Numerical Simulation of Combined Reverse Combustion and Steamflooding for Oil Recovery in a Utah Tar Sand

This paper presents the design of the U.S. DOE Laramie Energy Technology Center's (LETC) Project TS-4, which involves numerical simulation of both in-situ reverse combustion and steamflooding. The simulator showed that the combustion could be limited and contained in a middle 10-ft [3-m] interval with a correlatable High-permeability streak within the 65-ft [20-m] pay zone of the upper Rimrock tar sand formation in Northwest Asphalt Ridge, Uintah County, UT. A high-transmissibility path was necessary to obtain adequate injectivity and sustain a stable reverse combustion. Combustion "echoes" developed and the front changed into a forward mode as the formation pressure increased and at very low air-injection rates. Oil recovery by steam injection was accelerated in a formation preheated by a reverse combustion. Introduction In 1973 LETC began a series of projects aimed at identifying feasible oil recovery techniques for the large deposits of tar sands in the U.S. Two previous combustion experiments have been reported by LETC: Land et al previous combustion experiments have been reported by LETC: Land et al reported the LETC TS-1C, and Johnson et al reported the LETC TS-2C. Both of these were conducted in the Northwest Asphalt Ridge tar sand deposit (T4S-R20E), in Uintah County, in 1975 and 1977, respectively. These were followed by a steamflood experiment, LETC TS-1S, in 1980 in the same area. Analysis of this steamflood experiment indicated that only 18.5% of the original oil in place (OOIP) was mobilized because of poor communication between the injector and the producers. It was clear at this point that the producers had to be stimulated to improve the oil mobility around the wellbores. Steam soaking was considered but discarded because of the lack of adequate reservoir pressure. Since LETC had been successful with its previous use of combustion, the use of reverse combustion to preheat the previous use of combustion, the use of reverse combustion to preheat the producers and possibly the entire sand was considered. A reverse producers and possibly the entire sand was considered. A reverse combustion is preferred to forward combustion because it eliminates the problem of plugging. Project TS-4, therefore, involves a combination of problem of plugging. Project TS-4, therefore, involves a combination of in-situ reverse combustion and steamflooding. The site selected for the test is about 200 ft [61 m] southeast of the location of the LETC TS-1S experiment. The project targets the 65-ft [20-m] pay zone of the upper Rimrock tar sand formation rather than the lower Rimrock targeted in all previous experiments. The sand is well confined and fairly continuous with previous experiments. The sand is well confined and fairly continuous with varying levels of shaliness. The formation bitumen saturation is about 80% compared with 35 to 65% in the lower Rimrock. The permeability of the unextracted core is less than 1 md in some parts and generally one or two orders of magnitude less than that of the lower Rimrock. Preliminary field tests ordinarily showed very poor injectivity without fracturing the formation. The in-situ reverse combustion is intended to preheat the formation rapidly before steamflooding the entire formation. It is confined to a 10ft [3-m] interval that includes a correlatable high-permeability streak to limit the air requirement. It also is expected that good communication can be established between the injector and producers while reducing the oil viscosity and, thus, improving the mobility of the oil. This paper reports a simulation study evaluating the feasibility of this project on a commercial scale and presents a conceptual study of the experiment using a numerical simulator previously described by Coats. Owing to the recent defederalization of LETC, the planned field test for Project TS-4 now has been abandoned. Geology The Northwest Asphalt Ridge is located at T4S-R20E in the Uintah Basin, Uintah County, UT. The geology of this area is described in a greater detail by Campbell and Ritzma. The ridge is separated from the major Asphalt Ridge by a northeast-trending fault. The strata dip southwesterly from about 9 to 350 Average dip angle at the TS-4 location is about 28. The Rimrock sandstone is a member of the Late Cretaceous Mesaverde formation. The other member of the group in this location is the Asphalt Ridge sandstone. Both are of marine origin and oil impregnated. The Rimrock sandstone is unconformably overlain by Tertiary Duchesne River formation of continental origin. It is underlain by the Asphalt Ridge sandstone and separated from it by a thin tongue of Mancos shale. SPEJ p. 227