Lsc Volume 33, Number 3, December 2010

iCoverTools Name : Measurement of CBM Gas Content in Rambutan FieldiISSN : 0126 - 3501Volume33,No. 3,December2010LEMIGAS SCIENTIFIC CONTRIBUTIONS (LSC) is a printing media to promote research and developmentactivities which have been done byLEMIGASResearch and Development Centre for Oil and Gas Technology.ChiefEditor : Dra. Yanni Kussuryani, M.Si. (Kimia)Managi ngEdi tor : AgusSalim,S.H.,M.H.(EconomicLaw)Ass.ManagingEditor :Drs. Heribertus J oko Kristadi, M.Si. (Geophysic)PeerGroup :1.Prof. Dr. Ir. Septoratno Siregar (Petroleum Engineering) 2.Prof. Dr. Wahjudi Wisaksono (Energy and Environment) 3.Prof. Dr. R.P. Koesoemadinata (Geological Engineering) 4.Ir. E. J asjfi, M.Sc, APU. (ChemicalEngineering) 5.Dr. Ir. M. Kholil, M.Kom. (Management of Environment)SeniorEditors :1.Dr. Ir. Noegroho Hadi Hs.,APU. (Chemical Engineering) 2.Prof. (R). Dr. Maizar Rahman (Chemical Engineering) 3.Prof. (R). Dr. Suprajitno Munadi (Geophysics) 4.Prof. (R). Dr. E. Suhardono (Industrial Chemistry)Editors1.Ir. Bambang Wicaksono T.M., M.Sc. (PetroleumGeology) 2.Dr. Ir. Ego Syahrial, M.Sc. (Petroleum Engineering) 3.Prof. (R). M. Udiharto (Biology) 4.Drs. Mardono, MM. (Chemical Engineering) 5.Dr. Ir. Usman, M.Eng. (Petroleum Engineering)6 Abdul Haris, S.Si., M.Si. (Chemistry and Environment) 7.Ir. Yusep K Caryana, M.Sc. (Gas Engineering and Management)Secret ari at :1.Ngadimun 2.RasikinPubl i sher :LEMIGASResearch and Development Centre for Oil and Gas TechnologyAfilliation DivisionPrintedby :Grafika LEMIGASAddressLEMIGASResearchandDevelopmentDivisionfor Afilliation,J l.CiledugRaya,Kav.109,Cipulir,KebayoranLama,P.O.Box1089/J KT,J akartaSelatan12230INDONESIA,STT:No.348/SK/DITJ ENPPG/STT/1987/May12,1977,Phone:7228614,7394422-Ext.1222,1223,Fax:62-21-7228614and7246150,e-mail:[email protected] Scientific Contributions has been published since 1977, 3 times a year. The editor receives scientificarticles about research results, related to the oil and gas research.LemigasScientificContributionsis published by LEMIGAS Research and Development Centre for Oil andGas Technology.ChiefEditor:Dra. YanniKussuryani,M.Si.ManagingEditor: AgusSalim,S.H.,M.H.iiISSN : 0126 - 3501Volume 33, Number3,December2010PageCONTENTS iABSTRACT iiiAPPLICATION OF NEW COMPOSITIONAL SIMULATION APPROACHTO MODEL GRAVITY SEGREGATION IN VOLATILE OIL RESERVOIRSBy: Ego Syahrial 155 - 164AN INVESTIGATION OVER ROCK WETTABILITY AND ITSALTERATIONONSOMEINDONESIANSANDSTONESBy: Bambang Widarsono 165 - 179TRACERTESTSFORHETEROGENEITYCHARACTERIZATIONANDSATURATIONDETERMINATIONONCOREFLOODINGBy: Sugihardjo, Usman, and Utomo Pratama I. 180 - 187MODELINGGRAVITYSEGREGATIONINSTRATIFIEDAND DIPPING RESERVOIR OF VOLATILE OILBy: Ego Syahrial 188 - 197INTEGRATINGPETROGRAPHYWITHCORE-LOG-WELLTESTDATAFORLOW PERMEABILITY SANDSTONE RESERVOIR CHARACTERIZATION:PRELIMINARYRECOMMENDATIONFORPRODUCTIONOPTIMIZATIONBy: J unita Trivianty Musu, Hadi Prasetyo and Bambang Widarsono198 - 203PERFORMANCETESTINGONMIXTUREOFKISAMIRPUREPLANTOIL(PPO) AND KEROSENE AS WICK STOVE FUELBy: Emi Yuliarita204 - 211STUDYONCOMPONENTSRATINGOFGASOLINEENGINEAS A PERFORMANCE QUALITY INDICATOR OF API SL LUBRICANTBy: Setyo Widodo, Shinta Sari H., Catur Yuliani R., and Subiyanto 212 - 219iii ISSN : 0126 - 3501Date of issue: 2010 - 12Ego Syahrial (Technological Assessor at LEMIGASR & D Centre for Oil and Gas Technology)APPLI CATI ONOFNEWCOMPOSI TI ONALSIMULATION APPROACH TO MODEL GRAV-ITY SEGREGATION IN VOLATILE OIL RESER-VOIRSLSC, December 2010, Vol. 33, No. 3, p. 155 - 164ABSTRACTInthispaper,weinvestigatetherecoveryperfor-manceofgasinjectionfromvolatileoilreservoirs.Cross-sectionalreservoirstudiesforinvestigatingthegravitysegregationduringdepletionandgascyclinginvolatileoilreservoirsisdiscussed.Fur-thermoretheeffectsofverticalpermeabilityongravitysegregationinahomogeneousandhori-zontalreservoirsareinvestigated.Anewefficientcompositionalsimulationapproachwasusedinthisstudytoinvestigatetheinfluenceofgravitysegregationandtheirmagnitudeinthecaseofleangasinjectionintoavolatileoilreser-voir.Thenewcompositionalsimulationapproachwasvalidatedthroughanalyticalandnumericalmethods,anditisunconditionallystableandasstableasfullycompositionalmodel.The results show that an increase in vertical to hori-zontalpermeabilityratiosresultsinanincreaseintheeffectofgravitysegregationandyieldearlygasbreakthrough.Ontheotherhand,thesmallerthe permeability ratios (vertical to horizontal), betteraretherecoveriesduetoresultingevenlayersweeps.Gravityforceshaveaconsiderableeffectonvolatileoilrecoveryviagasinjectionandtheneedfordeterminingnotonlythefluidcharacter-istics but also the reservoir heterogeneities was sig-nificant.(Author)Keywords:compositional,equationofstate,grav-itysegregation,volatileBambang Widarsono (Researcher at LEMIGAS R& D Centre for Oil and Gas Technology)ANI NVESTI GATI ONOVERROCKWETTABI LI TY ANDI TS ALTERATI ONONSOME INDONESIAN SANDSTONESLSC, December 2010, Vol. 33, No. 3, p. 165 - 164ABSTRACTWettabilityisareservoirrockpropertythatisnoteasytomeasureandquantifybuthasacrucialef-fectonotherrockpropertiessuchasrelativeper-meability,capillarypressure,andelectricalprop-erties.Problemthatmayoccurwithregardtothismatteristhatthosepropertiesareoftenmeasuredonalreadycleansedcoresamplesaspartofthestandardprocedure.Havingundergonethenor-mallyutilizedheatedcleansingprocessalterationin the rocks original wettability was often reported.Undersuchcondition,unrepresentativewettabilitycertainlyleadstounrepresentativemeasureddatawithallofconsequences.Thisarticlepresentsastudythatuses363sandstonesamplesretrievedfrom28oilandgasfieldsinIndonesia.Thestudyconsistsoftwostagesofanalysis.Firstanalysisisperformedondataobtainedfromthreewettabilitytestsresultswhilethesecondoneismadewithus-ing water-oil relative permeability data, that is usu-allymeasuredoncleansedcoresamples.Originalwettabilitydatashowsthatthesandstonesvarryinwettabilityfromwater-wettooil-wet(48.2%and30.2%oftotalsamples,respectively).Comparisonbetweendataofthetwoanalysesshowsthatorigi-nalwettabilitytendstodegradeinstrengthaftercleaningdowntoneutralwettability,amongwhichneutral wettability appears to be the largest in num-ber(49.1%oftotalsample).Resultsalsoshowthatweakwettabilitytendstoenduremorethanstron-gerones.Theoverallresultshavedemonstratedtheneedforcautionincorehandlingandformea-suresthatcanminimizetherisk.(Author)Keywords:wettability,sandstones,alteration,corecleansing,wettabilitydegradation,misleadingpetrophysicaldata,cautiouscorehandlingThe descriptions given are free terms. This abstract sheetmay be reproduced without permission or chargeivSugihardjo1), Usman1), and Utomo Pratama I.Researcher1)at LEMIGAS R & D Centre for Oiland Gas TechnologyTRACER TESTS FOR HETEROGENEITY CHAR-ACTERIZATION AND SATURATION DETER-MINATION ON CORE FLOODINGLSC, December 2010, Vol. 33, No. 3, p. 180 - 187ABSTRACTLowsweepefficiencyisthecommonproblemindisplacementprocessduetoheterogeneity,highpermeabilitystreaks,fractures,andthiefzonesexistingintheformation.Similarly,thesuccessorfailure of EOR implementations are always affectedbythoseproblemswhichcausesdisplacingfluidsfingeringandearlybreakthrough.Factorsofthistype,unlessproperlyidentifiedandunderstoodbeforethestartofEORprocess,willlikelycauseaprojectfailure.Corefloodingasthemodelofsmallscaleoffluidsmovementsinreservoirundergoessimilarcircum-stances.Approximatelyonefootlongoffour3.5inchesstackednativeandsyntheticcoresarenor-mally used in core flooding experiment. Tracer testwasperformedtocharacterizethecoreinaddi-tionofCTscananalysis.Onthisexperiment,lithiumsolutionwasselectedastracersolutiontobetheninjectedintocoreatconstantrate,4ft/day.Afterwards,theeffluentswerecollectedbyGilsonsamplecollectorineachtubeforfurtherdetermininationitsconcentrationusingAtomicAbsorptionSpectrometry(AAS).Responsecurvesoflithiumtracerwereabletodeterminecoreheterogeneitiesandthisshouldbedonetoavoidmisleadinginterpretationofcorefloodingresults.Besides,lithiumconcentrationreportedinsomeextentandsubsequentlyanalyzedbyemployingmethodoftemporalmoments.Thismethodprovidesnumericalcalculationtoestimateeffectivecoreporevolume(PV)andfluidsatura-tion.WeighingmethodwasalsousedtocomparethePVwithaforementionedmethodandthere-sultswerecomparable.(Author)KeyWords:Tracer,heterogeinity,fluidsaturation,andcorefloodingEgo Syahrial (Technological Assessor at LEMIGASR & D Centre for Oil and Gas Technology)MODELI NGGRAVI TYSEGREGATI ONI NSTRATIFIED AND DIPPING RESERVOIR OFVOLATILE OILLSC, December 2010, Vol. 33, No. 3, p. 188 - 197ABSTRACTInthispaper,weinvestigategravitysegregationinstratifiedanddippingreservoirofvolatileoilundergasinjection.Anewefficientcompositionalsimulationapproachwasusedinthisstudytoin-vestigatetheinfluenceofgravitysegregationandtheirmagnitudeinthecaseofgasinjectionintoavolatileoilreservoir.Theresultsshowthatinstratifiedanddippingreservoirswheretheper-meabilitydecreaseswithdepth,smallertheverti-caltohorizontalpermeabilityratio,lesseristheeffectofgravitysegregation,betteristhesweepefficiencyandhencebetteristherecovery.Inthecaseofincreasingpermeabilitywithdepthinstratifieddippingreservoirs,anup-dipgasinjec-tionintoavolatileoilreservoirwasfoundtobeafavourableconditionintermofrecovery.Gravityforceshaveaconsiderableeffectonvolatileoilrecoveryviagasinjectionandtheneedforde-termining not only the fluid characteristics but alsothereservoirheterogeneitieswassignificant.(Author)Key words: compositional, equation of state, grav-itysegregation,volatilevJunita Trivianty Musu1), Hadi Prasetyo2) and BambangWidarsono1) (Researcher1) at LEMIGAS R & DCentre for Oil and Gas Technology, Badan PelaksanaHulu MIGAS (BPMIGAS)2))INTEGRATING PETROGRAPHY WITH CORE-LOG-WELL TEST DATA FOR LOW PERME-ABILITY SANDSTONE RESERVOIR CHARAC-TERIZATION: PRELIMINARY RECOMMENDA-TION FOR PRODUCTION OPTIMIZATIONLSC, December 2010, Vol. 33, No. 3, p. 198 - 197ABSTRACTIntegratingpetrographiccoreinformationintocombined core petrophysics, log, and well test dataforunderstandingfaciesandenvironmentaldepo-sitioninrockcharacterizationhasproveditselfusefultoimprovingqualityandreliabilityoftherequiredconclusions.Thisintegratedapproachhasspecificallyshownitsuseinthecasesofcom-plexreservoirssuchonescharacterizedaslow-permeability sandstone reservoirs.It is in this spiritthat this paper demonstrates how this virtually costefficientanalysisprovidespreliminaryrecommen-dationsfortheexploitationofsuchreservoirs.Ascasestudy,twotypesofproducingreservoirs(Bekasap,Bangko,Pematang,andTanjungFor-mations)havebeentakenin2009.Thefirsttypeisstronglycontrolledbydepositionalenvironment.ItisfoundintheupperpartofBekasapandBangkoformations(1900-2300ft-ss),depositedinestuarinesystem,andmadeofveryfinetofinegrainedsandwithlowtomoderatebioturbation.Thismostlyfeldspathicandlithicgreywackeshavepermeabilityofupto200mD.ThesecondtypeisstronglydominatedbydiagenesisprocessandismainlyfoundintheUpperPematangandTanjungFormations(6200-7400ft-ss).Thisres-ervoirtypeischaracterizedbyitscoarse-grainedandconglomeraticsandstonesresultedfromfan-deltaandbraidedchanneldepositionalsystem.Diageneticeventssuchascompaction,recrys-tallizationofmatrixintomicrocrystallineclayminerals,precipitationofauthigenicmineralsinporesystemarealsowellidentifiedfromtheper-formedpetrographicanalysis.Thisisdominatedbysublithareniteandlitharenitesandstonesex-hibithorizontalpermeabilityofuptoseveraldozensmD.Thetwoproducingreservoirtypeshaveundergonecarefullyplannedexploitationandstimulationoperations,andthehorizontaldrillingandfracturingjobforthetype-1andtype-2reservoirs,respectively,areacknowledgedastwosuccessstoriesoftheirown.Thesesuc-cesseswouldnotprevailwithoutapplicationofwellintegratedcore-log-welltestapproachesinreservoircharacterization,inwhichinformationfromcorepetrographyplaysanimportantcon-tribution.(Author)Keywords:reservoircharacterization,sandstone,lowpermeability,petrographyviEmi Yuliarita (Researcher at LEMIGAS R & DCentre for Oil and Gas Technology)PERFORMANCE TESTING ON MIXTURE OFKI SAMI RPUREPLANTOI L(PPO)ANDKEROSENE AS WICK STOVE FUELLSC, December 2010, Vol. 33, No. 3, p. 210 - 217ABSTRACTPurePlantOil(PPO)whichismadefromkisamirseedhassmallerkineticviscosityvaluethanjatrophaandcoconutoil.Soithaspotentialtobeusedasalternativefuel/mixedkerosene.Thetestresultofsomemainphysical/chemicalcharacteristicsoffuelmadefromkerosineandpure plant oil (5% to 20%volume) are still in thelimitofkerosenespecificationasdecidedbythegovernment.However,themaximumpowertestresultofthemixtureofPPOandkerosenethathasbeentestedon16wicksstoveshowsthatthehighercontentofPPOinkerosenewilldecreasethe maximum stove performance as well as stovesefficiencyvalue.Butthebluecoloroffiregetsclearer,becauseoflessamountofsulfurbyadd-ingPPOinkerosene.TheuseofPPOupto20%willreducesulphurcontentupto20%.(Author)KeyWord:PPO,KerosineAlternatiffuel,Spesification,MaksimumPower,EfficiencystoveSetyo Widodo1), Shinta Sari H.2), Catur Yuliani R.1),and Subiyanto1) (Researcher1), Lubricant analyst2) atLEMIGAS R & D Centre for Oil and Gas Tech-nology)STUDY ON COMPONENTS RATING OF GASO-LINE ENGINE AS A PERFORMANCE QUALITYINDICATOR OF LEMIGAS FORMULATED APISL LUBRICANTLSC, December 2010, Vol. 33, No. 3, p. 218 - 225ABSTRACTPoorlubricationmaycausewearonthesurfacemovingpartsofenginecomponentssuchasbear-ings due to the metal-to-metal contact. Engine com-ponentsutilizedontheroad-testofgasolineengineslubricatingoilAPISLshowedwearandtearonsomepartsofthem.Thesumofwearoc-curredduringtheroadtestwerevaried.There-fore,ananalysisofwearquantityofenginescom-ponentswasanecessityinordertogetinforma-tion about lubrication condition on engine.Analy-sisofwearwasconductedbycomponentsratingbasedonthestandardspecificationssetoutforperformanceleveloflubricantoilAPISLandILSACGF-3(SNI06-7069-2005).AnalysisbasedonSeq.IIIFshowedthataveragevalueofthepis-tonskirtvarnishis10,lowtemperatureviscosityis 4673 cP, and cam wear lifter is 0.002 mm. It wasalsoshowedthattheminimumkinematicsviscos-ityincreasewasmanagedtobestay-in-grade.AnalysisbasedonSeq.IVAshowedthattheaver-agevalueofcamwearis0.0015mm.AnalysisbasedonSeq.VIIshowedthatthevalueofbear-ingweightlosswas0.010gandtherewasnode-positathightemperatures.Shearstabilityanaly-sisbasedonSeq.VIIIshowedthattheviscosityoflubricant oil is still in the range of allowed values.(Author)Keywords:rating;gasolineenginecomponents;APISLlubricatingoil155APPLICATION OF NEW COMPOSITIONAL SIMULATIONLEMIGAS SCIENTIFIC CONTRIBUTIONSEGO SYAHRIALVOL. 33. NO. 3, DECEMBER2010 : 155 - 164ABSTRACTSInthispaper,weinvestigatetherecoveryperformanceofgasinjectionfromvolatileoilreservoirs.Cross-sectionalreservoirstudiesforinvestigatingthegravitysegregationdur-ingdepletionandgascyclinginvolatileoilreservoirsisdiscussed.Furthermoretheeffectsofverticalpermeabilityongravitysegregationinahomogeneousandhorizontalreser-voirsareinvestigated.Anewefficientcompositionalsimulationapproachwasusedinthisstudytoinvestigatetheinfluenceofgravitysegregationandtheirmagnitudeinthecaseofleangasinjectionintoavolatileoilreservoir.Thenewcompositionalsimulationapproachwasvalidatedthroughanalyticalandnumericalmethods,anditisunconditionallystableandasstableasfullycompositionalmodel.Theresultsshowthatanincreaseinverticaltohorizontalpermeabilityratiosresultsinanincreaseintheeffectofgravitysegregationandyieldearlygasbreakthrough.Ontheotherhand,thesmallerthepermeabilityratios(verticaltohorizontal),betteraretherecov-eriesduetoresultingevenlayersweeps.Gravityforceshaveaconsiderableeffectonvolatileoilrecoveryviagasinjectionandtheneedfordeterminingnotonlythefluidcharacteristicsbutalsothereservoirheterogeneitieswassignificant.Keywords:compositional,equationofstate,gravitysegregation,volatileAPPLICATION OF NEW COMPOSITIONAL SIMULATIONAPPROACH TO MODEL GRAVITY SEGREGATION INVOLATILE OIL RESERVOIRSBy: Ego SyahrialTechnological Assessor at LEMIGAS R & D Centre for Oil and Gas TechnologyJ l. Ciledug Raya Kav. 109, Cipulir, Kebayoran Lama, J akarta Selatan 12230, INDONESIATromol Pos: 6022/KBYB-J akarta 12120,Telephone: 62-21-7394422, Faxsimile: 62-21-7246150First Registered on 27 September2010; Received after Corection on 5November2010;Publication Approval on :31 December 2010I. INTRODUCTIONA volatile oil is defined as a high shrinkage crudeoil near its critical point (Moses, 1986). In a phasediagram, it is recognised as a type between a black-oil and a gas-condensate fluid. With deeper drilling,more reservoirs containing volatile crude oil and gascondensates have been found and the need for accu-rate and economic methods for studying the perfor-mance of such reservoirs has become important. Inthe early 1950s several material balance methodswere used for reservoir performance predictions (e.g.,Cooketal.,1951;J acoby&Berry,1957;Reudelhuber & Hinds, 1957). Cook et al. (1951) pre-sented a method of estimating future reservoir per-formance and oil recovery of highly volatile type oilreservoirs. Later, Woods (1955) applied Cook et al.method to real field case study of that particular typeof reservoirs. Reudelhuber & Hinds (1957) presenteda compositional material balance method for the pre-diction of recovery from volatile oil depletion-drivereservoirs. Fluid compositions were determined fromlaboratory data and the actual reservoir study wasconducted. J acoby & Berry (1957) developed a tank-type model which used vapour-liquid equilibrium(VLE) and composition dependent densities and vis-cosities. This method used relative permeability dataand multi-component flash calculations to predict oiland gas production as a function of reservoir pres-sure. Later, Cordell & Ebert (1965) showed that thegreater accuracy of the volatile oil material balancewas due to the consideration of the oil recovered fromthe gas phase.156APPLICATION OF NEW COMPOSITIONAL SIMULATIONLEMIGAS SCIENTIFIC CONTRIBUTIONSEGO SYAHRIALVOL. 33. NO. 3, DECEMBER2010 : 155 - 164During the late of 1960s, the use of numericalcompositional methods increased significantly with therapid evolution of large scale, high speed, digital com-puters and the development of numerical mathemati-cal methods. Numerical simulators, in general, utilisefinite difference approximations to the rather com-plex partial differential equations that mathematicallydescribe the physics and the thermodynamics of fluidflow in porous media. Simultaneously solving the con-tinuity equations (after applying Darcys Law), andthe Equations of State for each phase, under the pre-scribed initial and boundary equations has become astandard method of developing a model for two-phasefluid flow in a porous media. Black oil simulators areused to simulate and predict reservoir performanceby considering hydrocarbon fluids as two lumpedcomponents (phases) namely oil and gas. In this ap-proach, inter-phase mass transfers were assumed tobe a function of pressure only. For volatile oils, andgas condensates, this assumption may not be valid(Daltaban, 1986). Compositional models are used tosimulate adequately the inter-phase mass transfer andpredict reservoir performancewhen compositionaleffects cannot be neglected.The development of compositional simulators canbe classified into three categories. The first categoryconcerns with the new formulations and efficientsolution schemes for the mass conservation equa-tions. In this category the formulations are dividedinto two basic schemes, namely IMPES and fullyimplicit schemes. The primary difference betweenthese two schemes is in the treatment of the flowcoefficient. The second category concerns with theefficiency of the phase equilibrium calculationschemes.Inthiscategory,theformulationisdifferenced whether or not they use the Equations ofState for phase equilibrium and property calculations.The last category concerns the representations ofphysical phenomena, such as the effect of interfacialtension to the shape of relative permeability curves.In general, the IMPES is inherently unstable andthefullyimplicitcanoverkilltheproblemcomputationally. To realise the problems, it is there-fore intended to propose a new formulation in orderto minimise the cost of the computational simulationwhile maintaining the thermodynamic consistency ofthe prediction. The formulation must be able to modelrecovery from volatile oil reservoirs in the presenceofheterogeneityunderdifferentrecoverymechanicms. In this paper the new compositionalsimulation approach is used to investigate the influ-ence of gravity segregation and their magnitude inthe case of lean gas injection into a volatile oil reser-voir.II. COMPOSITIONALSIMULATIONA new compositional simulation approach for avolatile oil reservoir modeling was presented in pre-vious publications (Syahrial & Daltaban, 1998; Syahrial& Daltaban, 1998; Syahrial 2010). The new formu-lation has an implicit equation for the oil-phase pres-sure and water saturation, an explicit equation forthe hydrocarbon saturation, and explicit equation forthe overall composition of each hydrocarbon compo-nent that satisfies thermodynamic equilibrium. Theformulation uses an Equation of State for phase equi-librium and property calculations. Interfacial tensioneffects are included in the formulation characterisethe thermodynamically dynamic nature of the rela-tive permeability. A two-dimensional relative perme-ability algorithm is included which handles lumpedhydrocarbon phase as well as individual phase flows.For each grid block two equations are required,namely total hydrocarbon and water-phase flow equa-tions. These equations are highly non-linear and theyare linearised by using Newton-Raphson method. Theresulting set of equations are solved by an efficientConjugate Gradient based iterative technique to ob-tain pressures and saturations simultaneously, andhydrocarbon-phase saturations are deduced from theirrespective equations.A.GeneralisedFlowEquationsThe general flow equation used in the formula-tion can be found equations by summing up all theequations, applying mole constraint, and convertingthe resulting expressions into finite difference formnamely:- Water equation:| |( ) ( ) ( )| |A AuAT qVtS Sw w w wrw wnw wn+ = + | |1, (1)- Oil equation:| |( ) ( ) ( )| |A AuAT qVtS So o o oro ono on+ = + | |1,(2)157APPLICATION OF NEW COMPOSITIONAL SIMULATIONLEMIGAS SCIENTIFIC CONTRIBUTIONSEGO SYAHRIALVOL. 33. NO. 3, DECEMBER2010 : 155 - 164- Gas equation:| |( ) ( ) ( )| |A AuAT qVtS Sg g g grg gng gn+ = + | |1, (3)where transmissibility term Tl in the x-direction,w g o lkxA kTllrll, , , =||.|

\||.|

\|A= The same expression exists for y- and z-direction. In this formulation, all transmissibilityterm are treated implicitly. To obtain the hydro-carbon equation, both sides of oil and gas equa-tions (Eqs. (2) and (3)) are multiplied bygn+1and on+1 respectively, and combined, hence:- Hydrocarbon equation:| | | | ( ) ( )( ) ( ) ( ) | |ng gnono ongnh g orng gnono ongngngnononongS S StVq q T T| | | 1 1 11 1 1 1 1 1 1 1+ + ++ + + + + + + + A=+ + Au A + Au A (4)where,( ) ( ) ( )1 1 1 + + ++ =ngnonhS S S- Water equation:| | ( ) ( ) ( ) | |nw wnw wrnw wnwnwS StVq T | | A= + Au A+ + + 1 1 1 (5)B.LinearisationandDiscretisationIt is clear that both water and hydrocarbonequations (Eqs. (4) and (5)) are highly non-lin-ear and analytical solutions are not possible.Consequently, numerical methods are required.To implement numerical techniques, however,the flow equations must be linearised and theresults are water and hydrocarbon equations inthe oil-phase pressure and the water saturationforms. Effects of capillary pressure are treatedexplicitly.The discretisation of water and hydrocar-bon equations is carried out by applying a finitedifference scheme using backward differenceintimeandcentraldifferenceinspace(Peaceman, 1967). This results in water andhydrocarbon equations having the form:- Water Equation:rw w S w Sw S w S w Sw S w S o Po P o P o Po P o P o PC S W S WS W S W S WS W S W P WP W P W P WP W P W P Wk j i k j i k j i k j ik j i k j i k j i k j i k j i k j ik j i k j i k j i k j i k j i k j ik j i k j i k j i k j i k j i k j ik j i k j i k j i k j i k j i k j i= ++ + ++ + ++ + ++ + ++ + + ++ + + ++ + + + 1 , , 1 , , , 1 , , 1 ,, , 1 , , 1 , , , , , , 1 , , 1, 1 , , 1 , 1 , , 1 , , 1 , , 1 , ,, 1 , , 1 , , , 1 , , 1 , , , ,, , 1 , , 1 , 1 , , 1 , 1 , , 1 , ,o oo o oo o oo o oo o o(6)- Hydrocarbon Equation:rh w S w Sw S w S w Sw S w S o Po P o P o Po P o P o PC S H S HS H S H S HS H S H P HP H P H P HP H P H P Hk j i k j i k j i k j ik j i k j i k j i k j i k j i k j ik j i k j i k j i k j i k j i k j ik j i k j i k j i k j i k j i k j ik j i k j i k j i k j i k j i k j i= ++ + ++ + ++ + ++ + ++ + + ++ + + ++ + + + 1 , , 1 , , , 1 , , 1 ,, , 1 , , 1 , , , , , , 1 , , 1, 1 , , 1 , 1 , , 1 , , 1 , , 1 , ,, 1 , , 1 , , , 1 , , 1 , , , ,, , 1 , , 1 , 1 , , 1 , 1 , , 1 , ,o oo o oo o oo o oo o o(7)The system of equations above can be written in matrixform:k kb x A =+1o (8)This particular matrix form can be solved in eachNewtonian iteration by either direct, or iterative methods inorder to obtain the required changes in pressure and satura-tion.C.CompositionandSaturationEquationsCompositions are computed explicitly by a method de-veloped by Tsutsumi and Dixon (1972). The overall compo-sitions of the components can be expressed as:| | ( ) ( ) ( ) { } | || | ( ) ( ) ( ) { } | |(((((

+A+ + + Au + Au A+A+ + + Au + Au A=+ + + ++ + + + + ++ng g o or ng g o ongngnonong g o on r ng g m o o mngngnmnononmnmS StVq q T TS S ztVq y q x T y T xz| | | | 1 1 1 11 1 1 1 1 11(9)Oil and gas saturations are calculated as the final result of aseries of computations form:| | ( ) ( ) | |( ) | |,11 11(((((

AA+ + Au A=++ ++norno orno onononotVStVq TS|| (10)| | ( ) ( ) | |( ) | |.11 11(((((

AA+ + Au A=++ ++ngrng grng gngngngtVStVq TS|| (11)D.ValidationProceduresThe equations presented in the previous section werecoded into a computer program that provides a field-scale158APPLICATION OF NEW COMPOSITIONAL SIMULATIONLEMIGAS SCIENTIFIC CONTRIBUTIONSEGO SYAHRIALVOL. 33. NO. 3, DECEMBER2010 : 155 - 164reservoir simulator which models the behaviour ofcompositionalprocessesandinparticularthebehaviour of volatile oil reservoirs. The results fromthe simulation procedure are validated by comparingthem against both analytical and numerical models.The Buckley-Leverett method (Buckley & Leverett,1942) was used as the analytical model.Numericalvalidation was provided by Eclipse300 (a fully com-positional simulator that is the de rigour industry stan-dard (GeoQuest, 1996)). The validation tests haveshown that this formulation gives sufficiently closeapproximations to the analytical Buckley-Leverettsolution and other numerical methods (Syahrial, 2010).The new model requires less number of equations tobe solved per time step than the fully implicit methodand only needs one to two iterations per time step,this formulation is as cheap as IMPES and is as ac-curate as fully implicit methods.III.APPLICATIONSAfter validating againstanalytical and numerical meth-ods, the simulator was used tomodel recovery from volatileoil reservoirs. The objective ofthis section is to carry out atheoretical investigation intothe recovery performance GasInjection from volatile oil res-ervoirs. Cross-sectional stud-ies for investigating the grav-ity segregation during depletionand gas cycling in volatile oilreservoirs is discussed. Fur-thermore the effects of verti-cal permeability on gravity seg-regation in a homogeneouesand horizontal reservoirs areinvestigated.A.Cross-SectionalStudiesIn this section, two-dimen-sional studies in which flow ispermitted in only the horizon-tal and vertical directions arediscussed. These studies areintended to illustrate the effectof gravity segregation on theoil recovery in the case of gasinjection into volatile oil reservoirs. The recovery ef-ficiencies of immiscible and miscible gas displace-ment due to gravity segregation are affected by:1. Increased permeability (either horizontal or ver-tical).2. Increased density difference.3. Increased mobility ratio.4. Decreasing production rates.A compositional simulation approach is fully imple-mented to investigate the influence of gravity segre-gation and their magnitude in the case of lean gasinjection into a volatile oil reservoir. Well productionperformance, gas saturation distribution and the com-position of the production stream will be monitored inview of gravity segregation in order to explore theways of maximising recovery. By knowing the fac-tors and the magnitude of the influence of gravitysegregation, design considerations of the injection fluidinto reservoir fluid can be properly evaluated.Property Field Units SI UnitsGrid System 40x1x20 40x1x20Reservoir Length, L 3000 ft 914.40 mReservoir Width, w 50 ft 15.24 mThickness of the Pay Zone, h 100 ft 30.48 mArea of Cross-Section, A5000 ft2464.52 m2Dip Angle, u 0 0Horizontal Permeability, kh200 mD1.97x10-13 m2Porosity, | 15% 15%Connate Water Saturation, Swc20% 20%Residual Oil Saturation, Sor30% 30%Residual Gas Saturation, Sgr5% 5%Initial Oil Saturation, Soi80% 80%Initail Gas Saturation, Sgi0% 0%Initial Water Saturation, Swi20% 20%Initial Reservoir Pressure at Datum, Pi2800.0 psia 19.31 MpaDatum 8500 ft 2591 mReservoir Temperature, Tr234F 112.2CProduction Point, Grid Block No. 1 1Injection Point, Grid Block No. 40 40Rock Compressibility, cr 4x10-6 psi-15.80x10-7 kPa-1Water Compressibility, cw 3x10-6 psi-14.35x10-7 kPa-1Table 1Data used for cross-sectional studies159APPLICATION OF NEW COMPOSITIONAL SIMULATIONLEMIGAS SCIENTIFIC CONTRIBUTIONSEGO SYAHRIALVOL. 33. NO. 3, DECEMBER2010 : 155 - 164B. GravitySegregationinHomogeneousandHorizontalReservoirsThe reservoir domain selected forthe purposes is a two-dimensionalcross-section reservoirs. The lengthof the reservoir is 3000 ft, width is50 ft and the thickness of the pay zoneis 100 ft. The average horizontal per-meability and porosity are 200 mDand 15% respectively. Initial reser-voir pressure at the datum is 2800psia with 20% water and 80% oilsaturations yielding 0.324 MMBBLof hydrocarbon pore volume. Initialoil-in-place, calculated by flashing theoil at stock-tank conditions of 14.69psia and 60F is 208 MSTB andstock-tank GOR is 922 SCF/STB.Table 1 shows the other relevant dataof this study.The fluid used in this study is thatof OIL-6 (Coats & Smart, 1982) andTable 2 shows the composition andproperties of that fluid. The fluid dataexhibits bubble point pressure of 2733 psia and the oildensity is 36.9 lb/ft3. The relative permeability curvesare shown in Figures 1 & 2. The reservoir domain isdiscretised by 4020 grid blocks with a productionwell and a injector well at the extremas. Total num-ber of active grid blocks is 800 and each gridblockcontains 401 RBBL of hydrocarbon pore volume. Itwas assumed that production and injection wells pen-etrate all layers with oil production rate of 200 STB/Day and minimum bottom hole pressure of 2000 psia.Lean gas injection with a composition listed in Table3is used as the gas cycling processes, and assumedTable 2Fluid compositions and properties at reservoir conditionsComponent Mole Fr ac. Tc(F) p c (psia) ZcMW P chCO20.0103 88.79 1071.33 0.2741 44.01 0.225 78.0N 20.0055 -232.51 492.31 0.2912 28.01 0.040 41.0C 10.3647 -116.59 667.78 0.2847 16.04 0.013 77.0C 20.0933 90.10 708.34 0.2846 30.07 0.099 108.0C 30.0885 205.97 618.70 0.2775 44.10 0.152 150.3C 40.0600 295.43 543.45 0.2772 58.12 0.196 187.2C 50.0378 378.95 487.17 0.2688 72.15 0.241 228.9C 60.0356 461.93 484.38 0.2754 84.00 0.250 271.0C 7 +0.3043 836.63 266.33 0.2398 200.00 0.648 520.0Lean Gas C 1C 2C 3C 4C 5C 6 C 7 +Mole Frac. 0.85 0.01 0.01 0.03 0.01 0.01 0.08Table 3Compositions of the lean injection gasFigure 1Water-oil relative permeability curves160APPLICATION OF NEW COMPOSITIONAL SIMULATIONLEMIGAS SCIENTIFIC CONTRIBUTIONSEGO SYAHRIALVOL. 33. NO. 3, DECEMBER2010 : 155 - 164that miscibility occurred during the in-jection. To compare the relative per-formance of gas cycling with differ-ent values of vertical to horizontal per-meability ratio, natural depletion is alsoconducted in this study.C. Effect of VerticalPermeability on GravitySegregationIn this section the effect of verti-cal permeability on gravity segrega-tion in homogeneous and horizontalreservoirs is investigated. Three dif-ferent ratios of vertical to horizontalpermeability kv/kh: 1.0, 0.1 and 0.01respectively are used in the modellingprocedure. Also, the depletion pro-cesses is modelled with a ratio of ver-tical to horizontal permeability of 0.1.Figures 3 through 5 show themethane saturation distribution foreach case studied after 18.8% HCPV(300 days) of lean gas injection. It isclear from these Figures that an in-crease in vertical to horizontal perme-ability ratios results in an increase inthe effect of gravity segregation andyield early gas breakthrough. Table 4shows the relative breakthrough times(defined as approximately 2% Meth-ane increase in the producing stream)and the corresponding C7+ (a compo-nent that characterises the liquid) re-coveries. It is also clear from this Tablethat smaller the permeability ratios(vertical to horizontal) better are therecoveries due to resulting even layersweeps.Table 5 shows the comparison ofC7+recoveries and gas-oil ratios after600 days (37% HCPV) of gas injec-tion. Figures 6 & 7 show the temporalvariations of C7+production rate andgas-oil ratio respectively for the sameperiod. The C7+ recovery for kv/kh:0.01 is about 1.6 times that can be ob-tained from kv/kh: 1.0. The gas-oil ra-Figure 2Gas-oil relative permeability curvesFigure 3Gas saturation in a horizontal reservoir, kv/kh = 1.0Figure 4Gas saturation in a horizontal reservoir, kv/kh = 0.01161APPLICATION OF NEW COMPOSITIONAL SIMULATIONLEMIGAS SCIENTIFIC CONTRIBUTIONSEGO SYAHRIALVOL. 33. NO. 3, DECEMBER2010 : 155 - 164tio for Figure 7 shows dramaticincrease after the breakthroughfor each injection case as ex-pected. At 600 days, their valuesas given in Table 5 demonstratethat the smallest permeabilitycase exhibit marked differencethan others. For the cases of kv/kh: 1.0 and kv/kh: 0.1, the resultsare similar due to not only the lesscontrast in the permeabilities butalso to the balance between vis-cous to gravity forces. For a kv/kh of 0.01, the viscous forceshave more control on the results.k v/k hBreakthrough Time (Days)C7 + Recovery (frac.)1.00 130 0.310.10 250 0.320.01 340 0.33kv/khC7 + Recovery (frac.)GOR MSCF/STB)1.00 0.16 5.500.10 0.18 5.400.01 0.25 4.43Table 4C7+recovery at breakthroughTable 5C7+ recovery and GOR after 600 daysFigure 5Gas saturation in a horizontal reservoir, kv/kh= 0.01Figure 6Temporal variation of C7+ in a horizontal reservoir162APPLICATION OF NEW COMPOSITIONAL SIMULATIONLEMIGAS SCIENTIFIC CONTRIBUTIONSEGO SYAHRIALVOL. 33. NO. 3, DECEMBER2010 : 155 - 164Figure 7Gas-Oil ratio versus time in a horizontal reservoirFigure 8Total mole fraction of methane after 300 daysFigure 9Totalmole fraction of methane after 1000 Days163APPLICATION OF NEW COMPOSITIONAL SIMULATIONLEMIGAS SCIENTIFIC CONTRIBUTIONSEGO SYAHRIALVOL. 33. NO. 3, DECEMBER2010 : 155 - 164Figures 8 & 9 show the distribution of Methanemole fraction after 300 and 1000 days respectivelyat the top and bottom of the reservoir for each casestudied. The Figure 8 shows that the total mole frac-tion of Methane increases rapidly in the top layer forkv/kh of 1.0 and 0.1 in all grid blocks. As injectionproceeds, the total mole fraction of Methane with kv/kh: 0.01 reached the same level as the model with kv/kh is 1.0 and 0.1. Figure 9 confirms that situation.In the case studies presented in this section, itwas demonstrated that the gravity forces have con-siderable effect on volatile oil recovery via gas injec-tion and the need for determining not only the fluidcharacteristics but also the reservoir heterogeneities.IV.CONCLUSIONS1. The model formulation developed has an implicittransmissibility term, an implicit for oil-phase pres-sure and water saturation and explicit equationfor the overall composition of each hydrocarboncomponent that satisfies thermodynamic equilib-rium. It is unconditionally stable like the FullyImplicit approach and can be as cheap as IMPES.2. The new model requires less number of equa-tions to be solved per time step than the fully im-plicit method and only needs one to two iterationsper time step, this formulation is as cheap asIMPES and is as accurate as fully implicit meth-ods.3. An increase in vertical to horizontal permeabilityratios results in an increase in the effect of grav-ity segregation and yield early gas breakthrough.The smaller the permeability ratios (vertical tohorizontal), better are the recoveries due to re-sulting even layer sweeps.4. Gravity forces have a considerable effect on vola-tile oil recovery via gas injection and the need fordetermining not only the fluid characteristics butalso the reservoir heterogeneities was significant.REFERENCES1. Buckley, S.E. and Leverett, M.C.: Mechanismof Fluid Displacement in Sands, Trans., AIME146, (1942) 107-116.2. Coats, K.H. and Smart, G.T.: Application Of ARegression Based EOS PVT Programe To Labo-ratory Data, paper SPE 11197, Proc. 57th An-nual Fall Technical Conference and Exhibition ofthe SPE of AIME, New Orleans, Los Angeles,(Sept. 26-29, 1982).3. Cook, A.B., Spencer, G.B. and Bobrowski F.P.:Special Considerations in Predicting ReservoirPerformance of Highly Volatile Type Oil Reser-voir, Trans., AIME (1951) 192, 37-46.4. Cordell, J .C. and Ebert, C.K.: A Case History -Comparison of Predicted and Actual Performanceof a Reservoir Producing Volatile Crude Oil, J PT.(Nov. 1965).5. Daltaban, T.S.: Numerical Modelling Of Recov-ery Processes From Gas Condensate Reservoirs,Ph.D. thesis, Dept. of Mineral Resources Engi-neering, Imperial College of Science, Technologyand Medine, London (1986).6. GeoQuest, Schlumberger : Eclipse 300: Refer-ence Manual, Version 96A, (1996).7. J acoby, R.H. and Berry, V.J . J r.: A Method forPredicting Depletion Performance of a Reser-voir Producing Volatile Crude Oil, Trans., AIME(1957) 210, 27-33.8. Moses, P.L.: Engineering Applications of PhaseBehaviour of Crude Oiland Condensate Sys-tems, J .P. Tech., (J uly 1986), 38, 715-723.9. Peaceman, D.W.: Fundamentals of NumericalSimulation, Elsevier Scientific Publishing Co.,Amsterdam Nedertland (1967).10. Reudelhuber, F.O. and Hinds, R.F.: A Composi-tional Material Balance Method for PredictionRevoveryfrom Volatile Oil Dpletion Drive Res-ervoirs, Trans., AIME (1957) 210, 19-26.11. Syahrial, E. and Daltaban, T.S.: A New Compo-sitional Simulation Approach to Model Recoveryfrom Volatile Oil Reservoirs, paper SPE 39757,the 1998 SPE Asia Pacific Conference on Inte-grated Modelling for Asset Management held inKuala Lumpur, Malaysia, 23-24 March 1998.12. Syahrial, E. and Daltaban, T.S.: Developmentof A Novel Compositional Simulation Approachto Model Recovery from Volatile Oil Reservoirs,presented in the 26th Annual Convention & GasHabitats of SE Asia and Australia Conferenceheld in J akarta, Indonesia, 27-29 October 1998.13. Syahrial, E. : A New Approach of CompositionalSimulation for a Volatile Oil Reservoir Modeling.Lemigas Sientific Contribution, Vol 33, Number1, May 2010.164APPLICATION OF NEW COMPOSITIONAL SIMULATIONLEMIGAS SCIENTIFIC CONTRIBUTIONSEGO SYAHRIALVOL. 33. NO. 3, DECEMBER2010 : 155 - 16414. Tsutsumi, G. and Dixon, T.N.: Numerical Simu-lation of Two-Phase Flow With Interphase MassTransfer In Petroleum Reservoirs, paper SPE,Proc. SPE-AIME 47th Annual Fall Meeting, SanAntonio (Oct. 8-11, 1972).15. Woods, R.W.: Case History of Reservoir Per-formance of aHighly Volatile Type Oil Reser-voir, Trans., AIME (1955) 204, 156-59.165AN INVESTIGATION OVER ROCK WETTABILITY AND ITSLEMIGAS SCIENTIFIC CONTRIBUTIONSBAMBANG WIDARSONOVOL. 33. NO. 3,DECEMBER 2010 : 165 - 179AN INVESTIGATION OVER ROCK WETTABILITY ANDITS ALTERATION ON SOME INDONESIAN SANDSTONESBy: Bambang WidarsonoResearcher at LEMIGAS R & D Centre for Oil and Gas TechnologyJ l. Ciledug Raya, Kav. 109, Cipulir, Kebayoran Lama, P.O. Box 1089/J KT, J akarta Selatan 12230INDONESIATromol Pos: 6022/KBYB-J akarta 12120,Telephone: 62-21-7394422, Faxsimile: 62-21-7246150First Registered on27 September 2010; Received after Corection on 3 December 2010Publication Approval on : 31 December 2010ABSTRACTWettabilityisareservoirrockpropertythatisnoteasytomeasureandquantifybuthasacrucialeffectonotherrockpropertiessuchasrelativepermeability,capillarypressure,andelectricalproperties.Problemthatmayoccurwithregardtothismatteristhatthosepropertiesareoftenmeasuredonalreadycleansedcoresamplesaspartofthestandardprocedure.Havingundergonethenormallyutilizedheatedcleansingprocessalterationintherocksoriginalwettabilitywasoftenreported.Undersuchcondition,unrepresentativewettabilitycertainlyleadstounrepresentativemeasureddatawithallofconsequences.Thisarticlepresentsastudythatuses363sandstonesamplesretrievedfrom28oilandgasfieldsinIndonesia.Thestudyconsistsoftwostagesofanalysis.Firstanalysisisperformedon data obtained from three wettability tests results while the second one is made with usingwater-oilrelativepermeabilitydata,thatisusuallymeasuredoncleansedcoresamples.Originalwettabilitydatashowsthatthesandstonesvarryinwettabilityfromwater-wettooil-wet(48.2%and30.2%oftotalsamples,respectively).Comparisonbetweendataofthetwoanalysesshowsthatoriginalwettabilitytendstodegradeinstrengthaftercleaningdowntoneutralwettability,amongwhichneutralwettabilityappearstobethelargestinnumber(49.1%oftotalsample).Resultsalsoshowthatweakwettabilitytendstoenduremorethanstrongerones.Theoverallresultshavedemonstratedtheneedforcautionincorehandlingandformeasuresthatcanminimizetherisk.Keywords:wettability,sandstones,alteration,corecleansing,wettabilitydegradation,misleadingpetrophysicaldata,cautiouscorehandlingI.INTRODUCTIONOne of the most important properties of reser-voir rocks is wettability. Wettability is basically aninclination of reservoir rocks to be wetted by certainfluids, either oil or water, due to which other rockphysical properties such as capillary pressure andrelative permeability are influenced. Reservoir rocksthat tend to be water-wet respond differently to oilflow compared to what is shown by oil-wet ones,which in turn controls capillary pressure and relativepermeability behavior hence governing hydrocarbondisplacement and ultimate hydrocarbon recovery.In oil saturated water-wet rocks the oil rests onthin film of water spread over the rocks interior sur-face area. When the rock is in contact with waterthe water imbibes and displaces the oil out. Watertends to fill all pores including the smallest ones. Onthe contrary, in oil saturated oil-wet rocks the oil tendsto act as water in a water-wet system. The oil dis-places water and enters into the finest pores. Thetwo different tendencies shown by the two differentpreferences to wettability certainly have differentconsequences on any attempt to produce the oil outfrom the rocks.The fact stated above has been long studied byengineers and earth scientists. It is known that sand-stones tends to exhibit neutral to water-wet charac-teristics (e.g. Block and Simms, 1967, as quoted in168AN INVESTIGATION OVER ROCK WETTABILITY AND ITSLEMIGAS SCIENTIFIC CONTRIBUTIONSBAMBANG WIDARSONOVOL. 33. NO. 3,DECEMBER 2010 : 165 - 179displacement is followed by water forced displace-ment to yield total produced oil volume of Vod (in-cludes Voi). Mathematically, the two indexes are ex-pressed as:WdWiOVVI = (3)andodoiWVVI = (4)Forced displacement is usually performed usingcentrifuge or core flow apparatus while imbibitionprocess is suggested to take at least 20 hours (Amott,1959) or much longer for rocks with neutral wettability(Anderson, 1986a).Interpretation using the two indexes is somewhatrelative in nature and there is no guideline for defini-tive judgment. Amott (1959) put 1.0 as strongwettability while a value of zero indicates neutralwettability, and values approaching zero are indica-tion of preferential wettability. Inclination towardseither wettability is judged from relative comparisonsbetween the two indexes.When wettability is put asO WI I then the Amott wettability index would varyfrom +1 for absolute water wet to -1 for absolute oilwet with zero indicating neutral wettability.For the purpose of clear classification and com-parison with other wettability indicator techniqueswettability in this study is divided into strong oil wet,oil wet, preferential oil wet, neutral, preferen-tial water wet, water wet, and strong water wet.Table 2 presents value ranges for the wettability cat-egories. The established value ranges are indeed sub-jective in nature but their assignments are consid-eredappropriatetoaccommodatereasonablediscretization on gradation in the wettability strength.USBM Wettability Index. The technique basi-cally uses capillary curves obtained through displac-ingoilandwaterusingcentrifugeequipment(Donaldson et al. 1969). The displacement is peformedalternately in a way similar to forced displacementprocess in the Amott technique, in which a water-saturated sample is spun under various rotationalspeeds while immersed in oil to reach Swirr. The pro-cess is repeated by spinning the now oil-saturatedsample in water immersion. Capillary pressures arecalculated based on the known rotational speeds.The fundamental principle of the method is thatdisplacement of a non-wetting phase by a wettingphase requires less force than the reverse. This re-sults in different capillary pressure curves with theTable 2Value ranges established for wettability classification used in the study170AN INVESTIGATION OVER ROCK WETTABILITY AND ITSLEMIGAS SCIENTIFIC CONTRIBUTIONSBAMBANG WIDARSONOVOL. 33. NO. 3,DECEMBER 2010 : 165 - 179by other factors such as imbibing liquidviscosity, permeability, porosity, artificialtension, and samples edge condition(Tiab and Donaldson, 2004). Acknowl-edging these factors, Ma et al. (1999)used scalling correlation for evaluatingimbibition-driven oil recovery in frac-tured water-drive reservoir introducedby Mattox and Kyte (1962) for evalu-ating wettability in this Direct Imbibi-tion method. Nevertheless, the mostcommon judgment for establishingwettability type is through relative com-parison between imbibition rates ofwater and oil, and the reported conclu-sion for the two sandstone samples (seeTable 3) is used without any further re-view.Deeper description is not spentand index categorization is not estab-lished for this technique.Water-oilrelativepermeabilitycurves. As wettability tendencies af-fect capillary pressure curves in theform of hysteresis, the tendencies alsoaffect water-oil relative permeability curves. Basi-cally, a core flow test designed to obtain relative per-meability curves is meant to observe on how a par-ticular rock sample pore system influences the multi-phase flow behavior. With presence of different wet-ting inclination shown by different reservoir fluids,however, this porous medium fluid interaction isbiased. Different degrees of wettability lead to dif-ferent fluid saturating characteristics within the rockhence changing the effective permeability of the flu-ids present.Figure 3 illustrates changes in relative permeabilitydue to different wettability. In comparison, water-wetsystem and oil-wet system become different eventhough the shape of curves remains the same. Atcondition of oil-wet system the flow tends to be ofearlier water breakthrough due to easier movementsof water compared to oil. In this condition, the pointof Kro =Krw occurs at lower values of water satu-ration with higher values of Krw and lower Kro val-ues at most values of water saturation. Change inwettability towards more water wettability shifts theKro =Krw point to higher water saturation pointsdue to the fact that the water tends to lose mobilityhence requiring higher water saturation to enable itto move under the same pressure difference (Amyxet al. 1960; Archer and Wall, 1986). Anderson (1986b)discussed further in more depth the influence ofwettability on relative permeability curves.Wettability is indeed not the sole factor that caninfluence water-oil relative permeability curves. Intheir report on a series of experimental works Geffenet al. (1951) put that variation in overburden pres-sures and the resulting changes in pore size distribu-tion may provide blocking effect to the two liquidphases movements and shifts the relative perme-ability curves. Increases in temperature also changewettability towards a more water-wet tendency.These all imply that any test for relative permeabilityhas to be performed under reservoir condition (i.e.overburden pressure, pore pressure, and tempera-ture). However, common industrial practices in thisregard rarely meet this ideal condition for variousreasons including equipment limitation and simplic-ity. All relative permeability data used in this studywas obtained under atmospheric temperature. Thisis also the case for measurements on wettability, forwhich both imbibition and forced displacement pro-cesses were carried out under atmospheric tempera-ture. Similarity in testing condition for the threeFigure 3Shift in permeability curves intersects due to changein wettability system. Relative permeability of an oil-wetsystem (dashed curves) tend to show higher watereffective permeability leading curve intersect at lowerwater saturations. On the contrary, higher oil effectivepermeability in water-wet system (solid curves) tendsto yield intersects at higher water saturations171AN INVESTIGATION OVER ROCK WETTABILITY AND ITSLEMIGAS SCIENTIFIC CONTRIBUTIONSBAMBANG WIDARSONOVOL. 33. NO. 3,DECEMBER 2010 : 165 - 179wettabilityindicatorsthereforesuggeststhatwettability remains the sole predominant factor in theshift of relative permeability intersects (i.e. Kro =Krw).Through the use of this laterconclusion, shift in intersect be-tween the two curves can there-forebeusedasindicatorforrockswettability. Water-wetrocks tend to have curves inter-sect to be at water saturation val-ues lower than 50%, and the re-verse is true for oil-wet system.No clear guideline has been givenby past studies regarding valuesor value ranges that representcertain degrees of wettability. Itis logical, however, that neutralwettability systems would havecurve intersect at around 50%water saturation, and strong wet-ting tendencies at water satura-tion values approaching Swirr andresidual oil saturation (Sor) for wa-ter-wet and oil-wet systems, re-spectively. Gradual degrees inwettingtendenciesforbothwettability systems naturally fallbetweenneutralandthetwostrong wetting tendencies.In order to make this indirectwettability indicator comparabletwo the other two standard tech-niques discussed earlier, a clearguideline is needed. Similarly, the seven-classwettability divison used for the other two techniquesare also used here, with water saturation ranges rep-resenting the permeability curves intersects as ref-Permeability Porosity(mD) (%) W-wet O-wet Interpretation1 23 25.4 0.4167 0.0000 0.4167preferential water-wet10 3251 32.5 0.4355 0.1719 0.2636preferential water-wet13 772 31.4 0.3352 0.0789 0.2563preferential water-wet19 22 22.6 0.4800 0.0000 0.4800preferential water-wet20 9 25.5 0.4857 0.4113 0.0744neutralSample numberWettability IndexI ATable 4Result example of wettability test using Amott technique.The generally preferential water-wet rocks are from BK 232 well, Central Sumatra BasinFigure 4Figure 4 Example of USBM wettability test graphical result for a coresample taken from T 105 well, Barito basin.The test yields I = log(Al/A2) value of -0.140 indicating preferentialtendency towards oil wetness (preferential oil-wet)172AN INVESTIGATION OVER ROCK WETTABILITY AND ITSLEMIGAS SCIENTIFIC CONTRIBUTIONSBAMBANG WIDARSONOVOL. 33. NO. 3,DECEMBER 2010 : 165 - 179Permeability Porosity(mD) (%)276 147 24.7 -0.336 oil wet265B 47 21.8 0.106 preferential water wet217 29 23.6 -0.346 oil wet216 34 24.6 -0.392 oil wet119B 844 28.1 -0.199 preferential oil wet105 62 27.5 -0.140 preferential oil wetInterpretation Sample No.|.|

\|=21AAlog ITable 5Result example of wettability test using USBM technique.The generally oil-wet rocks are from T 105 well, Barito BasinFigure 5Three pairs of relative permeability curves (solid and dashed ones for Kro and Krw,respectively) taken from;a) PP-CC5 well (N Sumatera Basin),b) KW P6 well (NE Java Basin), and FW-2 well (NW Java Basin). In accordance with thecriteria established in this study, the three exemplary data sets tend to exhibit wettabilitytendencies of oil-wet, neutral, and strong water-wet, respectively.173AN INVESTIGATION OVER ROCK WETTABILITY AND ITSLEMIGAS SCIENTIFIC CONTRIBUTIONSBAMBANG WIDARSONOVOL. 33. NO. 3,DECEMBER 2010 : 165 - 179erence. Table 2 presents thewatersaturationrangesas-signed to serve the purpose.IV.LABORATORYDATAWettability data used in thestudy was derived from variousLemigas Core Laboratory Re-ports of testing on 363 sand-stone core samples taken from28 oil and gas fields in Indone-sia. Table 3 presents list of datacovering sample origins andtheir type of wettability indica-tors. Amott technique appearsto make the bulk of wettabilitytestresults(113samples)among the three wettability testmethods while relative perme-ability, as a non-wettability testtechnique, is also available inevenlargernumber(224samples).All data was obtained fromLemigas Core Laboratory ar-chives and in the form of un-published reports. Amott test results are presented intabular form whereas the USBM and relative per-meability data is both tabular and graphical forms.Table 4 depicts an exemplary Amott test data (BK 232 well, Central Sumatra Basin) from which overallpreferential to water wetness is concluded. Table 5and Figure 4 present example (T 105 well, BaritoBasin) for USBM technique, the resulting I valuesindicate sufficiently strong inclination to be oil-wet.All wettability tests were performed using native cores i.e. uncleansed leading to results representingtheir unaltered wettability.For relative permeability data, most data avail-able to the study has complete curves to enable ob-servation on the curves intersects. Nevertheless, insome cases (less than 3% of overall data) with in-complete data, extrapolations were made so that thedesired information is obtained. Figure 5 exhibits threeexamples with three different wetness tendencies.All samples were cleansed using solvent prior to rela-tive permeability tests meaning that the resulting datais likely to represent un-restored or alteredwettability condition.Figure 6Wettability composition of the sandstone samples,which wettability test results are used in this study. Water-, neutral-, and oil-wet groups make 48.2%, 21.6%, and 30.2%of the total core samples, respectivelyV. ANALYSISIn analysing the data, observations were per-formed on two issues; original wettability as indicatedby wettability tests and wettability alteration due tocore cleansing.Originalwettability. In general, results fromthree wettability indicating techniques have exhibitedno strong preference towards specific wettabilitytypes. As depicted in Figure 6, preferential water-wet, water-wet, and strong water-wet are re-spectively represented by 39, 21, and 3 samples. Com-bination of these figures make 48.2% of all samplesare grouped into water-wetness tendency. On theother hand, combination of preferential oil-wet andoil-wet 32 and 10 samples, respectively tenden-cies establishes a correponding figure of 30.2% foroil-wetness tendency. No strong oil-wet result hasbeen observed.These oil-wet and water-wet compositions along with 21.6% of neutral wettability have shownthat Indonesian reservoir sandstones are not differ-ent to other sandstones from other places in the world.174AN INVESTIGATION OVER ROCK WETTABILITY AND ITSLEMIGAS SCIENTIFIC CONTRIBUTIONSBAMBANG WIDARSONOVOL. 33. NO. 3,DECEMBER 2010 : 165 - 179As put earlier, even though Blockand Simms (1967) showed that sili-cate glass tends to show strongwater-wetness but combined pres-ence of rock mineral impurities andoil pH preference as proved byTreiber et al. (1972) tends to ex-hibit even tendencies toward oil-and water-wetness (Table 1). Com-paringtheseresultsandthoseshown in Figure 7 comparable com-positions are obvious with strongsimilarity in water-wetness. Largeramount of samples on both sidesmay probably lead to more similarcompositions.Wettability alteration. As putearlier,coreplugsareusuallycleansed and extraxted of all saltsnormally present in native coresprior to measurement for rock ba-sic properties. This is often, and in-deed has become a recommendedpractice (API, 1960), for both prac-tical and objective reasons (e.g. airpermeability and helium porosityare measured on cleansed coreplugs). Therefore, it is expectedthat wettability alteration has oc-curred.In analyzing the alteration, as-sumptions are taken:1. For original wettability fromwettabilitytests,overallwettability of one sample set(i.e. from a well) is adoptedbased on majority in wettabilitytypeshownbythetestedsamples. This is due to the factthat samples used in wettabilitytests were not of same samplesused in relative permeabilitytest, even though they belongedto the same sample set. Thisoverall wettability was thencompared with relative perme-ability curves intersects fromindividual samples in order toobserve changes in wettability.Figure 7Wettability composition of samples that originallybelonged to strong water-wet class. Although most samplesstill retain water-wetness inclination some have lost theirpreferencetowater-wetnessFigure 8Wettability composition of samples that originally belongedto water-wet class. Most samples vave become neutralbut oddly enough some of them switch side into oil-wet group175AN INVESTIGATION OVER ROCK WETTABILITY AND ITSLEMIGAS SCIENTIFIC CONTRIBUTIONSBAMBANG WIDARSONOVOL. 33. NO. 3,DECEMBER 2010 : 165 - 1792. Relative permeability curves in-tersect (@ Kro =Krw) can beused as wettability indicatorbased on recognition that thepair of curves shift along watersaturation axis with changes inrock samples wettability type.3. The established index categori-zation for wettability classifica-tion serves well for the threewettability indicators (minus theDirect Imbibition technique) tojustify comparison among re-sults of all the four techniques.Using this three-point assump-tion, analysis was made through ob-serving the change of samples origi-nally belonging to each wettabilityclass. Figures 7 through 12 presentthe results.From the originally described asbelonging to a strong water-wetsample sets as indicated by thewettability tests no one of the 16samples tested for water-oil rela-tive permeability data indicatesstrongwaterdriveclassofwettability (Kro =Krw @ Sw >0.8)(Figure7).Thechangesinwettability, some samples still re-tain water-wetness at lesser de-grees, even extend to neutral (Kro=Krw @ Sw ~ 0.5) meaning thatthe samples of concern have lostaffinity tendency towards water(and also oil).Similarly to the case of strongwater-wet, all samples that origi-nally belonged to water wet cat-egory have degraded in wettabilitystrength against water (Figure 8)and most of the samples have be-come neutral and even switchside into the oil-wet group. Thecase is not entirely the same forpreferentially water-wet class, outof which some still retains theiroriginal wettability (18 samples)even though most of the samplesFigure 9Wettability composition of samples that originally belonged topreferentially water-wet class. Some samples retain their originalwettability but most samples have become neutral. Small portionof samples also become preferentially oil-wetFigure 10Wettability composition of samples that originally belongedto oil-wet class. None of the samples retain their originalwettability and most samples have degraded into softer wettability, and some even become inclined into the water-wet group176AN INVESTIGATION OVER ROCK WETTABILITY AND ITSLEMIGAS SCIENTIFIC CONTRIBUTIONSBAMBANG WIDARSONOVOL. 33. NO. 3,DECEMBER 2010 : 165 - 179have become neutral. Similarto the case of water-wet class,some of the originally preferen-tially water-wet samples havebecome preferentially oil-wet.The degradation in water-wet-ness due to core cleansing is un-derstandable, but change into oil-wetness indeed requires morethorough explanation.In the oil-wet group no origi-nally strong oil-wet samples areat disposal, which means onlytwo classes available; oil-wetand preferentially oil-wet. In amanner similar to the cases in thewater-wet class the samplesbelonging to oil-wet sampleshave degraded into preferentiallyoil-wet and neutral classeswithsomeevenswitchedwettability into more oil-wet ori-entation (Figure 10). A resem-blance in behavior to preferen-tially water-wet class samplesis also shown by its counterpartin the preferentially oil-wetclass.Many of the originallypreferentially oil-wet samplesretain their wettability while mosthave become neutral with theremaining few jump onto theother side of the wettability spec-trum (Figure 11).Although thiswettability switch occurred onlyon few samples (22% of total inthe class) this phenomenon re-quires attention.For neutral class (Figure12), the samples wettability be-havior differs significantly fromthe tendencies shown by thewettability groups on the twosides of the spectrum. This caseis characterized by the retainingof wettability by the bulk of thesamples (65% of total), and ifsamples of the two preferentialwetness are included on theFigure 11 Wettability composition of samples that originally belonged to preferentially oil-wet class. Similar to the case of preferentially water-wet class, many of the samples retain their original wettability and most became neutral.Some few samples have gone to oil-wet tendency, howeverFigure 12Wettability composition of samples that originally belongedto neutral class. Majority of samples remain neutral,and if samples of the two preferential wetness classes are includedon the ground of classification uncertainty this portionis even higher to reach 88.4% of total samples177AN INVESTIGATION OVER ROCK WETTABILITY AND ITSLEMIGAS SCIENTIFIC CONTRIBUTIONSBAMBANG WIDARSONOVOL. 33. NO. 3,DECEMBER 2010 : 165 - 179ground of uncertainty in boundaries between classes the portion is even higher (88.4%). This fact, com-bined with wettability degradation in strength aftercore cleansing, has led into a thought that rockwettability tends to move toward neutrality if thecauses of the original wettability have removed fromthe rocks surfaces.From individual analysis based on individualwettability class (Figures 7 through 12), overall fig-ures have shown that out of 224 water-oil relativepermeability samples only 67 (29.9%) retain their origi-nal wettability. If this group is expanded to includesamples that remain in their wettability group (e.g. awater-wet sample that was originally strong water-wet) the overall number becomes 80 (35.7%) only.These figures correspond to the total figure ofsamples that remain or become neutral after corecleansing of 110 or 49.1% of total samples. Thisfurther underlines the fact that cleansed core samplestend move toward neutrality in wetness tendency,along with all validity consequences on the data mea-sured afterward.VI.FURTHERDISCUSSIONAs put by Tiab and Donaldson (2004), rocks sur-face mineral composition and polar organic compo-nents in crude oil act as either weak basic or weakacidic compound depending on the amount of resinand aspalthene contents that can react to each otherto form a very thin layer of active compounds on therocks solid surface. This thin layer of active com-pound affects wetting characteristics of the rock-fluidsystem. During core cleansing prior to many labora-tory applications and tests, this thin layer is to be ei-ther completely or partially removed. The result isdegradation in wettability strength, or a full shift toneutral wettability if the thin layer is completely wipedout. This mechanism is likely to serve as an explana-tion over the wettability change commonly observedduring the study.One question related to wettability change re-mains. What actually causes the switch in wettability,from water-wetness to oil-wetness and vice versa?The only possible explanation at this stage is that hotsolvent (usually toluent and methanol) used in the coresample cleansing has somehow chemically reformedthe thin layer of wettability-affecting compound onthe rock samples surface to form an opposite wet-ting tendency. However, since this occurred on 50samples only (22.3% of the total 224 samples) andis further reduced to 10 samples (4.5%) if preferen-tially oil-/water-wet samples are excluded on classi-fication uncertainty ground this switch is likely tobe caused by other factor than reform of the thinlayer compound by hot solvent. Generalization oforiginal wettability on heterogeneous rocks and thefact that samples used in wettability tests are usuallydifferent from the ones used in water-oil relative per-meability test even though belonging to the samerock formation are probably the factors causingthis apparent wettability switch. Speculatively there-fore, the process of core cleansing using hot solventcauses degradation in wettability down to the point ofneutral wetness tendency at most.Regardless the real cause of change and switchin rock wettability, however, this occurrence may af-fect validity of the ensuing tests performed after thecore cleansing. As previously discussed changes inwettability affect relative permeability curves with allof its consequences. Furthermore, Widarsono (2008)pointed out in length the effect of wettability alter-ation on rock electrical properties, which in turnthrough well log analysis may affect severely anyestimation of water saturation. In the article, he alsounderlined the need to either restore rocks wettabilitythrough core-ageing or use cold core cleansing tech-nique that utilizes cold solvent flow in order to dis-solve oil and salts within samples. Through thesemethods, invalidity of laboratory test results causedby wettability alteration can hopefully be minimized.Attempts were initially made to see whether thereis any relation between wettability and sedimentarybasins. However, it was then realized that wettabilityis much governed by mineral and oil compositionrather than by any other factors specifically relatedto individual regions. Different basin may have accu-mulated similar minerals depending to depositionalenvironments to others, and vice versa. Nonethe-less, a more thorough study may have to be made inorder to investigate this matter.VII.CONCLUSIONSAnalyses and evaluations on all data used in thisstudy have led into some main conclusions, namely:- Like all reservoir sandstones throughout the worldIndonesian sandstones also tend to have bothwater-wet and oil-wet tendencies. Rock mineral-178AN INVESTIGATION OVER ROCK WETTABILITY AND ITSLEMIGAS SCIENTIFIC CONTRIBUTIONSBAMBANG WIDARSONOVOL. 33. NO. 3,DECEMBER 2010 : 165 - 179ogy appears to play an important role in determin-ing wetness characteristics.- Core cleaning, as a standard practice in labora-tory core analysis tends to weaken wettabilitystrength, which results in wettability degradation.However, complete change in wettability is likelyto reach no further than neutral wettability.- Wettability switch from water-wet to oil-wetand vice versa due to core cleansing probablydoes not occur. However, if it actually does morethorough study and investigation are required forbetter understanding.- Weak wettability i.e. preferentially oil-wet andpreferentially water-wet appear to be more re-silient against wettability change. This is likely dueto actual similarity between weak- and neutralwettabilities in a way that external factors suchas core cleansing cannot change much.- Strong proof that standard laboratory core han-dling (i.e. core cleansing) changes rock samplewettability has emphasized the need to utilize nec-essary measures to prevent/minimize its occur-rence. Core-ageing and cold core cleansing areamong the suggested methods to serve the pur-pose.REFERENCES1. Amott, E. (1959). Observation relating to thewettability of porous rock. Trans. AIME, Vol.216, pp. 156 162.2. Amyx, J .W., Bass J R, D.M & Whiting, R.L.(1960). Petroleum reservoir engineering: Physi-cal properties. McGraw-Hill Book Co., NewYork, p. 610.3. Anderson, W.G. (1986a). Wettability literaturesurvey Part 2: Wettability measurement. Soc.Petrol. Eng. J PT vol. 38, pp. 1246 1262.4. Anderson, W.G. (1986b). Wettability literaturesurvey Part 5: The effects of wettability onrelative permeability. Soc. Petrol. Eng. J PT vol.38, pp. 1453 1468.5. API (1960). Recommended practice for coreanalysis procedure API RP 40. The Ameri-can Petroleum Institute, August.6. Archer, J .S. and Wall, C.G. (1986). Petroleumengineering: Principles and practice. Graham &Trotman Ltd, Sterling House, 66 Wilton Road,London SW1V 1DE, UK, p.362.7. Block, A. & Simms, B.B. (1967).Desorptionand exchange of absorbed octadecylamine andstearic acid on steel and glass. J . Colloid andInterface Sci., Vol. 25, p.514.8. Chilingarian,G.V. & Yen, T.F. (1983). Some noteson wettability and relative permeabilities of car-bonate rocks. Energy Sources, Vol. 7, No. 1,pp. 67 75.9. Denekas, M.O., Mattax, C.C. and Davis, G.T.(1959). Effect of crude oil composition on rockwettability. Trans. AIME, vol. 216, pp. 330 333.10. Donaldson, E.C., Kendall, R.F., Pavelka, E.A. andCrocker, M.E. (1969). Wettability determinationand its effect on recovery efficiency. Soc.Petrol. Eng. J ., Vol. 9, No. 1, March, pp. 13 20.11. Geffen, T.M., Owens, W.W., Parrish, D.R. &Morse, R.A. (1951). Experimental investigationof factors affecting laboratory relative permeabil-ity measurements. Trans. AIME, vol. 192, pp.99 110.12. Ma, S., Morrow, N.R. and Zhang, X. (1999).Characterization of wettability from spontane-ous imbibition measurements. J . Can. Petrol.Tech. (Special Edition), Vol. 38, No. 13, p. 56.13. Mattox, C.D. and Kyte, J .R. (1962). Imbibitionoil recovery from fractured water drive reser-voir. SPEJ , J une, pp 177 184.14. Mennella, A., Morrow, N.R. and Xie, X. (1995).Application of the dynamic Wilhelmy Plate toidentification of slippage at a liquid-liquid-solidthree phase line of contact. J PSE, vol. 13, Nov.,pp. 179 192.15. Tiab,D.&Donaldson,E.C.(2004).Petrophysics: Theory and practice of measur-ing reservoir rock and fluid transport properties.Gulf Professional Publishing, 200 Wheeler Road,Burlington, MA 01803, USA, p. 889.16. Timmerman, E.H. (1982). Practical reservoir en-gineering Methods for improving accuracy orinput into equations and computer programs.PennWell publishing Company, 1421 SouthSheridan Road, Tulsa Oklahoma 74 101, p. 365.17. Treiber, L.E., Archer, D. & Owens, W.W. (1972).A laboratory evaluation of the wettability of fiftyoil producing reservoirs. Soc. Petrol. Eng. J .,Vol. 12, No. 6, December, pp. 531 540.179AN INVESTIGATION OVER ROCK WETTABILITY AND ITSLEMIGAS SCIENTIFIC CONTRIBUTIONSBAMBANG WIDARSONOVOL. 33. NO. 3,DECEMBER 2010 : 165 - 17918. Widarsono,B.(2008).PerubahanSifatKebasahanFluidadanSifatKelistrikanBatuanReservoir:IsuLama,ProblemAktual(Change in Reservoir Rocks Wettability and ItsInfluence on Electrical Characteristics: Old Is-sue, Ever Present Problem). (in Bahasa Indone-sia). Lembaran Publikasi LEMIGAS, Vol. 42, No.1, April, pp: 20 - 28.180TRACER TESTS FOR HETEROGENEITY CHARACTERIZATION LEMIGAS SCIENTIFIC CONTRIBUTIONSSUGIHARDJ O, ET AL.VOL. 33. NO. 3, DECEMBER2010 : 180 - 187TRACER TESTS FOR HETEROGENEITYCHARACTERIZATION ANDSATURATION DETERMINATION ON CORE FLOODINGBy: Sugihardjo1), Usman1), and Utomo Pratama I.Researcher1)at LEMIGAS R & D Centre for Oil and Gas TechnologyJ l. Ciledug Raya Kav. 109, Cipulir, Kebayoran Lama, J akarta Selatan 12230, INDONESIATromol Pos: 6022/KBYB-J akarta 12120,Telephone: 62-21-7394422, Faxsimile: 62-21-7246150First Registered on 3 November; Received after Corection on 22 November 2010;Publication Approval on : 31 December 2010ABSTRACTLowsweepefficiencyisthecommonproblemindisplacementprocessduetoheteroge-neity,highpermeabilitystreaks,fractures,andthiefzonesexistingintheformation.Simi-larly,thesuccessorfailureofEORimplementationsarealwaysaffectedbythoseproblemswhichcausesdisplacingfluidsfingeringandearlybreakthrough.Factorsofthistype,unlessproperlyidentifiedandunderstoodbeforethestartofEORprocess,willlikelycauseaprojectfailure.Corefloodingasthemodelofsmallscaleoffluidsmovementsinreservoirundergoessimilarcircumstances.Approximatelyonefootlongoffour3.5inchesstackednativeandsyntheticcoresarenormallyusedincorefloodingexperiment.TracertestwasperformedtocharacterizethecoreinadditionofCTscananalysis.Onthisexperiment,lithiumsolu-tionwasselectedastracersolutiontobetheninjectedintocoreatconstantrate,4ft/day.Afterwards,theeffluentswerecollectedbyGilsonsamplecollectorineachtubeforfurtherdetermininationitsconcentrationusingAtomicAbsorptionSpectrometry(AAS).Responsecurvesoflithiumtracerwereabletodeterminecoreheterogeneitiesandthisshouldbedonetoavoidmisleadinginterpretationofcorefloodingresults.Besides,lithiumconcentrationreportedinsomeextentandsubsequentlyanalyzedbyemployingmethodoftemporalmoments.Thismethodprovidesnumericalcalculationtoestimateeffectivecoreporevolume(PV)andfluidsaturation.WeighingmethodwasalsousedtocomparethePVwithaforementionedmethodandtheresultswerecomparable.KeyWords:Tracer,heterogeinity,fluidsaturation,andcorefloodingI.INTRODUCTIONEOR is the only technology which is capable forproducing the remaining of oil in the reservoirs afterprimary and secondary recovery processes. Successof secondary and tertiary oil recovery projects tar-geting the remaining oil in mature or partially depletedreservoirs strongly depends on adequate descriptionof reservoir heterogeneity. Processes that are well-understood in a laboratory environment and those alsoshould be properly designed for the reservoir scale.A number of procedures exist that can be used be-fore implementation of an EOR process in attempt todescribe the reservoir geology. One of these proce-dures is tracer test.Tracer technology plays an important role in im-proving the reservoir characterization before the ap-plication of EOR methods by providing qualitativeinformation on reservoir compartmentalization, pref-erential flow paths to improve understanding of fluidmovement in the reservoir, stratification, and hetero-geneities distribution. Basically, tracer is chemicalsthat can be added to fluids in small concentrationsand used to follows their movement without affect-ing their physical properties. It also can be used asan effective tool to detect and estimate of remainingoil saturation using two different types of tracers. Thetwo tracer types are differentiated into the conser-vative (ideal) tracer and the partitioning tracer. The181TRACER TESTS FOR HETEROGENEITY CHARACTERIZATION LEMIGAS SCIENTIFIC CONTRIBUTIONSSUGIHARDJ O, ET AL.VOL. 33. NO. 3, DECEMBER2010 : 180 - 187ideal tracers do not have solubility in other substances,this case in oil or by definition they do not interactwith the rock or other fluid phases present. Thus whenideals are injected in reservoir, they will flow only inthe water phase adopting the velocity of this phase.Some examples that can be used as ideal tracer areiso-propyl alcohol (IPA), bromide and lithium.In contrast, partitioning tracers are soluble in liq-uid hydrocarbon as well as water or gas phases. Themolecules of the partitioning tracers are moving backand forth between the water and oil phase, becausethey have high partition coefficient (absorb into therock) which determines the tracer solubility in otherphases. Consequently, when a pulse of aqueous so-lution containing a suite of partitioning tracers is in-jected into an oil reservoir, the tracer will continu-ously partition into and out of the oil phase contactedby aqueous solution (injected gas). Hence the mol-ecules of partitioning tracers are flowing with thewater velocity when they are in the water phase andoil velocity when they are in oil phase6. Thus, parti-tioning tracers propagate more slowly in an oil reser-voir than conservative tracers. This retardation of thepartitioning tracer is analogue to chromatographicseparation where this mechanism is utilized to esti-mate oil saturation in the reservoir. Several examplesfrom this type of tracers are n-butanol, rhodamine,and propanolTracer test is necessary to be performed in coreflooding experiment when we are deeply concernedwith core properties and validation seeking. BesidesFigure 2CT Scan on Stacked Core PlugsFigure 1Top Synthetic and Bottom WrappedStacked Native Cores182TRACER TESTS FOR HETEROGENEITY CHARACTERIZATION LEMIGAS SCIENTIFIC CONTRIBUTIONSSUGIHARDJ O, ET AL.VOL. 33. NO. 3, DECEMBER2010 : 180 - 187describing the heterogeneity within core including itsconnectivity, another advantage is as leak indicatorin core set up. Moreover, we could obtain effectivepore volume swept by the tracer.This paper will describe the results of tracer teston several core flooding succinctly including how touse the method of temporal moment analysis fromtracer response curve data to estimate effective corepore volume.A quick look observation on responsecurve to reflect core heterogeneity and experimentalset up was also presented concisely.II. COREPREPARATIONFORTRACERTESTSA native core sample was taken from the inter-est zone. Approximately one foot long of four 3.5inches stacked native core are normally used for rep-resenting the reservoir rock. In case of unavailablityof native core, a standard or synthetic core suchClasshach, Brial Hill, and Berea sands can be usedjust only for determining the efficiency of displacingmaterial for EOR project without any results of fluid-rock interaction. Running core flood using stackedcores is much more crucial and common misleadinginterpretation of the core flood results occurs unlessprior heterogeneity determinations by tracer test. Fig-ure 1 shows the two types of core i.e. stacked nativeand synthetic cores.Four 3.5 inches native core plugs were drilled forthis core tracer test. Prior to performing core floodtest, a CT scan must be done to sproperly select thecores for stacking which have similar qualitativelyrock properties and avoiding fractures and shale lami-nation. Figure 2 is the CT scan results of the candi-date cores for tracer test.Figure 3Coreflood Equipment Applied for Tracer Test183TRACER TESTS FOR HETEROGENEITY CHARACTERIZATION LEMIGAS SCIENTIFIC CONTRIBUTIONSSUGIHARDJ O, ET AL.VOL. 33. NO. 3, DECEMBER2010 : 180 - 187III.TRACERINTERPRETATIONSMETHODA host of tracer analysis methods consider thetemporal behavior of tracers. The methods were origi-nally developed for closed reactor vessel, but havebeen applied to more general conditions of openboundaries, characterization of fractured media un-der continuous reinjection and to estimates flow ge-ometry. The methods have rigorous mathematicalbasis. The methods and application mentioned aboveare all based on analysis of tracer residence times.The mean residence time or first temporal moment,is the most useful single property derived from tracertest, although other properties have been used as well.Levenspiel10 shows the total pore volume swept bythe tracer can be determined from its residence time.The method of temporal moments is a very simple,fast and robust method to estimate swept pore vol-ume and remaining oil saturation. As explained ear-lier, to calculate remaining oil saturation we need twodifferent types of tracer where the ideal tracer be-haves as the reference tracer and the partitioningtracer as the partitioned one. Because of the pres-ence of oil, partitioning tracers are retarded compoundto the non-partitioning. However, this paper aimsmerely on effective pore volume estimate.Effective pore volume can be estimated using onetracer that is swept by tracer. The pore volume wasdetermined from tracer mean resi-dence time which required steps aresummarized as follows:- Normalize the tracer history- Extrapolate the history to latetime- Calculate mean residence timeand swept volumeAlthough this method providessome advantages but it is necessaryto inform that moment analysis is ageneral method, and one that suf-fers from few limitations. Assumedconditions essentially state that theflow field is steady and tracersmoves with bulk fluid flow such thatthe information obtained from theanalysis is general bynapplicable.These conditions can be stated asfollows:1. Steady state injection and extraction2. The tracer is ideal and conservativeIV.EXPERIMENTALPROCEDURESeveral tracer tests were conducted by using highpressurized core cell with low dead volume. The ex-periment used both standard and native sandstonecore that condition set to reservoir state (Table 1).The core was prepared to contain residual oil satura-tion using standard procedure by altering flow of brineand crude oil. The experiment configuration is illus-trated in Figure 3. Core length 27.18 cm Core diameter 3.8 cm Pore volume 115.23 cc (?) Porosity 30.08 % Flow rate 0.3 cc/min Velocity 4 ft/day Temperature85 oC Mobile phase Synthetic brine Volume of tracer injected 24.36 ccFigure 4Tracer Response CurveTable 1General Experimental Data184TRACER TESTS FOR HETEROGENEITY CHARACTERIZATION LEMIGAS SCIENTIFIC CONTRIBUTIONSSUGIHARDJ O, ET AL.VOL. 33. NO. 3, DECEMBER2010 : 180 - 187First tracer test was conductedwhen the core was saturated fullywith the formation water. This firststep of core flooding experiment isto identify the heterogeneity of frac-tures in the core. In this stage a smallvolume of tracer was injected intothe flowing water phase close to thecoreinlet.OnthisexperimentLithium (Li), dissolved in LiCl solu-tion, was selected as conventionaltracer. The solution had been de-signed at 100 ppm Li+ ion. Water-flood resumed at stable flow rateuntil the tracer chemical was elutedthrough the core plug, thus the ma-jority of the injected tracer masscould be recovered.Similarly, when residual oil satu-ration had been established to quan-tify the right residual oil saturationafter water flood, the core wasflooded also with synthetic brine at constant flow rate,4 ft/day. During this period, a small volume of tracerwas injected into the flowing water phase. The sametracer test was also performed after chemical flood,again to calculate the residual oil saturation afterchemical flood and to recheck the oil recovery factorat chemical flood, although a dean stark could be runon core after flood to make a comparison.Aliquots of 5 ml of the effluent were continu-ously collected by Gilson sample collector fractionclose to the outlet of the backpressure regulator valve.Then each sample fractions were analyzed for itscontent of Li by atomic absorption spectrometry(AAS).V.RESULTSEach sample concentration was plotted versusits volume, as depicted in Figure 4. The calculation ofeffective swept pore volume will be determined fromtemporal moment method. The first step is normaliz-ing the concentration history by dividing measuredLi+ concentration to total Li+ injected concentration.Tracer response curves should be complete interms of outflow measured concentration in order toestimate effective pore volume precisely, becausemuch of the information is contained in the tails ofthe response curve. Unfortunately, the tracer responsecurve is often incomplete either due to dilution of thetracer concentration below detectable limit of appa-ratus or some other reasons. Therefore to overcomethis difficulty is with extrapolating the history to longtime. The tracer response curve can be extrapolatedwith an exponential function provided the derivationof the test is sufficient to establish this decline.The first moment of the tracer response curveswas obtained by dividing the data into two parts. Thefirst part represents the data from zero to the time tbwhere time becomes exponential, and the secondcovers the exponential part in which it goes from tbto infinity.After time tb, the tracer response is assumed tofollow an exponential decline given by:||.|

\| =at tbbe C C........................................ (1)4Where 1/a is slope of the straight line when thetracer response curves are plotted in semi-log scaleand Cb is the tracer concentration at time tb whencurve becomes exponential. The extrapolatedcurveis seen in Figure 5.The mean residence time, or first temporal mo-ment, of tracer is determined directly from the nor-malized and extrapolated tracer history as:Figure 5Li Extrapolated Curve185TRACER TESTS FOR HETEROGENEITY CHARACTERIZATION LEMIGAS SCIENTIFIC CONTRIBUTIONSSUGIHARDJ O, ET AL.VOL. 33. NO. 3, DECEMBER2010 : 180 - 187}}++ +=bbbbtattbateabCdtat eabtCdtt002) 1 (* ...................(2)4The constants a, b and tb aredetermined by curve-fitting latetime tracer data in spreadsheet.Pore volume estimates follow di-rectly from the mean residence timeas describe from the equation be-low:* t qMmVinjinjp =................(3)4Then each data was tabulatedand plotted against cumulative vol-ume as shown in figure 6. The ef-fective pore volume calculationfrom temporal moments and weigh-ing method is 118.14 cc and 115.23cc respectively. There is a fair dif-ference between the estimation ofeffective pore volume using tempo-ral moments and weighing method.A.CoreCharacterizationFigure 4 shows the tracer re-sponse and indicated that the coreis homogeneous curving with thesingle peak and having almost simi-lar front tail and end tail formation.But Figure 7 is heterogeneouscorereflectedfromtwopeakswhich formed in tracer responsecurve. These peaks mean the corewas stratified into different flowpaths.Another case depicted in Fig-ure 8 shows scattered and varyingnoise in response curve that mayindicate the leakage occurred dur-ing the flooding due to imperfectcore set up particularly in coreholder sleeve.B.SaturationDeterminationTracer test can be used in vali-dating fluid content during corefloodexperiment. We can estimate effective pore volumefrom first tracer response curve which is obtainedprior saturating the core with oil, this can be simpli-fied by following simple term:Effective Swept PV Tracer Response Curve. ..(4)Figure 7Heterogeneity Core Response CurveFigure 6Pore Volume Estimation Curve186TRACER TESTS FOR HETEROGENEITY CHARACTERIZATION LEMIGAS SCIENTIFIC CONTRIBUTIONSSUGIHARDJ O, ET AL.VOL. 33. NO. 3, DECEMBER2010 : 180 - 187Obtained effective swept PVform above term can be defined aseffective PV 1.Moreover,sequentialtracertests can give us estimates on re-sidual oil saturation (Sor) and recov-ery factor. To obtain estimates onSor secondary tracer test needs tobe undertaken after waterflood. Atthis stage we obtain another tracerresponse curve, herewith can bedefined as effective swept PV 2. Bysubtracting effective swept PV 1with effective swept PV 2 we canobtain Sor. This can be simplified byfollowing equation:SorValidation=Eff.PV1-Eff.PV2....................................... (5)RecoveryFactorValidation=Eff.PV3-Eff.PV2 ................(6)Third tracer test conducted af-ter chemical flood can be used alsoin validating oil recovery.Figure 9 exhibits the three tracertests responses running at initial con-dition, after water flood, and afterchemical flood. Then the saturationon each flood stephasbeencal-culated preciselyandshowninTable 2.VI.CONCLUSIONS1. Effective pore volume estimatesfrom temporal moments andweighing method are 118.14 ccand 115.23 cc, respectively.2. Tracer test provides helpful toolsto improve core characterizationby providing qualitative informa-tion on preferential flow, strati-fication, core connectivity, het-erogeneities distribution and ap-paratus set up.3. Method of First Temporal Mo-ment Analysis is simpler andfastertointerprettracerre-sponse curve.Figure 8Leakage on Core Holder Set up Response CurveFigure 9Tracer Response Curves at Each Step of FloodTable 2Saturation DeterminationSynthetic 60.95 39.05 26.24 80.65Native 61.54 38.46 27.59 95.76RFEOR%ROSCoreSample Soi%PV Swc%PVROSWF%PV187TRACER TESTS FOR HETEROGENEITY CHARACTERIZATION LEMIGAS SCIENTIFIC CONTRIBUTIONSSUGIHARDJ O, ET AL.VOL. 33. NO. 3, DECEMBER2010 : 180 - 1874. Tracer test possible to be used as initial assess-ment to core before proceeding core flooding.Hence, it will save times and cost in experimen-tal.VII.NOMENCLATUREC = tracer concentration, ppmq = flow rate, cc/minM = total tracer injected, ccROS = residual oil saturationRF = recovery factorREFERENCES1. Tang, J .S.,Extended brigham model for remain-ing oil saturation measurement by partitioningtracer test, SPE 84874.2. Sinha, R., K. Asakawa K, G.A. Pope, and K.Sepehnoori,Simulation of natural and partition-ing interwell tracer test to calculate saturation andswept volume in oil reservoirs, SPE 89458.3. Illiasov, P.A., A.D. Gupta, and D.W. Vasco,Field-Scale Characterization of Permeability andSaturation Distribution Using Partitioning TracerTests: The Ranger Field, Texas, SPE 71320.4. Shook, G. M., J . Hope Forsmann, Tracer Inter-pretation Using Temporal Moments on a Spread-sheetI dahoNationalLaboratorydocu-ments,(2005)5. Asakawa, K., A generalized analysis of parti-tioning interwell tracer test, dissertation, univer-sity of texas, (2005).6. Chatzichristos, C.,. Dugstad, A. Haugan, J .Sagen, J. Muller,Application of Partitioning Trac-ers for Remaining Oil Saturation Estimation: AnExperimental and Numerical Study, SPE 59369.7. J in,M.,R.E.J ackson,G.A.Pope,S.Talfinder,Development of partitioning tracer testfor characterization of non-aqueous phase liquidcontaminated aquifers, SPE 39293.8. Abidin, Z., Teknologi Perunut Untuk ManajemenReservoir Minyak bumi (EOR), Pusat TeknologiAplikasi Isotop dan Radiasi, BATAN.9. Bailey, R.E., and L.B. Curtis, Enhanced oil re-covery, National Petroleum Council, (1984).10. Levenspiel, O., @Chemical Reaction Engineer-ing@, 2nd edition, New York: John Wiley and Sons,Chapter 9, (1972).188MODELING GRAVITY SEGREGATION IN STRATIFIE