压裂液滤失（第二部分）.pdf

A SPE 36493Fracturing Fluid Leakoff Under Dynamic Conditions Part 2: Effect of Shear Rate, Permeability, and PressureS, Vitthal, SPE, and J.M. McGowen,SPE, HalliburtonEnergy ServicesCopyright1996, Societyof PetroleumEngmeecsInc.Th!s paperwaspreparedfor presenlal!onat the1996 AnnualTechnncalConferenceand Exhhbmon held n Danver,Colorado,U.S.A., 6-9October1996Thm paper was selectadfor presantatnonby an SPE ProgramCommmaefollowlngrev,ew o{ mformatloncon!amedm an abstractsubmittedby the author(s)ContentsO( the paper,as presanted,have no! baen revmwedby the Soctery of PetroleumEngineersand are Subpcl!0correctionby the author(s)Thematerial,as presentad,doesnot nacassar[lyreflectany position of me SocIefy of PetroleumEngmaers,Its offlcera, or membersPaperspresentedat SPE meetmgsara subjactto pubhcalmnrev$aw by Ed!tortalComm!tleesof the Societyot PetroleumEngme8rs.Permmsmnto copy IS restrictedto an abstractof not morelhan 300 words.lllus!raoonsmay nol be Cc!pled. Tha ab$lraclShou!d conlainconspicuousacknowl. adgmentof whereand by whomthe paperIS presentedWriteLlbranan,SPE,P OBox 833636,Richardson.TX 75083-3S36,U S A, lax 01-214.952.9435Abstract This paper discussesthe resultsof an industryconsortium establishedto understandand allow modelingof the fluid leakoff process in hydraulic fracturing applications.Over 1,000 laboratory experimentsstructured to study the effects of shear rate, permeability,differentialpressure, temperature,gel con- centration,fluid-lossadditives,and fluid type have been con- ducted.Fluid-1oss data was measuredwith a state-of-the-art experimentalsetup that includedfluid preconditioningloops for high shear and low shear regimes and a novel cell design that minimizedflow irregularities.Part 2 presentsthe resultsof experimentsinvolvinglinear hydroxypropylguar (HPG) gel, HPG/boratesystems, and HPG/titanatesystems at shear rates between Oand 200 see”’, pressures between 500 to 10,000 psi, a temperature of It30°F,and core permeabilitiesbetween 0.1 md and 1000 md. This study found that fluid loss under dynamic conditions can be significantlyhigher than under static conditions.Sig- nificant differencesin fluid-lossbehaviorwere observedbe- tween linear gels, transition-metalcrosslinked gels, and borate- crosslinkedsystems. The behavior of linear gels was sensitive to permeabilityand pressure, but insensitive to shear rate. The behavior of crosslinkedgels was more sensitive to shear rate, but less sensitive to permeabilityand pressure. The fluid loss of all fluids tested could be modeled though the mechanismsof non-Newtonianviscous invasion, classicalfilter-cakedeposi- tion, and filter-cake resuspension.’~This paper presen[s guide- lines for fluid-loss prediction for these systems with appropriate fluid-lossmechanisms.ISociatvnf Potrohmm EnoineersThefindingspresentedin thispaperare vitalfor the predictionof fracture fluid leakoff within fracture simulators. Understandingwhich fluid loss mechanismis predominantis essential in analyzing pressure decline data from minifracturing treatmentsand in using this informationin predictingtreat- ment placement.Introduction Traditionally,hydraulicfracturinghas been limitedto rela- tively low-permeability(vf..,(2b)A plot of the leakoff velocity on a log-log plot vs. leakoff timewill have a slope of -0.5 until VC is reached.When equilibriumis observed,the leakoff velocityshould become constant and equal to v,. Filter-CakeCompressibility.Asidefrom the fluid-1oss modelingfrom Eqs. 1 and 2, one additionalparameterwill be discussed in detail: the filter cake compressibility.For simplic- ity, the filter-cake compressibility(a),is defined as the pressuredependencyof the fluid-loss coefficients as expressed in Eqs. 3 and 4.C,,, K @~)a. .. ........................................ (3)Vn m (flp)~a.................................. .......... . (4)The value of the exponentawill vary from 0,0 for a completely compressiblecake to a value of 0.5 for a completely incompressiblecake. In broaderterms, the filter-cakecom- pressibilityindicates the variance of cake permeabilitywithin the cake thickness.If the cake permeabilityis a constant,as occurs with hard, solid particles, the filter cake will be incom- pressible (i.e., independentof pressure). Since most fracturing fluids generate a filter cake of flexible polymer strands,it is extremelyunlikely for them to behave other than as compress- ible materials. PreviousInvestigations.The effect of the parametersthat influencefluid leakoffhave been investigatedby many au- thors.5R-~~ However,it is clear that fluid leakoff is a complex phenomenon(hat is not fully understood and poorly modeled. Penny and Conwayn and McGowen and Vittha17provide com- prehensivereviews of fracturing-fluidloss, thus only literature relevant to discussion of the effects of shear rate, permeability, and pressure will be reviewed here. Effect of Shear Rate. The effect of crossflow velocity on fluid loss (dynamicfluid loss) has been commonlymodeled usingthe conceptof。