Copyright 2006, Society of Petroleum Engineers This paper was prepared for presentation at the 2006 SPE Annual Technical Conference and Exhibition held in San Antonio, Texas, U.S.A., 2427 September 2006. This paper was selected for presentation by an SPE Program Committee following review of informationcontainedinanabstractsubmittedbytheauthor(s).Contentsofthepaper,as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correctionbytheauthor(s).Thematerial,aspresented,doesnotnecessarilyreflectany position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPEmeetingsaresubjecttopublicationreviewbyEditorialCommitteesoftheSocietyof Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the Society of Petroleum Engineers is prohibited.Permissiontoreproduceinprintisrestrictedtoanabstractofnotmorethan300words;illustrationsmaynotbecopied.Theabstractmustcontainconspicuous acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435. Abstract Acidfracturingisperformedtoimprovewellproductivityin acid-solubleformationssuchaslimestone,dolomite,and chalk.Hydrochloric acid is generally used to create an etched fracture,whichisthemainmechanismformaintainingthe fracture open during the life of a well.Proppant fracturing is analternativeoptionthathasbeenappliedincarbonate formations.Incertainareas,proppantfracturinghasbeen usedasastandardstimulationmethodforcarbonate formations.Thereisnoquantitativemethodtoprovidean answer of whether acid fracturing or proppant fracturing is an appropriatestimulationmethodforagivencarbonate formation. In proppant fracturing, proppant is used to sustain the effect of theminimumhorizontalstressfromclosingthefracture.In acid fracturing the etched, non-smooth, surface with sufficient roughnessshouldleaveopenchannelsuponclosing.The effectofelastic,plastic,andcreepingdeformationsinacid fracturingandtheproppantcrushingandembedmentin proppantfacturing,onreducingfracturepermeabilityis investigated.Theviscouseffect,creeping,isaslow displacementthatincurredoveralongperiodoftime.The creeping effect on fracture closure following an acid fracturing treatment is demonstrated in this paper. Laboratory experiments have been performed to simulate acid andproppantfracturingtreatments.Theeffectofelastic, plastic and viscoelastic rock behavior on fracture conductivity was studied for acid and proppant fracturing treatments, using full core samples. Comparison of acid vs. proppant fracturing conductivity in carbonate formation is also presented. Introduction Hydraulicfracturing(acidorproppant)isusedtocreatea conductivefractureintheformationtoenhancewell productivity. The induced fracture will tend to close due to the effectoftheminimumhorizontalstress.Fractureclosureis controlledbyelastic,plastic,andviscousrockproperties.In acidfracturingtheetched,non-smooth,fracturesurfaces wouldleaveopenpathwaysuponclosinginadditionto wormholesandchannelscreatedfromthefractureintothe formation.Fractureconductivityisgeneratedbythequantity ofrockremovedandthepatternofrockremoval.Depending on the pattern of natural fracture system, acid solubility of the formation,magnitudeoftheminimumhorizontalstress,and reservoirtemperature,acidfracturingvs.proppantfracturing shouldbeevaluatedtoselectthemosteffectivestimulation treatment for a given formation. Interestingobservationsrelavanttostimulationof carbonatereservoirs,havebeenreportedintheliterature. Fractureconductivitydoesnotincreasewithincreasing amountsofdissolvedrock1.Aftersuccessfulapplicationof proppant fracturing in a chalk formation, it was concluded that proppantfracturingyieldedsustainedproductionrateand became the standard stimulation treatment2.Chalk formations areusuallysoftwithBrinellhardnesslessthan10Kg/mm2 andtherefore creepingispronounced.The effect ofincreased effectivestress,duetoreservoirdepletion,onfractureand matrix permeabilities, was reported3. Proppants importance in sustainingfractureconductivityincarbonateformationwas demonstrated4. Althoughlongercontactperiodsofacidwithformation resultsinmoreetchedsurfaceandthushigherfracture conductivity, it lowers compressive strength of the formation5. Itwasclaimedthatinhighreservoirtemperatures,fastacid reactioninformationscontaininghighconcentrationsof calcite,acidfracturelengthisalotshorterthanpropped fracturelength6.Itwassuggestedthatforreservoirswitha minimumhorizontalstresse(fractureclosurestress)higher than 5,000 psi, proppant fracturing is the optimum stimulation methodbecauseetchingcausedbyfractureacidizingcannot supportsuchhighstress7.Inchalkformations,itwasshown thatproppantfracturingyieldedbetterresultsthanacid fracturing8.Fracturelengthinacidfracturingandproppant fracturingwillbedifferentduetothedissimilarfracture mechanicsinvolvedinthesetechniques.Inproppant fracturing, the fracturing gel is not reactive with the formation, andthereforecanpenetratedeeperascomparedtoacid fracturingforagivenfracturing-fluidvolumeespeciallyat high reservoir temperature. Therefore it is anticipated to create longerfracturesinproppantfracturingascomparedtoacid fractures.Thispaperpresentsarockmechanicsviewof SPE 102590 Acid Fracturing or Proppant Fracturing in Carbonate Formation?A Rock Mechanics View H.H. Abass, A.A. Al-Mulhem, M.S. Alqam, and K.R. Mirajuddin, Saudi Aramco 2SPE 102590 fractureclosureofproppedandacidetchedfractureas follows: 1)Inacidfracturing,fractureclosureisduetoasperities embedment,asperitiescrushing,andviscousflow (creeping).2)Inproppantfracturing,fractureclosureisdueto proppantembedment,proppantcrushing,andproppant flowback. Fracture Closure in Acid Fracturing Productionincreaseduetoanacidfracturingtreatmentis generatedfromtwofactors;fracturelengthandfracture conductivity.Fracturelengthiscontrolledbyacidconvection (injection rate), acid-reaction rate, and acid-loss rate. Fracture widthisaresultofthedifferentialetchingoccurringasthe acidreactswiththewallsofthecreatedfracture.Thiswill createanunevenfracturesurfacethatwilldeterminethe fracturewidthuponfractureclosure.Therefore,fracture conductivityisdeterminedbytheamountofrockdissolved, fracture-surface roughness, closure stress, and the stress-strain characteristicsofrockformation.Ifreservoirtemperatureis toohigh,optimizationoftheinjectionratebecomesvery critical to create a long conductive fracture. If the reaction rate is low, uniform etching may be resulted leading to insufficient fractureconductivity.Theindustryhasfocusedonreducing fluidlossandacidreactionbyincreasingacidviscositysuch as using emulsified and gelled acid systems.Uponcompletionofanacidfracturingtreatment,three factors will contribute to a reduction in fracture conductivity: 1)Elastic response. 2)Compressive failure of contact points (asperities). 3)Creeping effect. Theelasticclosureresponseoccurswhenthenet effectiveminimumhorizontalstressincreasesasaresultof reservoirdepletion.Theelasticresponsetoclosethefracture followsHookeslawofelasticityanditiscontrolledby Youngsmodulusoftheformation.Theelasticresponsewill decreasetheapertureofthefracturewhichreducesfracture conductivity.Ifweassume50ftoftherockperpendicularto thefracturewillcontributetofractureclosure,thenfora Youngs modulus of 3 x 106 psi, the decrease in fracture width correspondingtoadecreaseinreservoirpressurefrom7,000 to 4,000 psi will be 0.05 inches. The fracture will not close by 0.05inchesratherthecontactpointswillcarrytheapplied stresstopreventfractureclosureiftheyarestrongenoughto withstand the stress. The compressive strength of the asperities willdeterminetheseverityoftheirfailureonfracture permeability.Thereductioninconductivityisduetoa combined effect of elastic response and compressive failure of theasperities.Compressivefailurealsogeneratesrock particlesandfinesthatwillfurtherreducefracture conductivity. Thecreeping(viscous)effectisaslowtimedependant displacement.Thetotaldisplacementobtainedfromapplying a constant stress is the sum of two components: ( ) te t + = ......... (1) Thecreepfunction,( ) t characterizestherheological propertiesoftherockformation.Thisfunctionisbest describedexperimentallyforagivenrangeofstress, temperature,andlithology.Creepingmodelsincludethe elasticresponsedescribedbyHookeslawforHookean substances(springmodel)andtheviscousresponsefor Newtoniansubstancesasdescribedbyadashpotmodel.The elasticstrain,e ,isdefinedintermsofYoungsmodulus,E, and effective stressfollowing Hookes law: Ee = . (2) While the viscous effect as represented by a dashpot model is described by: =.. (3) Bothoftheseeffectsareactingwhenthereservoir pressuredecreases;theelasticdisplacement(springeffect)in response to the increase in effective closure stress, and a time dependantdisplacementfunction(dashpoteffect).All viscoelastic models include both effects to simulate a creeping phenomenon.Figure1showsatypicalcreepingbehaviorof certainrocksthatflowsundergivenconditionsofstressand temperature9. Figure1:Acreepingbehaviorofarockthatflowsundergiven conditions of stress and temperature (J eager & Cook, 1979). Three regions are usually observed in some rocks that creep to failure: 1)Primarycreep:Itisalsocalledatransientcreep,which representsanincreaseintheobservedstrainbutina decreasing rate with time (slope is decreasing).2)Secondary creep: It is also called a steady-state creep that exhibits constant strain rate (constant slope). 3)Tertiarycreep:Itisalsocalledacceleratingcreepin which some rocks and under certain combination of stress and temperature,exhibitacceleratedstrainrate(increasingslope) approachingviscousdeformation,leadingtostructural collapse of the rock frame. Experimental Design and Simulation An experimental procedure was designed to simulate acid and proppantfracturingprocessesusingtherockmechanics loadingframe.Twotypesofgeometrieswereusedthat Time Transient Steady Tertiary StrainSPE 1025903 simulateradialandlinearflowregimes. Wholecoresamples wereselectedwithdimensionsof4-in.diameterand approximately4-in.long.A-in.diameterholeisdrilledin thecentertoallowforaradialflowthatcanbeestablished through the rock matrix or through an induced fracture (Figure 2). Figure 2: Radial flow experimental model. Agivensamplewasthencuthorizontallyintotwopiecesto simulateafracture.Thesurfacessimulatingafracturewere surfacegroundedandexposedstaticallyto15percentacid from both sides either by dipping the sample in acid or placing acidonthesurfaceuntilnomorechemicalreactionis observed.Thefracturesurfacebeforeandafteracidizingis shown in Figure 3.Specialcarewasexercisedaroundthecenterholeandthe sampleexternalboundarytopreventlosingsamplecontact due to excessive etching at these boundaries. The sample was boundedtogetheragainwiththesamealignmentbefore acidizing by matching two marked lines drawn on the sample before cutting. Figure3:Fracturesurfacebeforeacidizing(above)andafter acidizing(below)showingtheetchingeffectandgenerated asperities. Ascreenwithtwoscrewclampswasmountedonthesample toallowtheconfiningfluidtoflowradiallythroughthe simulatedetchedfracturetothewellbore.Thefinalgeometry ofthesimulatedexperimentwasaverticalwellborewitha horizontalfracture.Incaseofaproppedfracture,thesame design wasapplied;however,aproppantlayerwas placedon onesurface.Thesamplewasthenpositionedinsidetherock mechanicsloadingframethatprovidesthefollowing measurements: 1)Verticalstressappliedperpendiculartothefracturethat simulated the minimum horizontal stress. 2)VerticalstrainwasdeterminedfromtwoLVDTsthat measure the axial strain vs. time for a given stress. 3)Externalpressureusingtheconfiningfluidwhich simulatedthereservoirpressurebecausethesamplewas not jacketed.4)Wellborepressurewhichwasbasicallywhenitwasput on production. 5)Reservoirtemperaturewhichwassetwithintheloading frame. 6)Productionratewasmeasuredbytimingaproduced volume of oil. 4SPE 102590 Theothersamplegeometrywascreatedbyapplyinga Brazilian tensile failure to split a 4 whole core in two halves. Thentheconfiningpressure wasappliedaroundthefractured sample and a linear flow was established to determine fracture conductivity of an etched or propped fracture (Figure 4). Figure 4: Linear flow experimental model. Acreepingtestwasdesignedbyapplyingin-situ conditionsoftemperatureandstressforagivensample. Progressiveloadssimulatingastresspathexposedona fractureduringproduction,wasappliedandmaintained constantastheresultingdeformationwasmeasured.Fracture conductivitywascalculatedfortheappliedprogressive stressestodetermineitsvariationduetotheelastic,plastic, and viscous effects. Creeping Test of Acid-Fracture SamplesA creeping test was designed to study rock deformation under constant stress as a function of time. This test simulates the in-situreservoirconditionswhereafractureisexposedtothe effectiveminimumhorizontalstress.Atypicaltestinvolves loading the sample to three progressive stresses; 4,000, 6,000, and8,000psi.Ateachstresslevel,theelasticandviscous displacementsaremeasureduntilanexplanatorytrendis obtained.Ifweconsidertheresultsofatypicalsampleas shown in Figure 5, the elastic strain for loading the sample to 4,000psi,thestrainwas0.00064in/in.Thestresswasthen maintainedconstantat4,000psifor71hourstoobtainthe creeping characteristics for the sample at this stress level. The creeping profile suggests that the sample exhibits the primary and secondary creeping phases but has not shown any sign of tertiarycreeping.ThisisexpectedforsuchahighYoungs modulus sample. The elastic, primary, and secondary creeping responsescanbeobservedinFigure6inastress-time representation.Theaccumulatedstrainattheendof71hours was 0.00082278in/in.Thestresswasthenincreasedto 6,000 psi; where the cumulative strain increased to 0.00100201 in/in. This means that the elastic strain generated from the additional 2,000 psi is 0.0001793 in/in. The stress was then kept constant at6,000psiforabout41hoursto have atotaltestingtimeof 112hours.Thetotalstrainatthistimewas0.00108228in/in. Theprimaryandsecondarycreepingyieldedastrainof 0.00008027in/induringthisperiod.Thestresswasthen increasedto8,000psi;wherethecumulativestrainincreased to0.00126152in/in.Thismeantthattheelasticstrain generatedfromtheadditional2,000 psi was0.00017924.The stress was then kept constant at 8,000 psi for about 118 hours to have a total testing time of 230 hours. The total strain at this time is 0.00138577 in/in. The primary and secondary creeping yielded a strain of 0.00012425 in/in.

00.00020.00040.00060.00080.0010.00120.00140 50 100 150 200 250Time, HoursStrain, in/in 4,000 PSI 6,000 PSI8,000 PSIFigure5:Strainbehaviorforthreecyclesofloadingshowing elastic behavior and time dependent creeping. Linear fitting of the secondary creeping portion for three stress levels is described by the following equations: 4 710 871 . 7 10 78 . 4 + = x t x at 4,000 psi 4 710 99 . 9 10 41 . 7 + = x t x at 6,000 psi 3 710 306 . 1 10 5338 . 3 + = x t x at 8,000 psi To predict the creeping strain as a function of time, a plot wasmadetodeterminethefunctionbetweenthecreeping strainandlog(time).Astraightlinewasresultedand interpolationbecamepossibletodeterminecumulativestrain at any time (Figure 6).Creeping Prediction at 4000 psi and 228 F0.000760.000810.000860.000910.000961 10 100 1000Time, hrsStrain, in/in Figure 6: Creeping extrapolation for any given time. SPE 1025905 Creep ModelingTomodelthecompletecreepingresponse(primaryand secondary),Burgersmodelwasusedtodescribetheaxial strain as a function of time for a sample subjected to constant axial stress10: t eG G G Ktt G2) / (1 1 23 3 3 3 92) (1 1 + + + =.. (4) Thismodelincludestheinstantaneousstrain,transient creep, and steady state creep. The experimental creep data for 4,000 psi axial stress was matched by Burgers model using the following parameters. = 4000 psi K= 3.75 x 106 psi G1 = 16 x 106 psi G2= 2.9 x 106 psi 2= 2.2 x 109 psi.hr 1= 40 x 106 psi.hr t= time, hrs Figure 7 shows the experimental and model prediction for the 4,000psicreepingtest.The modelclearlyillustratesthenon-linear time-dependant behavior. Themodel parameters reflect intrinsic properties of a given rock formation. 4000 psiAxi alStress0.00070.000740.000780.000820 10 20 30 40 50 60Ti me, hrsAxial strainExperimental dataBurgers modelY=5 E-7 X +0.0008 Figure 7: Modeling experimental creeping data using Burgers model. Fracture WidthFracture width varies significantly between acid fracturing and proppantfracturing.Fracturewidthinacidfracturingis createdfromtheetchingmechanismanduponclosing; channels will be left open because of the non-smooth surfaces of the created fracture. In proppant fracturing, a fracture closes onaproppantbedleavingacontinuoushighlypermeable fracture (not channel) connecting the reservoir to a wellbore.The average width of an induced fracture subjected to an applied net pressure, P, is given by10: ( )EA P mWav21 = (5) WhereAistheareaforafractureandmisanumerical geometryfactorrangingfrom0.71to0.95dependingona givenfracturelength.Ifthisequationiscomparedtosimple plainstrainequation,theeffectofapressurizedfracturein developingfracturewidththroughrockdisplacementis determined by the factor A . If we assume a square fracture, thedistancethroughtherockmassperpendiculartothe fracture that contributes to fracture displacement is equivalent tofractureheightorlength.Thissuggeststhattheapplied stressistransferredintotheformationanditsdepthis proportionaltotheloadedarea.Thismeansthatthereisa criticaldistanceperpendiculartothefracturewithinwhicha rockmassisdeformingandbeyondthisregiontherock formationdoesnotexperiencetheappliedstress.Therefore, theformationbeyondthecriticaldistancedoesnotexertany elasticrebounddeformationtowardthefractureupon removing the pressure in the fracture. Thisverydistancecanbeassumedtocontributetothe fractureclosureuponremovaloftheappliedfracturing pressure.ThestrainfunctionpresentedinFig.7determines the strain at a given time which is basically defined by: LwStrain= . (6) Wherewistherockdisplacementthatcausesfracture-widthdevelopmentorclosure,andListhecriticaldistance thatwillbecontributingtofractureclosure.Thecritical distancecanalsobetherockmasscontributingtothetime-dependantclosureincludingtheprimaryandsecondary creeping phases. Thedisplacementduetocreepingcomparedtoelastic responsebecomessignificantwithtime.Thisdisplacement willnotclosethefracturedirectly,butitismanifestedinto stressappliedonthecontactpoints(asperities)inacid fracturingorontheproppantgrainsoftheproppantpackin proppantfracturing.Theconductivityofaproppedfracture will be compared to the acid fracture as follows: 1)Theelasticandcreepingforceswillbeappliedonthe proppant grains or asperities. In general a single grain of a proppedfracturewillexperiencelessstressthanan asperity of an acid fracture. 2)Thestrengthofasphericalproppantgrainshouldbe compared to an irregular asperity in an acid fracture. On the other hand direct displacement thatcloses the fracture duetocreepinghappenswithinarockspacebetweentwo consecutive contact points. 6SPE 102590 Conductivity of acid & proppant fracture (Rock A) To evaluate the effect of elastic and creeping displacements on fracture conductivity, flow testing was conducted using a type of mineral oil. The production rate decreased from 180 cc/min toabout20cc/min.Thestresswasthenmaintainedat4,000 psi to evaluate the creeping effect as depicted in Figure 8. The productionratedeclinedfrom20cc/mintoabout5cc/min after 100 hours. Figure8.Timedependentcreepingeffectonproductionrate at 4000 psi. The effect of creeping can be less dramatic if the fracture can transferthecreepingforcethroughcontactpointswithout failure.Asclosurestressincreasessomecontactpointsfail andacontinuousproductionratedeclineisanticipated.Ina proppedfracture,thecreepingforceistransferredifthe proppantgrainstrengthissufficient,otherwiseproppant crushing occurs.Theeffect ofconstantclosure stress of5000 psi left for 67 hours, on flow rate through a propped fracture is showninFigure9.Theeffectofcreepingonproppant-fractureconductivitywasnotsignificant.Thistestshouldbe performed for the given formation and the proppant to be used under in-situ conditions. Time Dependant Effect at 5000 psi and 228 0FProppant Concentration 0.365 Ib/ft2 0204060801001201401601802000 10 20 30 40 50 60 70 80Time, hrsRate cc/min Figure9.Theeffectofcreepingonproductionrateofapropped fracture at 5,000 psi axial stress. Anormalizedfractureconductivityrepresentedasa percentage of the initial conductivity is presented in Figure 10 for the acid and proppant fractures. The figure indicate that for thisformationaproppantfracture,ifcanbeperformed,will sustainwellproductivitywhileacidfracturingwillsuffer production decline with time. 0.00.20.40.60.81.00 2,000 4,000 6,000 8,000Closur e St r ess, psiNormalized Frac Conductivityacidprop Figure 10. Comparison of acid and proppant (1-layer 12/20 ISP, 0.365 Ib.ft2). Conductivity of acid & proppant fracture (Rock B) In this case the sample is from a different formation and only flow testing at room temperature was performed. The effect of stressonpermeabilitywasevaluatedforthematrix,tensile fracture, 100 mesh layer of 0.12, one layer of 30 mesh RCP, and acid fracture (Figure 11). Creeping effect on rate, 4000 psi at Room Temp.05101520250 20 40 60 80 100 120Time, hrsRate, cc/min040801201602000 1000 2000 3000 4000 5000Stress, psiProduction Rate, cc/minReservoir d l tiCreepinSPE 1025907 0.000010.0010.11010000 1000 2000 3000 4000 5000 6000Effective Confining Pressure, psiTotal Permeability, mdMatrix mesh100 (0.12)Tensile fracture 30mesh RCPacid Figure11.Stressdependantpermeabilityofmatrix,tensile fracture, 100mesh sand (.12), 30 mesh RCP (one layer). Thepermeabilityofaonelayer30meshRCPdecreased drasticallyandalotoffineswasgeneratedataneffective closure of 4,000 psi. This is an important criterion to consider whendecidingonthetypeofproppanttobeusedinthe proppant fracturing treatment. Inthisformation,theacid fractureexhibitedmilddecreasein permeabilityasafunctionofincreasingstressasshownin Figure 12. 0.000.200.400.600.801.000 1000 2000 3000 4000 5000 6000Effective Confining Pressure, psiNormalized permeabilitymesh100 (0.12) 30mesh RCP (1 layer) acid Figure12: Normalized permeability for 100 mesh sand (0.12), 30 mesh RCP (one layer), and acid fracture. Mechanical Strength of Fracture Surface LongAcid-contacttimemaynotbebeneficialtoobtain fracture conductivity as it can weaken the fracture surface and maymakeitmorevulnerableforcreepingandcompressive failureofthecontactpoints.Itwasshownthatthe conductivitycreatedby20minutesofacidcontacttimewas higherthanthatcreatedby40minutesforbothdolomiteand limestonesamples1.Theacidexposureweakenedtherock structurealongthefracturesurfaceresultingingreater sensitivitytoclosurestress.Thefracturesurfacebecomes moreplasticandthecontactpointswillfailathigherclosure stress.Additionallythesecontactpointsneednottobesharp and long as their failure becomes more apparent. This effect is more pronounced near the wellbore as the acid contact time is themaximum.Begetal.11,thereforerecommendedover-displacing the acid into the fracture to prevent high dissolution near the wellbore.

TheBrinellHardnessNumber(BHN)wasusedtodecideon whetheracidorproppantisusedinstimulatingchalk formation; proppant fracturing is considered when BHN is less than 10 Kg/mm2 (Cook, 2004). The acid may reduce the BHN asmuchas50%.Thisdecreaseinhardnessweakensthe contactpointsandcausesthemtofailundertheeffectof closurestress.Additionallyitcreatesamoreductilefracture surface that becomes more vulnerable to creeping effect. Proppant flowback Proppantflowbackisoneofthedrawbacksofproppant fracturing.Averysmallamountofproppantproductioncan cause a major loss of communicating a created fracture to the wellbore.Thereforeseveraltechniqueshavebeenusedinthe industrytopreventproppantflowback.Thesetechniquesare ForcedFractureClosure,CurableResinCoatedProppant (CRCP),on-the-flyRCP,andsupportingmaterials.Among these methods the CRCP and on-the-fly RCP are the optimum choiceiftestedanddesignedforcorrectly.CRCPispartially curedduringmanufacturing,andwheninjectedinthe reservoir,thecuringiscompletedundertheeffectof temperatureandstress.Uponcuringagrain-to-grain cementationiscreatedandthecuredproppantdevelopsa compressivestrengththatpreventsproppantflowbackduring production.A new testing method has been designed in which acontinuousacousticvelocitythroughaCRCPsampleis measuredunderconstantconfiningpressureandtime-dependanttemperaturefunction.Figure13showsthearrival time for both P & S waves as a function of temperature.The p-waveisindicatedbytherightaxiswhiletheS-Wavesare plottedagainsttheleftaxis.TheP-wavetravelsfaster comparedtotheS-waves.Thetraveltimeofbothwaves decreasesastemperatureincreases.Thetraveltimedecreases rapidly, reaching 22.0 -sec and 35.7 -sec for P and S waves respectively,forthetemperatureof250 oFatwhichthis specificCRCPcuredasshowninFigure13.Itiscrucialto determinethetimeittakestoreachthistemperatureinthe field,followingthecoolingeffectinagivenfracturing treatment.8SPE 102590 Strength Development as a Function of Arrival Time353637383940414270 90 110 130 150 170 190 210 230Temperature, 0FS-Waves Travel Time, micro-sec2020.52121.52222.52323.52424.525P-Wave Travel Time, micro-secS1 travel timeS2 travel timeP travel time Figure13:LaboratorydatashowingCRCPstrength development as a function of temperature. Figure14showsfielddataforthecoolingandtemperature recoveryrelatedtoafracturingtreatment.Itisrecommended to establish this temperature behavior in the field. The bottom-hole-temperature profile determined in the field should reflect thewholeinjectionphasessuchthatitprovidesthe temperature basebeforethetreatment,thecoolingeffect, and temperaturerecoveryuntilbottomholetemperaturereturns backtotheoriginalreservoirtemperature.Thisvitaldata shouldbeperformedineachreservoircorrespondingtoa giveninjectionprocess(acidfracturing,proppantfracturing, water injection, etc.) to determine the thermal environment pre andposttreatmentinagivenformation.Inthisspecificfield case,ittookabout21hoursafterthecoolingperiodtoreach the required curing temperature of 250 oF.This is translated to designedclosuretimewhenCRCPisusedforproppant flowbackcontrol.Itisimportanttorealizethatthetesting procedureshouldreflectnotonlytheincreasingtemperature but also the temperature function determined in the field. The test must simulate the exact temperature recovery effect due to injectingatypicalvolumeoffracturingfluidinagiven reservoir. Figure 14: Field data showing the cooling effect of fracturing fluid and temperature recovery following the treatment. Conclusions 1)Creepingtestisintroducedtoprovideadditionalcriterion tomakeadecisiononselectingaproppantoracidfracturing treatment for a given formation and in-situ conditions. Primary and secondary creeping, but no tertiary behavior was observed in the tested carbonate rock at in-situ conditions. 2)Productivitydeclineinanacid-fracturedwellisan integratedresponseoftheelastic,plastic,andcreeping responsestotheappliedstress.About30-40%ofproduction ratedeclineoccursduringashorttimeasaresultofprimary creepingoftheacid-softenedcarbonateformation.Proppant fracturingsustainsproductionratebecausetherearemore supporting points to distribute the increasing closure stress. 3)Theeffectoflumpedfactorsaffectingacidandproppant conductivity,asrelatedtoelastic,plastic,andviscousfailure mechanisms,isbestevaluatedinthelabusingformation samples and same materials planned for a fracturing treatment. Thein-situconditionsofstress,stresspath,temperature,and temperature history should be carefully simulated.4) Atestingmethodologyandcalculatedshut-intimeis presented to achieve maximum compressive strength of RCP. Thegrain-to-graincontactoftheRCPisaverycritical parameter for strength development and the shut-in time must allowforthismechanismtotakeplacewithoutany disturbance.NomenclatureA : Fracture area E: Youngs modulus G1: A rock property that controls the amount of delayed elasticity, psi G2: Elastic shear modulus, psi K: Bulk modulus L: Distance perpendivular to fracture that contribute tofracture closure, ft m : Numerical geometry factor Pe: External pressure, psi Pw: Wellbore pressure, psi P: Pressure drawdown (Pe-Pw), psi re: External radius, inch rw: Wellbore radius, inch t:time, hrWav: Average fracture width w: Fracture displacement during width development : Biots constant :Effectivegrain-to-grainstressactingtoclose fracture. h : Total minimum horizontal stress ( ) t : The creep function t: Total displacement due to creeping e : Displacement due to elastic response : Shear stress : Shear strain rate SPE 1025909 : Dynamic viscosity : Poissons ratio : Axial stress, psi 1: A parameter that determines the rate of delayed elasticity, psi.hr2: The rate of viscous flow, psi.hr References 1.Navarre,R.C.,Miller,M.J.,andGordon,J.E.:Laboratoryand TheoreticalStudiesforAcidFractureStimulation Optimization, SPE 39776 presented at the Permian Basin Oil & Gas Recovery, Midland, TX. 23-26, 1998. 2.Cook,C.C.,andBrekke,K.:ProductivityPreservationvia HydraulicProppedFracturesintheEldfiskNorthSeaChalk Field,SPEpaper73725presentedSPESymposiumon Formation Damage, Lafayette, Louisiana, 20-21 Feb. 2002. 3.Lorenz, J.C.: Stress-Sensitive Reservoirs, JPT, Jan. 1999. 4.Fredd, C.N., McConnell, S.B., Boney, C.L., and England, K.W.: ExperimentalStudyofHydraulicFractureConductivity DemonstratestheBenefitsofUsingProppants,PaperSPE 60326presentedatthe2000SPERockyMountainRegional Conference, Denver, Colorado, 12-15 March 2000.5.Gong,M.,Lacote,S.,andHill,A.D.:ANewModelofAcid FractureConductivityBasedonDeformationofSurface Asperities,paperSPE39431presentedatthe1998SPE International Symposium on Formation Damage Control held in Lafayette, Louisiana, Feb 18-19, 1998. 6.Ben-NaceurK.andEconomidies,M.J.:TheEffectivnessof AcidFracturesandTheirProductionBehavior,paperSPE 18536presentedattheSPEEasternRegionalMeeting, Charleston, WV, 1-4 Nov. 1988. 7.Valko,P.,Norman,L.,andDaneshy,A.:PetroleumWell Construction,Chapter17,WellStimulation,Wiley,1998.p. 506. 8.Cook,C.C.andBrake,K.:ProductivityPreservationthrough HydraulicProppedFracturesintheElfishNorthSeaChalk Field, SPE Reservoir Evaluation & Engineering, April 2004.9.Jeager,J.C.andCook,N.G.W.:FundamentalsofRock Mechanics,ChapmanandHall,London,UK,1979,pp.309-325. 10.Goodman, R.E.: Introduction to Rock Mechanics, Jones Wiley & Sons, New York, 1980, p.199. 11.Beg,M.S.,Kunak,A.O.,Gong,M.,Zhu,D.,Hill,A.D.:A Systematic Experimental Study of Acid Fracture Conductivity, SPE Production & Facilities, Nov. 1998, pp. 267-271. 12.Olson,K.E.,Olsen,E.,Haidar,S.,Boulatsel,A.,andBrekke, K.:ValhallField:HorizontalWellStimulationsAcidvs. PeoppantandBestPracticesforFractureOptimization,SPE 84392presentedattheSPEAnnualTechnicalConferencein Denver, Colorado, 5-8 October, 2003.