Gas Compressor Station Economic Optimization

7/30/2019 Gas Compressor Station Economic Optimization

1/18

ResearchArticle

GasCompressorStationEconomicOptimization

RainerKurz,1MattLubomirsky,

1andKlausBrun

2

1SolarTurbinesIncorporated,9330SkyparkCt.,SanDiego,CA92123,USA2SouthwestResearchInstitute,6220CulebraRoad,SanAntonio,TX78238,USA

Received23September2011;Accepted6December2011

AcademicEditor:MauroVenturini

Copyright2012RainerKurzetal.Thisisanopenaccessarticledistributedunderthe CreativeCommonsAttributionLicense,whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.

Abstract

Whenconsideringgascompressorstations forpipeline projects, theeconomic successof theentire operationdepends to a significantextenton theoperation ofthe compressors involved.In thispaper,thebasic factors contributingto theeconomics are outlined,withparticularemphasisonthe interactionbetweenthepipelineandthecompressorstation.Typicalscenariosaredescribed,highlightingthefactthatpipelineoperationhastotakeintoaccountvariationsinload.

1.Introduction

The economic success ofa gascompression operation depends to a significant extenton the operation ofthe compressors involved.Important criteriainclude first cost, operating cost (especiallyfuelcost),lifecyclecost, and emissions.Decisionsaboutthe layoutofcompressorstations(Figure 1)suchasthenumberofunits,standbyrequirements,typeofdriver,andtypeofcompressorshaveanimpactoncost,fuelconsumption,operationalflexibility,emissions,aswellasavailabilityofthestation.

Figure1:Typicalcompressorstationwith3gasturbinedrivencentrifugalcompressors.

2.CapitalCost:FirstCostandInstallationCost

Capitalcostforaprojectconsistsoffirstcostandinstallationcost.Firstcostincludesnotonlythecostforthedriverandcompressor,andtheirskidorfoundation,butalsothenecessarysystemsthatarerequiredforoperatingthem, includingfilters,coolers,instruments,and

valves, and, ifreciprocatingcompressors areused,pulsationbottles.Capital spares, operational spares, andstart-upandcommissioningsparesalsohavetobeconsidered.

Althoughnotintuitivelyobvious,thisisalsotheareathatisaffectedbydriverderatesduetositeambienttemperatureandsiteelevation:thepowerdemandofthecompressorhastobemetatsiteconditions,notatISOorNEMAconditions.

Installationcostincludesalllaborandequipmentcosttoinstalltheequipmentonsite.Itisdeterminedbycomponentweights,aswellastheamountofoperationsnecessarytobringtheshippedcomponentstoworkingcondition.

3.MaintenanceCost

Maintenancecostincludesthepartsandlabortokeeptheequipmentrunningatoraboveacertainpowerlevel.Thisincludesroutinemaintenance(likechangeoflubeoilandsparkplugsingasengines)andoverhauls.Maintenanceeventscanbescheduleorconditionbased.A cost related tomaintenance effort is the cost due to the unavailability of the equipment (see below). Maintenance affectsavailabilityintwoways.Many,butnotall,maintenanceeventsrequiretheshutdownoftheequipment,thusreducingitsavailability.Scheduledmaintenancehasusuallylessofaneffectthanunscheduledevents.Forexample,ascheduledoverhaulofagasturbinemaykeep

theequipmentoutofoperationforonlyafewdaysifanengineexchangeprogramisavailable.

Ontheotherhand,insufficientorimpropermaintenancenegativelyaffectstheavailabilityduetomorerapidperformancedegradationandhigherchanceofunplannedshutdowns.

4.Efficiency,OperatingRange,andFuelCost

Theperformanceparametersofthecompressoranditsdriverthatareimportantfortheeconomicevaluationareefficiencyandoperatingrange. Efficiencyultimatelymeansthe cost offuelconsumed tobring acertainamount ofgasfroma suction pressure toa dischargepressure.Intechnicalterms,thiswouldbeapackagewithahighthermalefficiency(orlowheatrate)forthedriverandahighisentropicefficiencyofthecompressor,includingallparasiticlosses(suchasdevicesto dampenpulsations,butalsopressuredropduetofiltrationrequirements) combined with low mechanical losses. This factor determines the fuel cost of the unit, operating at given operatingconditions.

Operatingrangedescribesthe rangeof possibleoperating conditionsin termsof flowand headat anacceptableefficiency,within thepowercapabilityofthedriver.Ofparticularimportancearethemeansofcontrollingthecompressor(e.g.,speedcontrolforcentrifugal

machines,orcylinderdeactivation,clearancecontrol,andothersforareciprocatingmachine)andtherelationshipbetweenheadandflowofthesystemthecompressorfeedsinto.Operatingrangeoftendeterminesthecapabilitytotakeadvantageofopportunitiestosellmoregas.Itshouldbenotedthatthereisnoreallowflowlimitfor stations,duetotheconceptualcapabilityforstationrecycleortoshuttingdownunits.Unitshutdown,inturn,hastobeconsideredwithregardstostartingreliabilityoftheunitsinquestion,aswellastheimpact

Page 1 of 8Gas Compressor Station Economic Optimization

05/01/2013http://www.hindawi.com/journals/ijrm/2012/715017/


2/18

onmaintenance.Inthisstudy,operatingrangeandtheupsidepotentialarenotspecificallyconsidered,mostlybecausenodataregardingfrequencyandvalueoftheseupsidepotentialswasavailable.Upsidepotentialcanbe realized,iftheequipmentcapabilitycanbeusedtoproducemoregas,thustakingadvantageofadditionalmarketopportunities.

Thecostofthefuelgasisnotautomaticallythesameasthemarketpriceofthetransportedgas.Itdepends,amongotherthings,onhowfuelusageandtransporttariffarerelated.Thecostoffuelisalsoimpactedbywhethertheoperatorownsthegasinthepipeline(whichmakesfuelcostaninternaloperatingcost)ortheoperatorshipssomeoneelsesgas.Insomeinstances,thefuelcostmightbeconsideredvirtually nil.Thus, the ratio between fuel cost andmaintenance cost canvarywidely. Inmost installations, however, the fuelcostmayaccountformorethantwo-thirdsoftheannualoperatingcost.

Thevalueofoperationalflexibilityissomewhathardtoquantify,butmanypipelinesystemsoperateatconditionsthatwerenotforeseenduringtheprojectstage.Operationalflexibilitywillresultinlowerfuelcostunderfluctuatingoperatingconditionsandinaddedrevenueifitallowstoshipmoregasthanoriginallyenvisioned.

5.Emissions

Anynatural-gas-poweredcombustionenginewillproduceanumberofundesirablecombustionproducts.NO x istheresultofthereactionbetweenNitrogen(inthefuelorcombustionair)andoxygenandrequireshighlocaltemperaturestoform.LeanpremixgasturbinesandleanpremixenginesreducetheNO xproduction.Catalyticexhaustgastreatments,suchasSCRs,canremoveabigportionofNOx intheexhaustgas,butalsoaddammoniatotheexhaustgasstream.

ProductsofincompletecombustionincludeVOCs,CO,methane,andformaldehyde.FuelboundsulfurwillformSO x inthecombustionprocess.

Thecombustionproductsaboveare usually regulated.For theeconomic analysis, thecost ofbringingthe equipmentto meetlocal orfederallimitshastobeconsidered.

CO2 istheproductofburninganytypeofhydrocarbons.CO2,andsomeothergases,suchasmethane,areconsideredgreenhousegases.

Typically,allgreenhousegasesarelumpedtogetherintoaCO2 equivalent.Inthiscontext,itmustbenotedthatmethaneisconsidered

about20timesaspotentagreenhousegasasCO 2.Thus,1 kgofmethanewouldbeconsideredasabout20 kgCO2 equivalent.Insome

countries,CO2 productionistaxed,andthereisapossibilitythatothercountries,includingtheUSA,adoptregulationsthatwouldgive

CO2 avoidanceaneconomicvalue.Inthiscase,theamountofCO 2 ormethanethatisreleasedtotheatmospherehastobeconsideredasa

costintheeconomicevaluation.

Itfurther needsto be considered that theengine exhaust isnotthe only source of emissions in a compressor stationrelated to thecompressionequipment.Therearealsosourcesofmethaneleaksinthecompressionequipmentthatmayhavetobeconsidered.Inthiscase,onehastodistinguishleakagethatiseasilycapturedandcanthusbefedtoaflareandleakagethatcannotbecapturedeasily.Further,itmayhavetobeevaluatedhowfrequentlythestationhastobeblowndown.Forexample,whetherthecompressionequipmentcanbemaintainedandstartedfromapressurizedholddeterminestheamountofunwantedstationmethaneemissions.

Lastly,otherconsumablesmayhavetobeconsidered.Thefrequencyandcostoflubeoilchanges,aswellaslubeoilreplacementsdueto

lube oilconsumption, generate costs on various levels: first, the replacement cost for lube oil, second, if thelube oil is used in thecombustionprocess,theresultingemissions,andthird,ifthelubeoilentersthepipeline,thecostduetothepipecontamination,includingpossiblytheincreasedmaintenancecostofdownstreamequipment.

6.Availability

Availabilityistheratiobetweenthehoursperyearwhentheequipmentissupposedtooperateandactuallycanoperateandthehoursperyearwheretheequipmentissupposedtooperate.Availabilitythereforetakesintoaccounttheentireequipmentdowntime,bothduetoplanned and unplanned maintenance events. Inother words,if the operator needs the equipmentfor 8760 hours per year, and theequipmentrequires3shutdowns,lasting2dayseachforscheduledmaintenance,andinadditionalsohadtobeshutdownfor3daysduetoaequipmentfailure,theavailabilityoftheequipmentwouldbe97.5%.OtherthanMTBF,whichdescribesthefrequencyofunplannedfailures,theavailabilityhasadirectimpactonthecapabilityofaninstallationtoearnmoney.Besidesthetypeofequipment,thequalityofthemaintenanceprogramandthemeasurestakentodealwithenvironmentalconditions(air,fuel,etc.)haveasignificanteffectontheavailability(andreliability)oftheequipment.

Thecostassociatedwithavailabilityisthefactthatthestationmaynotbeabletoperformitsfulldutyforcertainamountoftime,thusnotearningmoney.Thelossofincomecanbeduetothereductionofthepipelinethroughput,theunavailabilityofassociatedproducts(oilonan oil platform, condensates in a gas plant), or due to penalties assessed because contractual commitments are notmet. The valueassociatedwith thelostproductionis notnecessarily themarketvalueof thelostproduction.Itmayalsobe thelossof income fromtransportationtariffs,thecostfrompenaltiesfornotbeingabletosatisfydeliverycontracts,orthecostforlostopportunity(i.e.,duetotherequirementtokeepsparepowertocompensateforpooravailability,insteadofbeingabletousethesparepowertoincreasecontractuallyagreeddeliveries).

Stationavailabilitycanbeimprovedbyinstallingspareunits,butthisisanadditionalfirstcostfactor.

7.CompressorOperation

Therelationshipbetweenpressuresandflowsinanygivenapplicationthatemploysgascompressorsaswellassomeotherfactorsmayinfluence the arrangement of compressors in a station as well as the type of equipment used. The question about series or parallelarrangementsinastationhastobeconsideredbothinthelightofsteady-stateaerodynamicperformance,aswellasregardingtransientbehavior, spare strategy, and growth capabilities.Not only the full load behavior of thedriver, but also itspart-load characteristics

influenceareaslikefuelconsumptionandemissions.SincegreenhousegasemissionsfromCO 2 foragivenfuelgasareonlyinfluencedbytheoverallefficiencyoftheoperation,thereisastrongtiebetweenefficiencyandemissions.

Differentconceptssuchasthenumberofunitsinstalled,bothregardingtheirimpactontheindividualstationandtheoverallpipelineandthenecessityofstandbyunitsarediscussed.Thenumberofcompressorsinstalledineachcompressorstationofapipelinesystemhasa




3/18

significant impacton theavailability, fuel consumption, and capacityofthe system.Dependingon theloadprofileof the station, theanswersmaylookdifferentfordifferentapplications.

Theoperatingpointofacompressorisdeterminedbyabalancebetweenavailabledriverpower,thecompressorcharacteristic,andthesystembehavior.Thecompressorcharacteristicalsoincludesthemeansofcontrollingthecompressor,suchas

Ifvariablespeedcontrolisavailable,forexample,becausethedriverisatwo-shaft-gasturbineoravariablespeedelectricmotor,thisisusually thepreferred control method. It is often augmented by the capabilityto recycle gas. A typical compressor mapfor a speed-controlledcompressorisshowninFigure(Figure 2).Itshowstheareaofpossiblecompressoroperatingpoints.Thelowestflowpossibleisdeterminedbythesurgeline.Ifthestationrequiresevenlowerflows,gashastoberecycled.Onanypointofthemap,compressorspeedandpowerconsumptionaredifferent.

Figure2:Systemcharacteristicsandcompressormap.

Whereonthemapthecompressorwillactuallyoperateisdeterminedbythebehaviorofthesystem,thatis,therelationshipbetweenhead

(pressureratio)andflowenforcedbythesystem(Figure 2).LineBdepictsasystemwheresuctionanddischargepressurearemoreorlessfixedandthuschangeverylittlewithchangesinflow.Examplesarerefrigerationsystemsorsystemswherethesuctionpressureisfixedbyarequiredseparatorpressure,andthedischargepressureisfixedbytheneedto feedthatgasintoanexistingpipeline.LineAshowsthetypicalbehaviorofapipeline,whereanychangeinflowwillimpactthepressuredropduetofrictioninthepipeline.

Line C is typical for storage applications, where the pressure in the storage cavity increases with the amount of gas stored. If thecompressorisoperatedatmaximumpower,theinitialflowwillbehighduetotheinitiallylowpressureratio.Themoregasisstoredinthecavity,thehigheritspressure,andthustherequireddischargepressurefromthecompressorbecomes.Beingpowerlimited,theoperatingpointthenmovestoalowerflow.

Incaseofapipeline,theoperatingpointofthecompressorisalwaysdeterminedbythepoweravailablefromthedriver(Figure3).Inthecaseofagasturbinedriver,thepoweriscontrolledbythegasproducerspeedsettingandthepipelinecharacteristic.Wefindthispointonthecompressormapattheintersectionbetweenthepipelinecharacteristicandtheavailablepower.Increasingtheflowthroughapipelinewillrequiremorepowerandmorecompressorhead.

Figure3:Operatingpointanddriverpower.

Thechangeofoperatingconditionsovertimeisduetomanyreasons,suchasdropinfieldpressure,thepipelinegetsloopedorhastomeetanincreaseincapacity.Loopingthepipelinewillchangethecharacteristicofthepipelinetoallowmoreflowatthesameheadrequirement.So,any change in the pipeline operationwill impactpower requirements, compressor head orpressure ratio,and flow. Operationalchangesmaymove thecompressoroperatingpointover timeinto regimeswithlower efficiency.Fortunately, centrifugalcompressorsdrivenbygasturbinesareveryflexibletoautomaticallyadapttothesechangestoadegree.

Asindicated,variablespeeddriversallowforefficientoperationoveralargerangeofconditions.Ifaconstantspeeddriverhastobeused,theadjustmentofthecompressortotherequiredoperationusuallyhastorelyontheuseofrecyclingandsuctionordischargethrottling.Eithermethod,whileeffective,isnotveryefficient:throttlingrequiresthecompressortoproducemoreheadthanrequiredbytheprocess,thusconsumingmorepower.Onacompressorspeedline,throttlingwillmovethecompressortoapointatlowerflowthanthedesignpoint.Recyclingforcesthecompressortoflowmoregasthanrequiredbytheprocessandthusalsoconsumesmorepower.Itthusallowsthecompressortooperateatalowerheadthanthedesignhead.Ingeneral,anelectricmotorforaconstantspeeddrivehastobesizedforalargerpowerthanamotorforavariablespeeddriveunderotherwisethesameconditions.Furthermore,startingofconstantspeedmotorsrequirestooversizetheelectricutilities,especiallyifastartofapressurizedcompressorisdesired.

8.OperationalFlexibilityandStandbyRequirements

Operational flexibility undera larger numberof differentoperating scenarios hasto be studied. Demand varieson anhourly,daily,monthly,orseasonalbasis.Also,theavailablegasturbinepowerdependsontheprevalentambientconditions(LottonandLubomirsky[1]).Similarly,transientstudiesonpipelinesystems(Santos[2])canrevealtheoftenlargerangeofoperatingconditionsthatneedstobecovered bya compressorstationand thoroughanalysis can often revealwhichtypeof concept yieldslowerfuelconsumption.Lastly,scenariosthatarisefromfailuresofoneormoresystemshavetobeconsidered(OhanianandKurz[ 3]).Inanyapplication,operatinglimitsduetospeedlimitsareusuallyundesirable,becausetheymeanthattheavailableenginepowercannotbeused.But,becauseagasturbinecanproducefarmorepoweratcolderambienttemperatures,designsbasedonworstcaseambientconditionsmaynotbeoptimal.Optimizationconsiderationscanalsobefoundin[4,5].

Thequestforoperationalflexibilitycanbesatisfiedonvariouslevels:thecompressorandthedrivershouldhaveawideoperatingrange.Usingmultiplesmallerunitsperstationratherthanonelargeunitcanbeanotherway.Here,thearrangementinseriesorinparallelwillimpacttheflexibility.Thegasturbinesallowforimmediatestartingcapabilityiftheneedarises.

(i)(ii)(iii)(iv)

variablespeedcontrol,adjustableinletguidevanes,suctionordischargethrottling,recycling.




4/18

Inupstream and midstream applications, configurations usually need to cover a large range of operating points. Depending on theapplication,theoperatingpointscanvaryonanhourly,daily,weekly,monthly,orseasonalbasis.Contributingfactorsaresupply(e.g.,depletedfields ornewwells) and demand, changes ingas composition orsite availableengine power.Often, flow demandand headrequirementsarecoupled.Thisisveryobviousforpipelineapplications,wherethepressuredropbetweenstations(andthusthepressureratioofeachofthestations)isdirectlyrelatedtotheflow(Figure 4).

Figure4:Typicalsteadystatepipelineoperatingpointsplottedintoacompressorperformancemap.

Inotherapplications,thecompressoroperatingpointsarelimitedbythemaximumavailableenginepower.Thisis,forexample,thecaseforstorageandwithdrawaloperations.Here,thegoalistofillthestoragecavityasfastaspossible,whichmeansthatengineoperatesatitsmaximumpower.Sincethefillingofthestoragecavitystartsatverylowpressuredifferentials,theflowisinitiallyveryhigh.Asthecavitypressureandwithitthecompressorpressureratioincrease,theflowisreduced.Forapplicationslikethis,compressorarrangementsthatallowtooperatetwocompressorsinparallelduringtheinitialstage,withthecapability(eitherwithexternalorinternalvalves)toswitchtoaseriesoperation,areveryadvantageous.Incidentally,back-to-backmachinescannotbeusedinthiscaseduetotheirdelicateaxialthrustbalance.

Dynamic studiesofpipelinebehavior reveal adistinctlydifferentreactionof apipeline changes instation operatingconditionsthan asteady-statecalculation(Figure5).Insteadystate(orforslowchanges),pipelinehydraulicsdictateanincreaseinstationpressureratiowithincreasedflow,duetothefactthatthepipelinepressurelossesincreasewithincreasedflowthroughthepipeline(Figure 4).However,ifacentrifugalcompressorreceivesmoredriverpowerandincreasesitsspeedandthroughputrapidly,thestationpressureratiowillreactveryslowly tothischange.This isduetofact thatinitially theadditional flowhastopack thepipeline (withitsconsiderablevolume)untilchangesinpressurebecomeapparent.Thus,thedynamicchangeinoperatingconditionswouldlead(inthelimitcaseofaveryfastchangeincompressorspeed)toachangeinflowwithoutachangeinhead(Figure 5).

Figure5:Typicaloperatingpointsiftransientconditionsareconsidered,inthiscaseduetotheshutdownofoneunitinatwo-unitstation.

Becausethefailureorunavailabilityofcompressionunitscancausesignificantlossinrevenue,theinstallationofstandbyunitsmustbeconsidered.Thesestandbyunitscanbearrangedsuchthateachcompressionstationhasonestandbyunit,thatonlysomestationshaveastandbyunit,orthatthestandbyfunctioniscoveredbyoversizingthedriversforallstations.(Oversizingnaturallycreatesanefficiencydisadvantageduringnormaloperation,whentheunitswouldoperateinpartload.)Itmustbenotedthatthefailureofacompressionunitdoesnotmeanthattheentirepipelineceasestooperate,butratherthattheflowcapacityofthepipelineisreduced.Sincepipelineshavea

significant inherent storage capability(line pack), a failure of one ormore units does nothave an immediateimpact on the totalthroughput.Additionally,plannedshutdownsduetomaintenancecanbeplannedduringtimeswhenlowercapacitiesarerequired.

Standbyunitsarenotalwaysmandatorybecausemoderngas-turbine-drivencompressorsetscanachieveanavailabilityof97%andhigher.Astationwithtwooperatingunitsandonestandbyunitthushasastationavailabilityof100 (becausetwounitshavetofailatthesametimeinordertoreducethestationthroughputto50%).Astationwithonestandbyunitandoneoperatingunitalsoyieldsa99.91%stationavailability.However,whilefailureoftwounitsinthefirstcasestillleavesthestationwith50%capacity,theentirestationislostifbothunitsfailinthesecondcase.Arguably,installingtwosmaller50%unitsratherthanonelarger100%unitcouldavoidtheneedforinstallingastandbyunit.

Ithasoftenbeenassumedthatfortwo-unitstationswithoutastandbyunit,aparallelinstallationofthetwounitswouldyieldthebestbehaviorifoneunitfails.However,OhanianandKurz[3]haveshownthataseriesarrangementofidenticalcompressorsetscanyieldalowerdeficiencyinflowthanaparallelinstallation.Thisisduetothefactthatpipelinehydraulicsdictatearelationshipbetweentheflowthrough thepipeline and the necessarypressure ratio at the compressor station. For parallel units,the failure ofone unit forces theremainingunittooperateatornearchoke,withaverylowefficiency.Identicalunitsinseries,uponthefailureofoneunit,wouldinitiallyrequirethesurgevalvetoopen,buttheremainingunitwouldsoonbeabletooperateatagoodefficiency,thusmaintainingahigherflow

thanintheparallelscenario.Giventhefactthatthegasstoredinthepipelinewillhelptomaintaintheflowtotheusers,aseriesinstallationwouldoftenallowforsufficienttimetoresolvetheproblem.

Operatingmultipleunits(eitherinparallelorinseries)canbeoptimizedonthestationorunitlevelbyloadsharingcontrols.Iftheunitsare fairly similar inefficiencyandsize,controlschemes that shareloadsuchthatbothunitsoperateat thesamesurgemarginof thecompressorscanbeadvantageousandwillusuallyresultinagoodoverallefficiency.Similar(oridentical)unitsingeneralachievethelowestoverallfuelconsumptionifbothareaboutevenlyloaded.Thefuelsavingsfromrunningoneunitathighload(andthushighergasturbineefficiency)ismorethancompensatedbyrunningtheotherunitatlowloadandlowerefficiency.Inotherwords,runningtwounitsat75%loadresultsinloweroverallfuelconsumptionthanrunningoneunitat100%andoneunitat50%load.

Ontheotherhand,iftheunitsaredissimilarinsize,orofverydifferentefficiency,itmaybebestifthe larger,orthemoreefficient,unitcarriesthebaseload,whilethesmaller,orthelessefficient,unitisresponsibletoprovidepowerforloadswings.

Pipelineswithloadswings(Figure6)canoftenbenefitfromusingmultiplesmallerunitsasopposedtosinglebigunits.

Figure6:AveragedloadvariationforfourstationsofaninterstatepipelineinSouthAmericaduringsummerand

winterscenarios.




5/18

Whileanalyzingtheperformanceoftheentirepackageitisimportanttounderstandhowtodistinguishtheunitwiththebestoverallefficiency. The turbocompressor performance depends on efficiency of two main componentsgas turbine and the centrifugalcompressor.Thebestefficiencycompressordoesnotalwaysprovidetheoveralllowestunitfuelconsumptionsastheveryimportantpieceoftheequationistheturbineefficiency.Also,therelationshipbetweenthecompressorrunningspeedandthepowerturbineoptimumspeedattherequiredgivencompressorloadmustbeconsidered.Themaingoalduringthecompressorselectionsprocessistofindthestageselectionsthatnotonlyprovidesthehighestcompressorefficiencybutalsowouldyieldthehighestoverallpackageefficiency,whichisachieved,whenthefuelconsumptionforthedutyisminimized.AsweknowfromdiagraminFigure 7 thelowertheturbinepartloadis-thelowertheturbineefficiencywillbe.Assuch,thehigherloadedturbineshouldgenerallyprovidethebetterturbineandoverallpackageefficiency.

Figure7:TypicalChangeofEfficiencywithpartloadfor3differentindustrialgasturbines.

Therelationshipbetweencompressorefficiencyandfuelconsumptioncanmathematicallybederivedasfollows.

For a given operation requirement, defined by standard flow, suction pressure,suction temperature,gas composition, and dischargetemperature,thisoperationrequirementpreciselydefinestheisentropichead andthemassflow .

Thepowerconsumptionofthecompressor thenonlydependsonitsisentropicefficiency andthemechanicallosses ,as

follows:

In otherwords, for a given duty, the compressorwith thebetter efficiency will always yield a lower power consumption, since themechanicallossestendtobeverysimilarfordifferentcompressors.

isthepowerthatmustbeproducedbythegasturbine.Inturn,thegasturbineefficiencyisafunctionoftherelativeload,thatis,

whatpercentageofgasturbinefullpower availableatthegivenambientconditionsandtherequiredcompressorspeedisrequired

bythecentrifugalcompressor(KurzandOhanian[6]):

Therelationshipbetweengasturbinepower,fuelflowFF,heatrateHR,andgasturbineefficiency isasfollows:

Forthecompressorapplicationsthismeansthat,forabettercompressor,thereducedpowerconsumptionindeedcausesasmallincreaseingasturbineheatrateorreductioningasturbineefficiency.However,theresultoflowerloadandhigherheatrateisalwaysalowerfuelconsumption.

Otherimportantissuesmustbeconsideredwhenanalyzingoverallpackageefficiency.Inmanyinstancesithasbeenrequestedbytheenduserto use thegasturbineheat rate asan indication oftheoverall packageperformanceefficiencies and asthebasis for thepackageguarantees.Higherturbineloadwillcorrelatewithahigherturbineefficiency.Thisrelationshipiscorrecthowever;theturbocompressorusershouldrecognizesomeunderlyingcircumstancesthatcanleadtothewrongconclusion.

Thepeculiarthingisthatoperatingatlowercompressorefficiencythecentrifugalcompressorwillrequiremorecompressionpowertodotheirdutythatwillultimatelyincreasetheturbineload.Ifweaccountthathigherpartloadwillleadtothebetterengineefficiencywediscoverthatcompressorwithlowerefficiencywillforceturbineoperateatbetterheatrate!ThisisthefactthatisbeingoverlookedinmanyinstancesifthecomparisonbetweenthevendorsisdonesolelybasedondriversHeatRate.Thequestionishowtoavoidthistrap.Thesimpleandthestraightforwardansweristhatoverallunitcomparisonshouldbedonenotonturbinesheatrate,whichissimplyjustanumber,butratherondirectvalue-turbinefuelconsumption.Inthiscasethelowercompressorefficiencywilldrivehigherturbinepower,and,despitethefactthattheturbinesheatrateandefficiencywillbeimproving,theactualturbinesfuelconsumptionwillbegoingup.

Intheend,theonlymeasurethatistakenintoaccountwhencalculatingpackageoveralloperatingcost(OPEX)isthefuelconsumptionwhereastheturbinesheatrateortheturbinessimplecycleefficiencyremainsonlynumbersonthepaper(Figure 8).

Figure8:ImpactofcompressorefficiencyonGTLoad,heatrate(HR),andfuelconsumption(FF).

9.PipelineSizingConsiderations

Kurzetal.[7]evaluateddifferentoptionsforpipediameters,pressureratings,andstationspacingforalongdistancepipeline.A3220km(2000mile)onshoreburiedgastransmissionpipelinefortransportingnaturalgaswithagravityof0.6wasassumed(Figure 9).




6/18

Figure9:Optimumnumberofstationsandoptimummaximumoperatingpressure(MAOP)forthe3220 km(2000

mile),560000 Nm3/hsamplepipeline.ThelowestcostconfigurationsforeachMAOPsolutionaremarked(from[ 7]).

Assumingthatpipeswillbe availablein 2-diameter incrementsfrom pipemills, thenearest evenincrementsof theabove-mentionedtheoreticaldiameterswereselected(24 ,28,and34 for152bar,103bar,and69bar(2200,1500,1000psia)pressures,resp.)andanalyzed

byvaryingthenumberofstationsalongthepipeline.TheresultofthisrefinementisshowninFigure 9,wherepresentvalueisplottedagainstnumberofstationsforeachpressurelevel.Theminimaforeachisshowninthechartwithpresentvaluetotalhorsepower,andnumber of stations. Inthis study,the 69bar (1000psia)pipeline hasthe lowestpresent value thuswouldbe themost cost-effectivesolution.

Inactualpractice,forcommonalityreasons,identicalsizeunitswillbeinstalledinthestations.Inordertohaveidenticalpowerateachstation,thestationspacingwillbeadjusted(dependentonthegeography)sincethestationsatthebeginningofthelinewillconsumemorepowerthanthestationsattheendofthelineduetothepowerrequiredforfuelcompression.Identicalpowerateachlocationalsodependsonsiteelevationanddesignambienttemperature,whichwoulddefinethesiteavailablepowerfromacertainengine.

Oneofthekeyfindingsisthattheoptimumisrelativelyflatinallcases.Thismeansinparticularthatcertainconsiderationsmayfavorlarger stationspacing,withhigher stationpressureratios and higher MAOP insituationswherepipelinesare routed through largelyunpopulatedareas.

10.TypicalApplication

Foracasestudyweconsideraninternationallongdistancepipeline.Thetotallengthofthelineisabout7000 km.Thepipelineconsistsoftwo42 parallellineswhichturnintosinglea48 lineatthecrossingofaninternationalborder.Thepipelinedesignthroughputis30billion

Nm3peryearandmaximumoperatingpressureofthispipelineis9.8 MPa.Thereare10compressorstationsplannedinoneareaandover

20stationsinthereceivingcountry.Afterfirstgas,ittakes5yearstobuilduptofullcapacity.

Whenwecompareoperationsofthecompressorstationweneedtorecognizetwomainapproaches.Wecaneitheroperatewithfeweroflargerturbocompressorunits(CaseA,2largeunits)orwithahigherquantityofsmallerturbocompressorunits(CaseB,4smallunits).Thefollowingfactorsneedtobeconsideredwhenselectionofeitheroptionisdecided.Inevaluatingthesystemreliabilityandmaximumthroughputtheimpactofunitoutagesneedstobeconsidered.Ifweweretoconsidertwolarge30 MWunitsthefailureofoneofthemwillresultin50%reductionofpoweravailablewhereasifweconsider4smaller15MWunitsthefailureofoneofthemwillresultinonly25%powerreduction.Figure10 outlinesthebasicfact that,if thesurvivingunits runatfull loadtomakeup asmuchflowaspossible,theoperatingpointfortheCaseBwillbeclosetothehighestefficiencyislandsotheremainingon-linecompressorswillbeworkingmoreeffectively comparedto CaseAwhenthe single remaining largeunitwillbeworkingin the stonewallarea. Itis obviousthatpipelinerecoverytimewillbeshorterinCaseB.

Figure10:Impactoflossofoneunitforthe4unitandthe2unitscenarios.

BasedonananalysisbySantosetal.[ 8,9],CaseAcanrepresentevenmoreproblems.Theamountofgasthatthesingleremaining30 MWunitwillhavetoprocessissobigthatitwillputthisremainingunitintochoke,andthusforpracticalpurposesoutofoperation.Theamountoffuelthattheremainingunitis goingtoburnwillnotjustify thatnegligible increasein headthatthisunitwillprovide.So,practically,whenone larger turbocompressor will beoutof operation,the secondwill have to be shut down and thestationwill bebypassed.Stationconfigurationswiththesingleoversizedriverandeithernostandbyorstandbyoneachsecondorthirdstationareoftenadvocated.Theargumentsinfavorofthismethodareveryhighpipelineavailability(99.5%)andhighefficiency(4042%)ofthelarger30MWturbocompressorunits.Infact,designingforaturbineoversizedby15%willleadtonormaloperationsatpartloadconditionsalmostallthetime(99.5%)wheretherewillbenegativeimpactonturbineefficiencyand,asaresultofit,increasedfuelconsumptionAnother

negative impact of this approach is that normally this pipeline would operate at lower than MAOP pressure, whereas the highestoperationalpipelinepressureproduceslesspressurelossesand,therefore,lowerrequirementsfortherecompressionpower.Thereasonforthat isthemaintenancescheduleforthe turbocompressorson the stationswith the single unitswithout standby. Inorder toperformmaintenanceon theseunitsthe pipeline,linepackwillhaveto bemaximized uptoMAOP,so thatunitcanbetakenofflineandthepipelinethroughputwillnot beimpacted. Therefore, thenormalpipelineoperationshave tobe basedon alowerMAOP.Alsoworthmentioning is thepipeline capacitywhenconsidering the single turbocompressorapproach.Manypipelines transport gasownedandproducedbydifferentcommercialentities.Assuchthegasfieldsdevelopmenttimeandgasavailabilitydependonmanytechnologicaland,lately, political factors thatmay potentiallyhavenegative impacton pipeline predictedcapacity growth. In theseconditionsthe singleoversizedturbocompressorwilleitherbeworkingintodeeprecyclingmodeuntiltheexpectedamountofgaswillbecomeavailableorstartoperationwithsmallercapacitycompressorstageswhichwillsubsequentlyrequireacostlychangeoftheinternalbundle.

11.FuelComparison

Itisincreasinglyimportanttoevaluateallseasonsconditionswhenmakingacomparisonbetweentwodifferentstationlayoutcases.Forthesubjectpipelinedifferentdesignorganizationswereinvolvedinthepipelinefeasibilitystudy.Oneofthemhasusedsummerconditions

onlyandcametotheconclusionthatlargerturbinesarepreferredoption.Anothersourceusedannualaverageconditionsandcametotheoppositeconclusion.Thereasonforthatwasthefactthatduringwinter,fall,andspringmonths,whichcovertotalof9outof12monthsofoperation,oneofthesmallerturbocompressorswasputinthestandbymode.Duetolowerambienttemperaturetheamountofpoweravailablefromtheremaining3unitswasenoughtocoverthe100%dutiesduetohighcompressorefficiency.ThiswasnottrueforCase1(basedonsameexplanationabove)andboth30 MWunitshadtoworkinthedeeppartloadwithunsatisfactoryturbineefficiency.Thefact




7/18

thatoperationalmodebecame3+2 forCaseBgaveadditionalbenefitsworthmentioning.Sincetwoturbocompressorswereinstandbymodetherewasanopportunitytodoallmaintenanceworkduringthistimeoftheyear.ItmeansthatavailabilityofthissystembecomessuperiorcomparedtoCaseA,especiallyifweweretoconsidersummermonthsofoperations.

12.MaintenanceandOverhauls

Anotheradvantage ofoperating only3outof5 unitsfora significantpartof the year (i.e.,9 outof12months)isthe extendedtimebetweenoverhauls.Basedonthecalculationsbelow,thetotalnumberofhoursforeachturbocompressorunitperyearwasreducedfrom7008to5694,and,therefore,thetimebetweenoverhaulscouldbeextended.Basedon unitsoperatingduring3summermonthsand3

+2unitsoperatingduringtherestoftheyear(9months),iftheunitswereusedsothattheyallranexactlythesamenumberofhourseachyear,eachunitwouldrunfor5,694hourseveryyear.Whereasifweaccountfor4workingunitswithonestandbythroughouttheyearthenumberofworkinghourswillbeasfollows:8,7604unitsrunning=35,040/5unitsavailable=7,008totalhoursperunit/year.

Notethatallunitsrunforanequalnumberofhourstomakethecalculationsimple.However,thecustomercouldpushleadmachinestoreachtheagreedtimebetweenmajorinspections(TBI)first,sothatallenginesdonotcomeupforoverhaulatthesametime,andthiswouldhelpwiththeoverhaulcost,helpingtodistributetheoverhaulcostoverthe30yearcycle.Wecanevenmakestepfurtherandwillseeadditionalbenefitsofthisapproach.Accountingforthenormalyeararoundoperationwith4unitsonline,eachturbocompressorwillget7,00830year=210,240requiredhoursofoperations,whereasconsidering3+2setupfor9monthsthetotalnumberoftherequiredhoursofoperationsreduceddownto170,820hours.Withmodernturbinestechnologyitisnotuncommontoseethatlifetimeoperationreaches150180,000hours.Itmeansthatforthelifetimeofthisproject(30years)therewillbenoneedtobuynewsetofequipment.Thisalonemakeshugefavorableimpactonprojectseconomics.

13.StationversusSystemAvailability

Itisimportanttorecognizethedifferencebetweenstationandpipelineavailability.Foreconomicassessments,misunderstandingthisissuecanleadtothewrongconclusion.Stationavailabilitycalculationsareeasy,straightforwardandbasedonsimplestatisticalequations.Itiseasytoseethatfewerunitsonacompressorstationwillyieldhigheravailability,assumingthethresholdforavailabilityis100%oftheflow.But is this true for the entire pipeline system? The answer is not easy and requires additional investigation including extensivehydrodynamic analysis usingof thestatisticalmethodology.TheMonteCarlomethod[9]hasprovedtobethegoodmethodologytodeterminethepipelinesystemavailability.Thestatisticalportionconsistsofgeneratingmultiplerandomcasesofequipmentfailureonsingleortwoconsecutivecompressorstations.Thehydrodynamicportionwillcalculatethemaximumthroughputthatpipelineisavailabletocarrywhenthesefailuresoccur.Basedonthisextensiveandin-depthanalysisitcanbeshownthatavailabilityofapipeline,configuredwithsmallermultipleunits,deliversbetteroverallresults.Themainreasonforthatoutcomeisthefactthatshutdownofthesmallerunitmakeslesserimpactonthebehavioroftheentirepipeline.Ofcourse,tohavefairresults,theavailabilityofthesingleturbocompressorunit,eithersmaller15 MWorlargersize30 MW,wasidentical.It iseasyto understandthatinourparticularcasetheavailabilityof thestationsetupwithsmallerunits(CaseB)wasgreatlyenhancedbecauseofthepresenceofextrastandbyunitduringwinterandfall/springmonthswhenstationssetuphas3+2configuration.

14.EffectofLargeUnitShutdown

Examplesofthevulnerabilityaredemonstratedbasedona typicalpipelinescenariowith4stations.Eachstationhas2compressortrainswithout spares. If one unit in station2 is lost, thepipeline flow is reduced by 12%. However, thesame 12% flow reduction can bemaintainedbyalsoshuttingdownthesurvivingunitinstation2.Thisisduetothenecessarilyinefficientoperationofthesurvivingunitinstation2,whichisforcedtooperateinchoke.Ifbothunitsareshutdown,stations3and4willbeabletorecovertheflow,butatamuchhigheroverallefficiency.Thus,shuttingbothunitsdownreducesthepipelinefuelconsumptioncomparedtothescenariowithonlyoneunitshutdowninstation2.Thepointofthisexampleis,thefailureofoneoftwolargeunitsinacompressorstationhasmoresignificantconsequencesthanthefailureofasmallerunitinastationwiththreeormoreoperatingunits.Or,inotherwords,scenarioswith3oremoreunitsperstationwithoutspareunitstendtohaveahigherflowifoneoftheunitsfailsorhastobeshutdownformaintenance,thanscenarioswith2unitsperstationwithoutspareunits.

15.Conclusion

Thepaperhasillustratedthedifferentinfluencefactorsfortheeconomicsuccessofagascompressionoperation.Importantcriteriaincludefirstcost,operatingcost(especiallyfuelcost),capacity,availability,lifecyclecost,andemissions.Decisionsaboutthelayoutofcompressorstations such as the number of units, standby requirements, type of driver and type of compressors have an impact on cost, fuel

consumption,operationalflexibility,emissions,aswellasavailabilityofthestation.

References

1. J.H.LottonandM.Lubomirsky,GasTurbineDriverSizingandRelatedConsiderations,BratislavaPublishers,Slovakia,2004.

2. S.P.Santos,Transientanalysis- amustingaspipelinedesign,inProceedingsofthePipelineSimulationInterestGroupConference ,Tucson,Ariz,USA,1997.

3. S.OhanianandR.Kurz,Seriesofparallelarrangementinatwo-unitcompressorstation,JournalofEngineeringforGasTurbinesandPower,vol.124,no.4,pp.936941,2002. ViewatPublisher ViewatGoogleScholar ViewatScopus

4. W.Wright, M. Somani,andC.Ditzel, Compressor station optimization, inProceedingsof the30th Annual Meeting,PipelineSimulationInterestGroup,Denver,Colo,USA,1998.

5. M. Pinelli, A.Mazzi, and G.Russo, Arrangement and optimization or turbocompressors in off-shore natural gas extractions

station,Tech.Rep.GT20005-68031,ASME,2005.6. R.KurzandS.Ohanian,Modellingturbomachineryinpipelinesimulations,inProceedingsofthe35thAnnualMeeting,Pipeline

SimulationInterestGroup,Berne,Switzerland,2003.

7. R.Kurz,S.Ohanian,andK.Brun,Compressorsinhighpressurepipelineapplications,Tech.Rep.GT2010-22018,ASME,2010.




8/18

8. S.P.DosSantos,M.A.S.Bittencourt,andL.D.Vasconcellos,Compressorstationavailabilitymanagingitseffectsongaspipelineoperation,inProceedingsoftheBiennialInternationalPipelineConference(IPC'07),vol.1,pp.855863,2007.

9. S.P.Santos,MonteCarlosimulationakeyforafeasiblepipelinedesign,in ProceedingsofthePipelineSimulationInterestGroupConference,Galveston,Tex,USA,2009




9/18

Figure1:Typicalcompressorstationwith3gasturbinedrivencentrifugalcompressors.

Page 1 of 1Gas Compressor Station Economic Optimization : Figure 1

05/01/2013http://www.hindawi.com/journals/ijrm/2012/715017/fig1/


10/18

Figure2:Systemcharacteristicsandcompressormap.




11/18

Figure3:Operatingpointanddriverpower.




12/18

Figure4:Typicalsteadystatepipelineoperatingpointsplottedintoacompressorperformancemap.




13/18


14/18

Figure6:AveragedloadvariationforfourstationsofaninterstatepipelineinSouthAmericaduringsummerandwinterscenarios.




15/18

Figure7:TypicalChangeofEfficiencywithpartloadfor3differentindustrialgasturbines.




16/18

Figure8:ImpactofcompressorefficiencyonGTLoad,heatrate(HR),andfuelconsumption(FF).




17/18

Figure9:Optimumnumberofstationsandoptimummaximumoperatingpressure(MAOP)forthe3220 km

(2000mile), 560000Nm3/h sample pipeline. The lowest cost configurations for each MAOP solution are

marked(from[7]).




18/18

Figure10:Impactoflossofoneunitforthe4unitandthe2unitscenarios.


Gas Compressor Station Economic Optimization

Documents