• / 13
  • 下载费用:5 下载币  

papers_ETLP6_Saturation_Alvarez-MacBeth-Brain_JPG_2016

关 键 词:
papers_ETLP6_Saturation_Alvarez MacBeth Brain_JPG_2016
资源描述:
Quantifyingremainingoilsaturationusingtime-lapseseismicamplitudechangesatfluidcontactsErickAlvarez1,2*,ColinMacBeth1&JonathanBrain21InstituteofPetroleumEngineering,Heriot-WattUniversity,EdinburghEH144AS,UK2ShellUKLimited,1AltensFarmRoad,Nigg,AberdeenAB123FY,UK*Correspondence:erick.alvarez@shell.comAbstract:Ourstudyshowsthattime-lapsechangesintheamplitudeoftheseismicreflectionatanoil–watercontact(OWC)and/orproducedOWCcanbeusedtoestimatedirectlythedisplacementefficiencyofwaterdisplacingoil,ED,withouttheneedofarockandfluidphysicsmodel.Fromthisvalue,itispossibletodeterminetheremainingoilsaturationifrequired.Apreliminaryapplicationisperformedusingseveralpublishedliteratureexamples,whicharereinterpretedtoassesstheaverageEDandensurethatthetheoryisconsistentwithexpectations.Next,aNorthSeafieldmodelwithaknownEDisusedtocreatefluid-flowpredictionsandthecorrespondingsynthetictime-lapseseismicdata.Applicationtothesedataagainconfirmsthebasicprinciplesofthemethodanddefinestheaccuracywhenappliedto4Dseismicdata.Finally,anobserved4DseismicdatasetfromaproducingfieldintheNorthSeaisanalysed.Theresultssuggestadisplacementefficiencyofbetween21and65%withanaccuracyof3%duetodatanon-repeatability(withaNRMSofbetween11and13%).Givenanaverageirreduciblewatersaturationof0.32,thiscalculatestheremainingoilsaturationsatbetween24and53%forthisfield.AprerequisiteforuseoftheproposedOWCapproachisthatadiscretecontactbeinterpretedoneitherthe3Dor4Dseismicdatasets.Therefore,successfulapplicationofthistechniquerequiresmoderate-tohigh-qualityseismicdataandafairlythickreservoirsequencewithoutsignificantstructuralcomplexity.Received25February2016;revised31August2016;accepted1September2016Inthiswork,weconsidertheevaluationoftheoilsaturationleftbehindinahydrocarbonreservoirafteroilisdisplacedbywaterasaconsequenceofnaturaldriveoilproduction,orproductionsupportedbywaterinjectionintheoilleg.Thisisanimportantobjectiveforreservoirmonitoringandsurveillance,andunderstandablymuchefforthasgoneintothestudyofthistopicinthepublishedreservoirengineeringliteratureoverthepast50years(Al-Harbi2011).Remainingoilsaturation(ROS)valuesimpactourunderstandingofthereservoir’srecoveryfactor–akeyparameterthatdefinesthestrategyforproductionoptimization.(Herewemustdistinguishbetweenresidualoilsaturationandremainingoilsaturation.Residualrelatestotheendpointsoftherelativepermeabilitycurves,whereasremainingincludesalloiltrappedinthereservoirheterogeneityatavarietyofscales.)ThedistributionofROSdeterminesthesweepefficiencyofthewaterdisplacementduetoeitheredgeorbottomwatermechanisms(Dake2001).Moregenerally,itidentifiespossiblezonesofunswept,bypassedoilforfuturedrilling,andalsotheamountofoilleftbehindinthesweptzoneafterproduction.TheexactlocationandamountofROSneedstobeunderstoodinordertoassessthefeasibilityofenhancedoilrecovery(EOR)methods(suchasmisciblegasinjection,wateralternatinggas,polymerfloodingandsurfactants)tofurthermobilizeandextracttheremainingoil.Fromaproductionperspective,knowledgeofROSisimportantformaterialbalance,calculatingproductionforecastsandisanessentialinputforthenumericalsimulationmodel.Therearemanyfactors,actingatthemicroscopic(porescale)andmesoscopic(outcrop)ormacroscopic(reservoir/field)scale,thataffectthemagnitudeofROS.Atthesmallestscale,residualoiltrappedwithintheporespaceisimmobilizedbycapillaryforcesunderflowconditions(Fig.1a).Thisresidualoilsaturation(Sorw)definesoneoftheend-pointsoftherelativepermeabilityofoiltowater,andisthepointatwhichoilnolongerflowsandoilistrappedbytheinvadingwater.Thisbehaviourdependsonarangeofrock-andfluid-dependentfactors,suchasmineralogy,pore-sizedistributions,texture,geometry,roughness,claycontent,andoilandformationwatercomposition.Atthispore-scale,thewettingbehaviouroftherockshasalsobeenidentifiedtobeofparticularimportance.Forexample,researchbySkauge&Ottesen(2002)hasshownthatSorwincore-floodingexperimentsdecreasessignificantlywhenwemovefromawater-wettoanoil-wetreservoir(e.g.Sorwmovesfrom0.45to0.04).ROSvaluesatthemesoscopicormacroscopicscaleareprincipallycontrolledbythegeologicalheterogeneity(Fig.1b),andhencearestronglyrelatedtothedepositionalenvironment.Clearly,ROSatthemesoscopic/macroscopicscaleisgreaterthanpore-scaleSorwinafullysweptregionofthereservoir.ImportantfactorscontrollingthevalueofROSatmesoscopic/macroscopicscaleincludethedegreeofdiagenesis,cementationandtrappingmechan-isms,suchaslaminaeorcross-bedding(Pickup&Hern2002).Faults,fracturesandstratigraphicbarriersalsorequireconsideration.Finally,atthefield(macro)-scale(Fig.2),zonesofunswept(bypassed)oilcanbeidentified.Unsweptzonesmayexistduetoadiversionofthewater-displacementprocess,suchasfingeringandinterbeddingordiversionatabarrierorfaultedstructure.Suchbypassedareasarestillintheinitialoilsaturationstateandrepresentopportunitiesforimprovedoilrecovery(IOR)throughinfilldrilling,betterprojectmanagementorbyachangeinthedepletionstrategy,suchasproductionratereductions.Suchbypassedareashavebeensuccessfullyidentifiedfromconventionalapplicationoftechniquessuchas4Dseismic(seelater).Importantly,withinthesweptzone,thereisoilnotrecoveredbythewater-displacementmechanismandthisrepresentsapotentialEORtarget,asdiscussedabove.Thefocusofthiscurrentstudyistoextendthe4Dseismictechniquewiththegoalofidentifyingremainingoilsaturationsinapracticalmannerthatdoesnotrequireadetailedrockandfluidphysicsstudy.MeasurementofROSfromengineeringpracticeCurrentmethodstomeasureROSrangefromlaboratoryexperimenttoopenhole,andrepeatedcased-holewirelineorproductionlogging.Tekluetal.(2013)providesareviewofthesemethods.©2017TheAuthor(s).PublishedbyTheGeologicalSocietyofLondonforGSLandEAGE.Allrightsreserved.Forpermissions:http://www.geolsoc.org.uk/permissions.Publishingdisclaimer:www.geolsoc.org.uk/pub_ethicsResearcharticlePetroleumGeosciencePublishedonlineNovember1,2016https://doi.org/10.1144/petgeo2016-037|Vol.23|2017|pp.238–250Laboratorycore-floodingtestsarecommonlyused:wherearocksampleintheshapeofacylinderisextractedfromthereservoir,saturatedwithoilandthenthisisdisplacedbywaterorgas.Othertypesofmeasurementincludecentrifugeexperiments,whichestablishthetrueresidualoilsaturationfromthelowestachievablecapillarypressurewhennofurtheroildisplacementispossibleundercapillarydominatedflowconditions.Wellmeasurementscanalsobemadeusingtoolsthatmeasureoilsaturationinreservoirzonesthatproduceonlywater.Howeverthereissomeuncertaintyassociatedwiththesemeasurements.Forexample,resistivitytechniquesaredifficulttointerpretinamixedsalinity(formationandinjectedwater)environment,orinmixedsand-shalelithologies.Singlewellchemicaltracertestscanalsoprovideawayofestimatingtheconcentrationsofinjectedwater,andhenceresidualoilvalues.Another,morerecentdevelopment,istheuseofporenetworkmodels,calibratedfromdigitalCTimagestonumericallycomputetheflowphysicsandhencerelativepermeabilities(Ryazanov2012).DespitearangeofmethodsavailabletomeasureROS,itisobservedthatagenerallyacceptedwaytopredictROSacrossthefieldandbetweenthewellsdoesnotexist,asitisdifficulttocapturethefieldvariationswithmeasurementsbasedatthewellbore(Pathaketal.2012).Nosinglemeasurementachievesadefinitiveresult,andacombinationinanintegratedstudyprovidesthemostsatisfactorywayofassessingtheoilinplace(Pathaketal.2012).Atthefieldscale,awholereservoiraveragemeasurementcanbeprovidedbymaterialbalanceandproductionanalysis(Sharma&Anil1996).Finally,numericalsimulationprovidesthebestcurrentapproachtocapturearealandverticalvariations,andthefieldproduction-scalevalues(forexample,Raoetal.2013).HowevermostapproachesunderestimateROSastheydonotadequatelycaptureheterogeneityanduseconservativeinputvaluesofROS(Valentietal.2002).Overall,valuesofROShavebeenquotedaslowas0.04andashighas0.45,forbothcarbonateandclasticreservoirs(SkaugeVrachliotis2012).ROSdeterminedfromseismicdataTheuseof3Dseismicdataforreservoirmanagementiswellestablished,throughtheinterpretationofreservoirstructure(faultsandmajorboundingsurfacesofthereservoirflowunits)andfluidcontacts,whichinputdirectlyintothedeterminationofinitialoilvolumesinthereservoironcesaturationsandporositieshavebeenassigned.Themonitoringofreservoirsundergoingproductionandrecoveryusingrepeatedseismicsurveying(the4D,ortime-lapse,seismicmethod)isnowalsofirmlyestablishedbynumerouscasestudiesovermorethanadecade(Jack1998;Johnston2013).Inparticular,ithasbeendemonstratedthatwater-flooding,andassociatedcontactmovements,areoftenvisibleinmost4Dseismicdatadespitetheoverlappingeffectsofpressure,temperature,salinity,geomechanicaleffectsandacquisitionnon-repeatability.Indeed,watersaturationchangeshavebeendetectedinFig.1.(a)Exampleofpore-scaletrappingofoilduetoviscous-dominated(fast)waterdisplacementofoil.Averagewatersaturationmovesfromconnatewatertopredominantlywaterplusresidualoilsaturation.(b)Exampleofoutcrop-scaletrappingofoilinindividualbedlaminaeafterawaterflood(afterPickup&Hern2002).(Saturationvaluesarecolourcodedinpdfasfollows:green/yellow,oil;lightanddarkblue,water).Fig.2.Illustrationofthefield-scaleremainingoilsaturationtargetedinourcurrentstudy.Thewater-flood(inblueonpdf)progressesfromtwoinjectors(I1andI2)inthenorthernandsoutherndowndipareasofthecrestalstructure,andbypassessomeofthereservoiroil(zoneA).Thetargetforourstudiesisnotthisunsweptoil,buttheoilremaininginthezonesalreadysweptbywater(zoneB).(FigureadaptedafterCalvert2005.)239Quantifyingremainingoilsaturationacquisitionnon-repeatabilityconditions,rangingfromexcellenttopoor(anormalizedrootmeansquaremetricofbetween5and40%,respectively),bothinonshoreandoffshoredata,andforclastic(Marshetal.2003),chalk(Byerleyetal.2006)andhardrockcarbonatereservoirs(Al-Jenaibietal.2006).Theprincipaluseofsuchinformationtodatehasbeentoidentifyeconomicinfilltargetscreatedbybypassedpocketsofmobileoilduetounexpectedwaterflow(Kosteretal.2000)orcompartmentalization(Staplesetal.2006).Otherusesof4Dseismicdataincludeevaluationofinjectorperformanceandsubsequentintervention(Røsteetal.2009),andsimulationmodelupdating(Gonzalez-Carballoetal.2006;Oliveira2008).LesscommonlymentionedistheneedtoestimateanaccuratevalueofROSwithinthealreadywater-sweptzones(Johnston2013;AlvarezObiwulu&MacBeth2015)thatimpactstheeconomicalevaluationofthedisplacementprocessandunderpinsfutureEORstrategies.Athoroughquantitativeanalysisofthe4Dseismicdatamustdrawalinkbetweenthemagnitudeofthe4Dseismicsignatureandtheremainingoilsaturations,andisthetopicofthisstudy.The4Dseismicsignatureofanoil–watercontactmovementduringproductionInthiswork,weattempttoestimateremainingoilsaturationbymonitoringtheoriginalandproducedOWCsfromoilreservoirsdepletingwithnaturalbasalaquiferdriveorwithsupportfromwaterinjectors.WeexaminethesteadyupwardsriseofahorizontalOWCthroughbottomwaterinflux,anddeterminethemagnitudeofthefluidcontactreflectivitybeforeandafterproductionusing4Dseismicdata.Thetechniqueissuitedtoreservoirsinwhichthepre-andpost-productionfluidcontactsarevisibleontheseismicdata,andthesecontactsexhibitadetectableupwardsmovement.SelectedfieldexamplesofsuchreservoirsinthepublishedliteratureincludeGannettC(Kloostermanetal.2001),Nelson(McInallyetal.2001;MacBethetal.2005)andAlba(Turaetal.2009),allofthemfromtheNorthSea.Seismicvisibilityofthecontactsrequiresareasonablythickreservoirsequence,withgoodverticalconnectivityandamechanismforbasalwaterdrive.TheprocessofbottomwaterinfluxissimplifiedforourcalculationsasshowninFigure3a,b.Here,weneglectthethicknessofthetransitionzonebetweentheoilleg(withirreduciblewatersaturation)andthefreewaterlevel(at100%watersaturation)relativetotheverticalmovementofthewatercontactsandoverallthicknessofthereservoir.WealsocollapseboththeOWCandfreewaterlevelintoonesinglediscreteinterfacethatwedefinegenericallyasthe‘oil–watercontact’.Oursimple‘seismic’modelthereforeassumesthattheinitialoilzoneissaturatedbywaterattheirreduciblewatersaturationSwirr(thesaturationbelowwhichwatercannotflow),andignoresthefactthattheinitialwatersaturationoftheundisturbedreservoirpriortoproductionmaycontainformationwaterthatmayalsobeproducedwiththeoil.Inthismodel,wealsoneglectthefluidpressurechanges,ascalculationsshowtheseareafactorof10smallercomparedtothesaturationchanges(Alvarez&MacBeth2014).Theproductionstateisrepresentedbythecreationofanadditional‘producedoil–watercontact’.Theassumptionsabovearealljustifiableforseismicanalysispurposesprovidedthattheverticalpermeabilityofthereservoirisrelativelyhigh(>100mD),theAPIgravityoftheoil(API>30°)isalsohighandtheseismicwavelengthislong(largerthanthetransitionzone).However,asinpractice,transitionzonescanvaryfromafewmetrestoseveralhundredmetresdependingonthereservoirandthespecificcapillarypressurecurves(Ahmed2006),theapplicabilityofthismodelmustbeexaminedonacase-by-casebasis.Afinalassumptionforourtheoreticaldevelopmentisthatboththeoriginalandproducedcontactsareplanarandhorizontal,andthereforethattheverticalconnectivityisuniform–thisisrarelythecaseforanactualreservoirowingtothehydrodynamicgradientsandgeologicalheterogeneity.Furthermore,theseismicdatacouldexperienceseismictravel-timevariationsthatwouldpreventthecontactsfromappearingflatintwo-waytime.TheoriginalOWCisvisibleintheseismicdataasareflectioninthebaseline(pre-production)dataduetothefluidpropertycontrastsbetweentheoilleg(oilandirreduciblewater)andthewaterleg(100%watersaturation)(seeFig.3a).Thiscontrastmay,ormaynot,belarge,dependingontheoilgravity–thelightertheoil,thestrongerthecontrast.Duringproduction,astheOWCmovesupwardsitsweepsthroughtheoilcolumn(seeFig.3b).Theoilsaturationremainingafterthisprocesswilldependonthemanyfactorsoutlinedintheprevioussection.Thereisnowaseismiccontrastbetweentheunsweptoilleg(oilandtheirreduciblewatersaturation)andthesweptzone(remainingoil,invadingwaterandirreduciblewatercomponent),andalsobetweenthesweptzoneandthewaterlegwith100%water(MacBethetal.2005).ThesecontrastsleadtoseismicreflectionsattheoriginalandproducedOWC(OOWCandPOWC,respectively),themagnitudesofwhichvaryaccordingtotheremainingoil.Ifthetwo-waytime(TWT)differenceislargeenoughtoseparatetheseindividualreflections,thentheiramplitudecanbeusedasameasureoftheremainingoilsaturationasdescribedbelow.ReflectivitychangesatthefluidcontactsForthemodelandassumptionsdefinedintheprevioussubsection,theP–PwavereflectioncoefficientsfortheOOWCorPOWCdependpredominantlyonthesaturationconditionsaboveandbelowthecontacts,providedthattherearenosignificantfacieschangesacrossthecontactboundary.Thisconditionhasimplicationsforthevariationofthereflectioncoefficient,R,withincidenceangle,θ,ateachfluid–fluidinterfacewithinthesameporousrock.SomeofthesecharacteristicshavebeenexploredbyWright(1986)inaprevioustheoreticaldevelopmentbasedonthefluidsubstitutiontheoryofGassmann(1951)andthereflectioncoefficientapprox-imationsofShuey(1985).ThisworkshowedthattheP–Pamplitudevariationwithθinthecaseoffluid–fluidcontactstakesasimpleform.HeconcludedthatreflectioncoefficientsR(θ)atgas–brine,gas–oiloroil–brinecontacts(thatareallnaturallypositiveatnormalincidence)alwaysincreasemonotonicallywithθ:thatis,becomebrighterandmorepositivewithangle/offset.Tofirst-orderinimpedance,andforsmalldensitycontrasts(:þDSw5f1C0fC18C19rwC0rorminC18C19C27sec2uþSwirrf1C0fC18C19rwC0rorminC18C19(A4)andR
展开阅读全文
  石油文库所有资源均是用户自行上传分享,仅供网友学习交流,未经上传用户书面授权,请勿作他用。
0条评论

还可以输入200字符

暂无评论,赶快抢占沙发吧。

关于本文
本文标题:papers_ETLP6_Saturation_Alvarez-MacBeth-Brain_JPG_2016
链接地址:http://www.oilwenku.com/p-70409.html

当前资源信息

吾王的呆毛

编号: 20180607204146330472

类型: 共享资源

格式: PDF

大小: 4.50MB

上传时间: 2018-06-08

广告招租-6
关于我们 - 网站声明 - 网站地图 - 资源地图 - 友情链接 - 网站客服客服 - 联系我们
copyright@ 2016-2020 石油文库网站版权所有
经营许可证编号:川B2-20120048,ICP备案号:蜀ICP备11026253号-10号
收起
展开