Heavy Oil From Fractured Carbonate Reservoirs

Abstract

Summary This paper investigates the recovery mechanisms for steam injection intonaturally fractured, carbonate, heavy-oil reservoirs. Interim ideas and resultsof laboratory and simulation studies are presented and topics for furtherinvestigation are suggested. Results presented and topics for furtherinvestigation are suggested. Results so far indicate that both imbibition andinternal gasdrive are effective at driving oil into the fracture network. Introduction Heavy oil contained in carbonate reservoirs worldwide is estimated to be 1.6x 10(12) bbl in place. So far, this major resource has attracted littleattention from the petroleum industry and little has been produced. Thisprobably stems from the conceptually difficult task of recovering viscous oilfrom naturally fractured carbonate formations. Clearly, a means of reducing oilviscosity, such as a thermal recovery technique, is needed to allow flow toproducing wells. Steam injection is currently the most successful thermalrecovery technique available. However, steam might be expected to travelpreferentially through the fracture system and to recover little oilpreferentially through the fracture system and to recover little oil from thematrix; The chemical reactivity of the formation to the steam injectant alsomight be expected to cause problems in the forms of formation damage and scaleproduction. This paper describes recent research in identifyingpressure-cycling steam recovery strategies that could unlock this largeworld-wide resource. Mechanisms thought to offer potential include imbibition, viscosity reduction (from increased temperature and CO2 dissolution), oilswelling (from increased temperature and CO2 dissolution), gravity drainage, and internal depletion gasdrive (from flashing of solution gas, steamcondensate, and dissolved CO2). Initial experimental research studies have beenboth mechanistic and fundamental in nature. High-temperature/high-pressure(HTHP) mechanistic studies are identifying the contribution to recovery fromthe individual mechanisms listed above. The quantitative parameters needed tomodel the most significant processes mathematically are being determined atappropriate temperature and pressure conditions. Fundamental studies aredirected toward identifying scaling rules for modeling imbibition processes. The effects of viscosity ratio, temperature, and matrix/fracture geometry arebeing studied. To advance experimental research, special HTHP facilities havebeen constructed. These have been designed to match reservoir conditions duringcyclic steam processes. Further experiments are planned to study thetemperature dependence of these processes as the matrix heats and wettabilityand rock/fluid interaction contours traverse the rock matrix. An apparatus wasdeveloped to determine the modeling parameters, such as relative permeabilityand capillary pressure, under representative conditions. Effects of increasingtemperature on PV, PV compressibility, and matrix permeability are also beingstudied. PV compressibility, and matrix permeability are also being studied. Anapparatus also was developed to investigate displacement and depletionprocesses by use of X-ray computerized tomography (CT). Novel low-density coreholders with optical fiber temperature sensors are used. Concurrently, mathematical models are being developed to capture the essentialcharacteristics of the recovery processes for sensitivity studies, historymatching, multicycle prediction, and economic optimization. The simulationapproach to the field process that has been developed allows the main recoverymechanisms process that has been developed allows the main recovery mechanismsto be interlinked. At this point, we have used the simulator to investigaterecovery strategies for heavy oil recovery from fractured carbonatereservoirs. Results From Laboratory Program HTHP Mechanistic Studies. Initial HTHP mechanistic studies ere performedwith plugs cut from an outcrop block of Permian performed with plugs cut froman outcrop block of Permian magnesian limestone (dolomite). Table 1 gives themineralogical composition of the sample obtained from X-ray diffraction (xRD), and Table 2 gives the physical characteristics. Table 3 presents the propertiesof the live crude oil used for the studies. The experimental properties of thelive crude oil used for the studies. The experimental sequence was chosen torepresent a cyclic steam process. The clean, dry sample was saturated with10,000 ppm NaCl after measurement of the PV, porosity, and gas permeability. The sample was then confined in a core holder at a 1,500-psi overburdenpressure. Swi was achieved by flooding with the live crude oil with pressure. Swi was achieved by flooding with the live crude oil with a 200-psidifferential pressure across the sample and a 750-psi back-pressure. The samplewas then aged for 24 hours to allow equilibration of the fluids. The systemtemperature then was raised to 302 degrees F and any production from oil orrock thermal expansion was identified by allowing a period of 4 hours to elapsebefore the imbibition phase began. Spontaneous imbibition counterflowproduction was monitored by flowing brine across the top face of the rocksample and collecting effluent in a separator (Fig. 1). Initial oil production(over the first 10 minutes of brine flow) was associated with thermalexpansion. The backpressure then was reduced to ambient pressure and thedepletion production caused by internal gasdrive and/or steam flashing wasmonitored. The blind end of the core holder then was opened and the sample wassubjected to consecutive hot-water flood and steamflood. This sequence then wasrepeated on a fresh sample at 482 degrees F . Table 4 presents results of thesetests. The tests at 302 degrees F indicate that thermal expansion results inminimal recovery, imbibition plays a major role in producing oil from thematrix into the fracture network, depletion results in further appreciableproduction, and forced displacement by hot water or steam appears notproduction, and forced displacement by hot water or steam appears not to resultin further recovery. The tests at 482 degrees F indicate that the initial highproduction occurs as a result of gas evolution as the bubblepoint is breachedat the high temperature, further production occurs because of imbibition, anddepletion does not result in high production because the gas-drive mechanismhas already been exploited at the higher temperature. (Note that higherproduction may have been masked by distillation losses from the separator.)Note that firm conclusions cannot be drawn from the results of only two tests. Several further data points are required.