1.INTRODUCTIONAninvestigationofhistoricallostcirculationevents,particularlyindeepwaterenvironments,ledtotheconsiderationofthermaleffectsonformationstressasapossiblecause.Theevidencesupportedthepossibilitythathighdifferentialtemperaturesbetweenthedrillingmudandtheformationsmightbecausinghighthermally-inducedtensilestresses,reducingtheformationbreakdownpressuresufficientlytoallowtheformationofopenfracturesatthemudpressuresthatwereusedtodrillthewells.Inordertotestthetheoreticalpredictionsoftemperatureeffectsonnear-wellboreformationbreakdownpressures,aseriesofthreeleakofftestswasperformedatdifferentwelltemperaturesinanonshorewellinsouthTexas.Thegoalofthetestswastodetermineiftemperaturehadasignificanteffectonthebreakdownpressurebydirectlymeasuringthemagnitudeofthechangeoveratemperaturerangeof33oC(60oF).Thispaperpresentstheresultsoftheexperimentalongwithresultsofattemptstomodelthemeasurementsusingalinearelasticmodelwiththermalstressesandathermoporoelasticmodel.2.EFFECTOFTEMPERATUREONSTRESS2.1.PreviousPublishedWorkThebasicconceptsrelatingtotheeffectoftemperatureonstressesaroundacylindricalboreholehavebeenknownformanyyears.TimoshenkoandGoodier[1]describedtheeffectofthermalstressesaroundaninfinitelylongcylindercontainingacircularholein1951.Init,theyARMA/NARMS04-527EffectofDrillingFluidTemperatureonFractureGradient:FieldMeasurementsandModelPredictionsGregoryPepin,ManuelGonzalez,J.BenBloys,JohnLoftonChevronTexacoEnergyTechnologyCompanyJosephSchmidtFracDogs,Inc.CareyNaquin,ScotEllisLandmarkGraphicsCorp.Copyright2004,ARMA,AmericanRockMechanicsAssociationThispaperwaspreparedforpresentationatGulfRocks2004,the6thNorthAmericaRockMechanicsSymposium(NARMS):RockMechanicsAcrossBordersandDisciplines,heldinHouston,Texas,June5–9,2004.ThispaperwasselectedforpresentationbyaNARMSProgramCommitteefollowingreviewofinformationcontainedinanabstractsubmittedearlierbytheauthor(s).Contentsofthepaper,aspresented,havenotbeenreviewedbyARMA/NARMSandaresubjecttocorrectionbytheauthor(s).Thematerial,aspresented,doesnotnecessarilyreflectanypositionofNARMS,ARMA,CARMA,SMMR,theirofficers,ormembers.Electronicreproduction,distribution,orstorageofanypartofthispaperforcommercialpurposeswithoutthewrittenconsentofARMAisprohibited.Permissiontoreproduceinprintisrestrictedtoanabstractofnotmorethan300words;illustrationsmaynotbecopied.Theabstractmustcontainconspicuousacknowledgementofwhereandbywhomthepaperwaspresented.ABSTRACT:EvidencefromseveralChevronTexacodeepwaterwellssuggeststhatreductionoffracturegradientsduetocoolingaroundthewellboremaybeamajorcontributortolostcirculationeventscausingmillionsofdollarsinadditionalwellcosts.Afieldtestofthemagnitudeoftheeffectoftemperatureonthefracturegradientwasperformedinashallowsand/shalesequenceinaSouthTexaswell.Pre-testmodelingofthermaleffectsusingasimpleelasticmodelpredictedthatanincreaseinmudtemperatureof33oCwouldresultinanincreaseintheminimumstressofapproximately3.1MPaatadepthof915m.Aseriesofleakofftestswasperformedatdifferentmudtemperaturestotestthepredictedeffectoftemperatureonthefracturegradient.Thetestsshowedthatanincreaseinthemudtemperatureof33oCcausedanincreaseinthefracturegradientof1.0MPaatthetestdepth.Thisresultismuchlessthanpredictedbytheelasticmodelandabouthalftheeffectpredictedbyathermoporoelasticmodel.Differencescouldbeduetoexperimentalconditions,difficultyinobtainingaccuratevaluesformanymodelparameters,orinadequatemodelboundaryconditions.showedthatheatingtheinnerwallofthecylinderwillresultinanincreaseinthecompressiveforcesaroundthehole.PerkinsandGonzales[2,3]showedthatinjectinglargevolumesofliquidthatiscolderthanthein-situreservoirtemperaturecansignificantlyreducethefracturingpressuresintheformation.TangandLuo[4]presentedmodelpredictionsoftheeffectonthenear-wellborestressesofdifferencesintemperaturebetweenthemudinthewellboreandtheformation.Morerecentmodelswhichcombinethermaleffectswithporoelasticityhavebeenpresentedbyseveralauthors[5-8].Themodelsassumethattheheattransferfromthewellboreintotheformationisduesolelytoconduction,ignoringtheeffectsofconvectiveheattransferfromfluidmovement.Inaddition,manyofthemodelsassumeinstantaneouschangesintemperatureandpressurewithinthewellborewhenthewellisdrilled(time=0),withaconstantwellboretemperatureandpressureassumedforalltimesafterthecreationofthewellbore.Anumberofpredictionsofthetimedependenteffectsofmudtemperatureonwellborestabilityhavebeenmadeusingthethermoporoelasticmodels.Thedifferentauthorsconsistentlypredictthatthermaleffectsaregreatestatearlytimesafterthewellisdrilled,withthemagnitudeofthethermaleffectdecreasingandthelocationofthethermally-inducedinstabilitiesmovingawayfromthewellboreatlongertimes[5].Bothelasticandporoelasticmodelspredictthatmudwhichiscolderthanthereservoirtemperaturewillinhibitshearfailureoftheformation.Theporoelasticmodelsalsopredictthatthebeneficialresultofcoolingwilldissipateatlongtimes,resultinginapossibilityoftime-delayedformationcollapse[5,7].Wellboretemperatureshotterthantheformationtemperaturearepredictedtoincreasetheriskofshearfailure