Thisisa preprint of a paper intended for publication in a journal or proceedings. Since changes may be made before publication, this preprint is made available with the understanding that it will not be cited or reproduced without the permission of the author. DISCLAIMER This document _ _ u u aoco_t _ work Sl_m_! by *-_ of the United S_ _ Neither the Umlt_llS_ GovermmmdMr the Umlveml_ ' of Califorubt nor any d their employees, make8 uy warranty, expcum w Implied, o¢ assumesuy lepJ UsbUityer respoesiWUty f_ the sa:umcy,completme_ w useful-nes8 d any Isfommtioe, sppetstus, p¢odect, w l_eesu dJsd_e_ w rqwes_ts that its use would mot Infrinle Iwi_tely ewned rights. Rderesce berelu to amy specific eommercielp_oducts, process, w serviceby trade ssme, trademark, sumufscturtr, or otherwise,doesnot secemmrily coustitutew imply its eudorsemeat, recommeadatiou, or tsvoflng by the Uuited S_ttes Govetumeater the Uulversit7 of Cslifontht. The viewsand opiniom _ sutborsexpressedherein do uo(stecessarily state er reflect thoseof the United States Govemmeutor the University of Califoruis, and shall am(be used for advertising or product endorsement purposes.

No abstract

Summary The effectiveness of foam as a mobility-control agent has been tested by an immiscible CO2 coreflood displacement of a California heavy oil (14 degrees API [0.97 g/CM]). Foam was produced by simultaneous injection of 0.5 wt% surfactant solution and CO2. Foam injection recovered an incremental 33.6% of original oil in place (OOIP) by reducing CO2, mobility and diverting CO2. Introduction Many CO2 field projects have demonstrated a need for mobility control of the injected gas. Laboratory tests have indicated the possibility of improvements for CO2 mobility control and diversion with the use of foams, and some field tests have been carried out. Wellington and Vinegar studied CO2-foam mobility control by use of computerized tomography in corefloods. They used 0.05 wt% Neodol 25–12 ethoxy glycerol sulfonate in brine to prevent gravity and viscous instabilities. The basis of their coreflood was to saturate the core with surfactant solution and then inject CO2 continuously following waterflood. Casteel and Djabbarah studied the effectiveness of foam in improving sweep efficiency in a parallel-core system with a permeability ratio of 6.4. They initially waterflooded each core separately. This basis might not be representative of a typical reservoir process because lower-permeability strata may not be completely water-flooded. Casteel and Djabbarah showed that the highest total recovery is achieved by first injecting CO2 and then following with surfactant solution. Less recovery is achieved when the surfactant slug is injected after the waterflood and is then followed by CO2. Duerksen showed that a continuous foam generation is required to maintain foam resistance to steam flow effectively in a porous medium. Raza also showed that small slugs of surfactant solutions followed by gas are more effective in reducing gas mobility than a large slug of surfactant solution. The smallest slug size injection is obtained by simultaneous injection of gas and surfactant solution. In the study in this paper, CO2 and surfactant solution were injected simultaneously to generate and maintain foam to reduce CO2 mobility. The effectiveness of foam as a mobility control agent was tested by an immiscible CO2 coreflood displacement of a California heavy oil (14 degrees API [0.97 g/CM]). The study was conducted with a parallel-coreflood system consisting of two segmented unconsolidated reservoir cores with a permeability ratio of 3.7. This represented a physical model of two noncommunicating reservoir sand layers of different permeabilities. The coreflood work was completed in four stages:waterflood followed by CO2 water-alternating-gas (WAG) flood in a single core,waterflood followed by CO, WAG in a parallel-core system,a second waterflood following by CO, WAG in parallel cores to confirm Run 2, anda CO2-foam flood in the parallel cores following Run 3. The two base-case parallel corefloods (Runs 2 and 3) were performed by simultaneously waterflooding both cores, then injecting water and CO, alternately (WAG). Average oil recoveries for the two base cases for low-permeability core (LPC) were 0 and 0.4% OOIP for the waterflood and CO2 WAG, respectively. For high-permeability core (HPC), average oil recoveries were 46.4 and 6.7 % OOIP for waterflood and CO2 WAG, respectively. Foam was produced by adding 0.5 wt% surfactant to the brine solution and then simultaneously injecting the surfactant solution and CO2. The surfactant was Chevron proprietary surfactant CRSO 85/66, selected for the given reservoir fluids and rock, Injection of foam into the parallel-core system produced an incremental recovery of 19% OOIP from the HPC and 52% OOI from the LPC. The total incremental oil recovery over waterflood for CO2 and foam injection was 40% OOIP. Results of this parallel-coreflood study demonstrated three points:the ability of foam to reduce CO, mobility, which resulted in the incremental oil recovery from the HPC;the ability of foam to divert CO2 to the LPC, which showed little recovery from waterflood or CO, WAG; andthe ability of foam to recover 33.6% OOIP more heavy oil than the waterflood and CO2 WAG. Basis of Foam Mobility Control Foam potentially presents a more efficient method for reducing CO2 mobility than water. In a WAG process, injection of water reduces CO2 mobility by lowering its relative permeability, but it also traps oil, increases water flow, and decreases hydrocarbon extraction from the oil by CO2. The inherent advantage of foam over water for mobility improvement is that a foam is 70 to 90% gas. This means that a relatively small amount of water can be used to decrease CO2 mobility. Apparent foam viscosity is greater than the viscosity of its components, and foam viscosity increases with the capillary diameter. When gas contacts the surfactant solution, a foam-like dispersion of gas in the liquid phase occurs. This foam has a large volume of dispersed gas, which is encapsulated in continuous thin liquid films called lamellae. Foam quality is defined as the ratio of gas volume to total volume and normally varies between 70 and 90%. The conditions of the corefloods in this study were scaled to those common in heavy-oil reservoirs in California. The core material was an unconsolidated sandstone under overburden pressure. Pressure and temperature were 1,000 psia and 140 degrees F [6.9 MPa and 60 degrees C], respectively. Flow rates were about 1 to 2 ft/D [0.3 to 0.6 m/d]. At 1,000 psia and 140 degrees F [6.9 MPa and 60 degrees C], CO2 is a dense gas with a density of 0. 15 g/CM and produces a stable foam with a quality of 80 to 85%. The surfactant used in this study was Chevron CRSO 85/66, which was anticipated to have low adsorption on sandstones. The adsorption isotherm is Langmuir type and the adsorption level in a 100% brine-saturated core was measured at 0.5 to 0.6 mg/g rock. The adsorption isotherm for a similar reservoir rock and surfactant solution is given elsewhere. Core Preparation Core segments were cut parallel to the bedding plane from whole core frozen in liquid nitrogen. Wettability measurements were not made specifically for the core material, but repeated dynamic relative permeability measurements, shown in Fig. 1A, indicated that the cores were extremely water-wet throughout all the experiments. The gas/brine relative permeability for a reservoir core sample is given in Fig. 1B. Experimental Program Four corefloods were carried out to investigate the effects of foam on CO2 mobility. Run 1-A single-core waterflood followed by CO, WAG. Run 2-A parallel-core waterflood followed by CO2 WAG. Because less oil than expected was produced from the LPC during Run 2 (in fact, almost none), Run 2 was repeated as Run 3, and an iodide tracer experiment was done to validate the flow system. Run 3-A parallel-core waterflood followed by CO2 WAG. Run 4-Immediately after Run 3, alternating foam and CO, were injected, followed by continuous CO2/foam. Table 1 lists all the floods completed for this study. Table 2 lists the two parallel corefloods completed for this study and their key parameters. SPERE P. 136^

Summary A laboratory study of heavy oil recovery by CO2 injection was undertaken in support of the Wilmington Tar Zone CO2 Injection project operated by Long Beach Oil Development Company. The project operated by Long Beach Oil Development Company. The work included:–Phase behavior of Tar Zone reservoir oil and CO2.–Phase behavior of Tar Zone reservoir oil and the refinerygas (82% CO2 - 18% N2) used for the field project.–Viscosity measurements of oil-gas mixtures.–Reservoir condition displacements of oil by CO2 and byrefinery gas.–Equation of state characterization of phase behavior.–Computer simulation of gas-oil displacements. Saturation pressures and swelling factors were measured for oil-gas mixtures for up to 60 mol % CO2 and for up to 50 mol % refinery gas. These measurements show that N2 is substantially less soluble in oil than CO2. Viscosity measurements show that the viscosity reduction is a function of pressure and the total gas dissolved in the oil. pressure and the total gas dissolved in the oil. Four reservoir condition corefloods were completed:–Refinery gas injection at 0.22:1 WAG ratio, followed by waterflood–Continuous CO2 injection followed by waterflood.–Continuous refinery gas injection followed by waterflood.–Refinery gas injection at 1:1 WAG ratio, followed bywaterflood These floods showed that 1) the recovery efficiency of CO2 is higher than that of the refinery gas for continuous or low WAG injection and 2) the recovery efficiency of the refinery gas at 1:1 WAG is about twice that of continuous injection. The corefloods were modeled with a finite difference compositional simulator. Predictions agree with the experimental results. Introduction The Wilmington Tar Zone CO2 Injection Project has been designed and operated by Long Beach Oil Development (LBOD) Company. The project covers roughly 32,000 ac-ft in Fault Block V. The reservoir project covers roughly 32,000 ac-ft in Fault Block V. The reservoir is 2300 ft deep. The pay zone thickness is 50–100 ft. The porosity is 31%. The permeabilities are in the 100–10,000 MD porosity is 31%. The permeabilities are in the 100–10,000 MD range. The reservoir temperature and pressure are 120F and 1000 psia. The oil has 13–150 API gravity with a GOR of 50 SCF/BBL. psia. The oil has 13–150 API gravity with a GOR of 50 SCF/BBL. The solution gas composition is 96% methane and 4% CO2. The injection gas is refinery by-product gas purchased from Texaco refinery in Wilmington. The composition is 82% CO2 and 18% N2. The operation of the project has been described elsewhere. The Wilmington Tar Zone oil is too heavy to develop miscibility with CO2. However immiscible CO2 injection is effective in heavy oil recovery, because CO2 dissolves in heavy oils, causing swelling and a large reduction in oil viscosity. The reduced oil viscosity improves the fractional flow of the oil for a following water flood. We undertook an experimental and computational program to study the phase and displacement behaviors of this system. The results of the phase behavior measurements were used to tune the Peng-Robinson Equation of State (PREOS) characterization. We Peng-Robinson Equation of State (PREOS) characterization. We performed four reservoir state coreflood displacement using CO2 performed four reservoir state coreflood displacement using CO2 and refinery gas. The PREOS model and core properties were used in a finite difference compositional simulator properties were used in a finite difference compositional simulator to model the displacement experiments. P. 245

No abstract