson Inversion pressure Rule of-Thumb 7, Natural gases cool on expansion from pres- sures less than 6,00 sia; at more than 6,000 psi, the tempera- ture inereases on expansion, Virtually all oshore gas processes cool on expansion because only afew reservoirs and no current Pipelines or process conditions are at pressures greater than 6000 psia G.G. Browa,22 who consructed the first natural gas enthalpy/etro- py cha, determined Rule-of- Thumb 7. 2.8.1 Rapid Caleulation of Hydrate-Free Expansion Limits, Kaiz!® generated chars to determine the hydatefree limit 0 gas expansion, combining the gas-gravty char (Fig. 2.8) 0 obtain the hydkate-formation line in Fig. 2.20, with gas enthalpy-entropy chars by Brown?? to obiin the cooling line ‘Cautioning thatthe chars applied 10 gases of limited composi: sions, Kaz provided expansion charts for zases of 6, 0:7, and 08 gravities, shown in Figs. 221 through 2.23, respectively: The ab Scissa (ot « axis) in each figure represents the lowest downstream pressure without hydrate formation, given the upsteam pressure on {he ordinate (axis) and the upstream temperature (a parameter on cach ine) should be noted that maxima in Figs. 221 though 223 occur atthe upstream pressure of 6,000 pia the Joule-Thomson inversion pressure, This provides a further Validation of Ruleof-Thumb 7. The following three examples for chart use ae from Kat's! criginal work Example 2.11a: Maximum Pressure of Gas Expansion. To what pressure may a0.6-gravity gas a 2,000 psia and 100°F be expanded ‘without danger of hydrate formation? Solution. From Fig. 2.21, read 1,050 pia HYDRATE ENGINEERING.