碳酸盐岩石物理问题.docx

碳酸盐岩石物理问题.docx

《碳酸盐岩石物理问题.docx》由会员分享，可在线阅读，更多相关《碳酸盐岩石物理问题.docx（12页珍藏版）》请在冰豆网上搜索。

碳酸盐岩石物理问题

CarbonateRockPhysicsIssues

EnruLiu,MichaelA.Payne,ShiyuXu,GregorBaechle,ChristopherE.Harris/ExxonMobilUpstreamResearchCopyright2009,InternationalPetroleumTechnologyConference

ThispaperwaspreparedforpresentationattheInternationalPetroleumTechnologyConferenceheldinDoha,Qatar,7–9December2009.

ThispaperwasselectedforpresentationbyanIPTCProgrammeCommitteefollowingreviewofinformationcontainedinanabstractsubmittedbytheauthor（s）.Contentsofthepaper,aspresented,havenotbeenreviewedbytheInternationalPetroleumTechnologyConferenceandaresubjecttocorrectionbytheauthor（s）.Thematerial,aspresented,doesnotnecessarilyreflectanypositionoftheInternationalPetroleumTechnologyConference,itsofficers,ormembers.PaperspresentedatIPTCaresubjecttopublicationreviewbySponsorSocietyCommitteesofIPTC.Electronicreproduction,distribution,orstorageofanypartofthispaperforcommercialpurposeswithoutthewrittenconsentoftheInternationalPetroleumTechnologyConferenceisprhoehribitend.bPyermissiotntorepperroduacseirnprintisresrtirictLeidbtoriaanaIbstra,ct.of.notxm8ore3th3a,n3i0c0awrordos;,illuXstr5ation-sm83a6y,no.t.be,cfoapxie+d.The-9abstra4ctm.ustcontainconspicuousacknowledgmentofAbstract

Summary

Limestoneanddolomitereservoirsaccountforapproximately50%ofoilandgasproductionworldwide,yetseismicresponsesincarbonaterocksarepoorlyunderstood.Developmentofacarbonaterockphysicsmodelisextremelydifficultbecauseporesystemsaremorecomplexincarbonatesthantheyareinclastics.Carbonatescanhaveavarietyofporetypes,suchasmoldic,vuggy,interparticle,andintraparticle.Thecomplexporesystemcreatessignificantscatterintheporosity-velocityrelationship,asindicatedinexperimentalresults（e.g.,AnselmettiandEberli,2001）.Poreshapeappearstobethedominantfactorincarbonaterockphysics.Moldic,intraframe,andvuggyporestendtoberoundedandmaketherockstronger（faster）thanwhentheporesareinterparticle.Micropores（e.g.,microcracks）tendtobeflatandmaketherockweaker.Toeffectivelycharacterizecarbonatereservoirrocks,itiscriticaltodeveloparockphysicsmodelcapableofhandlingdifferentporetypes.

Inadditiontoporetypesandporeshapes,otherfactorsneedtobeincludedinaphysics-basedrockmodel.Somespecificadditionalfactorsarelithologyandgrainshapes,multiphasefluidsandwettingeffects,rock-fluidinteractions（poro-elasticity）,stresseffects,anisotropy,heterogeneityandscaleeffects,chemicalchangestotheframework,andcorrectionsforenvironmentaleffectsduetologgingconditions.Anyrockphysicsmodelshouldbecalibratedandvalidatedwithcontrolledlaboratoryexperiments,fieldmeasurements,andcomputationalrockphysics.Inthispaper,wedemonstratetheimpactandvalidationofseveralofthesefactors.

CarbonateRockPhysicsModel

Wepreferananalyticalrockphysicsmodelratherthanempiricalonesduetoboththepredictivepowerthatitprovideswithouthavingtoacquiresomuchanalogdataanditsphysics-basednature.Empiricalrockphysicsmodelsarewidelyusedduetotheirsimplicityinfittingarelationshipbetweenparameters.Theadvantageofananalyticalmodelisthatoncetheparametersdescribingthephysicalcontrollingfactorsaredetermined,themodelcanbeappliedanywherethatthecontrollingparameterscanbeestimated.Thisachievesbetterpredictionsanddeeperunderstandingofthesubsurfacethanempiricalmodelswithlessdata.Nonlinearphysics-basedinclusionmodels（e.g.,KusterandToksoz,1974）areattractivebecausetheyhandlevariousfactorsthatimpactseismicresponseinaninternallyconsistentmanner.

WehaveextendedtheXu-White（1995）modeltocarbonaterocks（XuandPayne,2009）.Ourmethodconsistsoffoursteps:

（1）mix（e.g.,theReuss-Voigt-Hillaverage）themineralstocreateasolidmatrix;

（2）addporestothematrixintwostepsusingtheDifferentialEffectiveMedium（DEM）process（XuandWhite,1996）andKuster-Toksoz（1974）theorytoaccountforthemechanicalinteractionbetweenthepores;themicro-poreswithboundwater（e.g.,claypores）areaddedfirstandwillbeincludedinthesolidmaterialforfluidsubstitution,followedbyallotherporesincludingwater-wetmicro-poresandempty（ordry）non-bound-waterpores,whichwillusedtocalculatethedryframeelasticproperties;（3）mixthewaterwhichisnotboundtomicro-poreswiththehydrocarbonsusinganappropriatefluidmixinglaw;and（4）useGassmann’sequationstoaddthefluidmixtureintotheporesystemtoyieldthefinaleffectiveelasticpropertiesforthesaturatedrock.Theexistingmodelispartitionedintoclay-relatedpores,interparticlepores,microcracks,andstiffpores.Wekeepclayporesinthemodeltomakeitapplicableinamixedcarbonate-clasticenvironment.Theintroductionofmicrocracksandstiffporesmakesiteasytostudythepore-typeeffectonseismicresponseincarbonaterocks.

PhysicalEffects

Itiscriticaltoproperlyaccountforphysicaleffectsintherockphysicsmodeltogeneratetheproperseismicresponse.Thisisparticularlytruesinceweusetherockphysicsmodeltoinvertgeophysicalmeasurementsforsubsurfacepropertiessuchasporosity,permeability,lithology,poretype,stressstate,andothersthatdescribehydrocarbonvolumesandproductionrates.Sinceproducibilityinheterogeneouscarbonatesettingsisverydependentontheporetype,itisimportanttohavemethodstopredicttheporetypefrominversions.Cross-plotsofvelocityversusdensityoftendemonstratesignificantscatter.However,ifonesegmentstheplotintothedifferentporetypes（Figure1）,thenthescatterforeachpopulationissignificantlyreduced.Inaddition,eachpopulationcanbemodeledseparately.Ourcarbonaterockphysicsmodelrepresentseachporetypebyaspecificaspectratio（shortaxistolongaxis）.FollowingKumarandHan（2005）,weassumethattheinterparticleporeisthemostcommonporetypeincarbonaterocksand,therefore,givesareferenceporosity-velocitytrend（Figure2）.Sampleswithhighervelocitiesthanthereferencetrendcanbemodeledasamixtureofinterparticleandrounded（stiff）pores.Thosewithslowervelocitiescanbemodeledasamixtureofmicrocracksandinterparticlepores.Wecandefineatemplateofcurvesfromouranalyticalrockphysicsmodelthatrepresentsporetype.Inthismanner,wecanpredicttheporetype（idealizedporeshape,e.g.,roundedorcrack-like）byinversion.

Duetothebrittlenatureofcarbonates,fracturesaregenerallyprevalentinthem.Fracturescreatesignificantanisotropy,asdemonstratedinfigure3.Seismicdatawereacquiredoverasurfacewithaprevalentfracturedirectionasindicated.Analysisofthesedatayielded19%anisotropyinadirectionconsistentwiththatofthefracturesystem.Thesefracturesystemsactasconduitsforfluidflowandcanconsiderablyenhancehydrocarbonproductioninlow-porositycarbonates.Althoughmappingthosefracturesystemsiscriticalforimprovingproduction,modelingtheeffectoffracturesontheseismicresponsecanbedifficult.WecaneasilyaccountforanisotropyusingDEMtheory（Xuetal.,2006）andtheanisotropicGassmannequationstodofluidsubstitution.Sinceourrockphysicsmodelhandlesanisotropyuptoorthorhombicsymmetry,itcansimulatebothfractureanisotropyandshaleanisotropyconsistently.XuandPayne（2009）describethefeaturesofthismodelasappliedtofractures.Itgenerallyreliesondistinguishingporetypes.Microcracksrepresentthemostcompliantcomponentinrocks,sotheyareextremelysensitivetostress.Thepreferredorientationdistributionwillgoverntheanisotropicbehavioroftherock.Otherporesaregenerallyinsensitivetostressandhavenopreferredorientation.

Whenvariationsoccuroverscaleslargerthanporesbutlessthanawavelength,differentmixinglawsneedtobeinvoked.Thesemeso-scaleeffectsareparticularlyimportantinheterogeneoussettingslikecarbonates.Figure4showslaboratorymeasurementsconductedbyPurdueUniversityundercontracttoExxonMobilthatdemonstratethesignificantchangesinvelocitythatcanoccuroververyshortdistances.Thecomplex,multi-scaleporesystemcreatesissuesastowhenGassmanntheorymaybeappliedincarbonatesandindicatesthatmeso-scaleeffectsmustbeincorporatedintorockphysicsmodels.

FluidSubstitutioninCarbonateRocks

GassmannfluidsubstitutionisapowerfultoolforAVOanalysisand4-Dinterpretation,butGassmann’stheoryrequiresmanyassumptions.Theseassumptionstendtobeviolatedmoreeasilyinrocksampleswithmulti-poretypes（XuandPayne,2009）.Ourrockphysicsmodeltreatsthevariousporetypesdifferently.Fluidsareaddedseparatelyintothesystemwithinthedifferentporesusingtheinclusion-basedtheory.Gassmanfluidsubstitutionisonlyappliedinthoseporeswhereitisappropriatetodoso.ThisapplicationgeneratesthecorrectresponseinamixedsystemconsistingofGassmann-consistentmacroporesandnon-Gassmann-consistentmicrocrackswithunrelaxedporefluid.

Conclusions

Theoptimalproductionofreservoirsrequiresanaccurateandpreciseassessmentofthehydrocarbonvolumesandoftheratesatwhichthosereservoirscanbeproduced.Thisassessmentisparticularlyimportantincarbonates,whichtypicallyyieldonlyabout30%oftheiroilduetothecomplexityofthemineralogyandporesystems.Goodrockphysicsmodelscanallowgeophysiciststoconvertfromgeophysicalmeasurementstothedesiredengineeringandgeologicrockproperties.Severalexamplesareshowntodemonstratetheimpactthatporetypes,fluidsubstitution,scaling,anisotropy,andenvironmentaleffec