A light oil steamflood pilot project was conducted from 1987 to 1991 in the Shallow Oil Zone of the Elk Hills field. A thermal simulator, employing a distillable component and a nondistillable component, was used to construct a steam-flood model and duplicate over three years of pilot performance. In addition, a conventional black oil model was used to duplicate pre-pilot primary production performance and quantify the distribution of oil, water and gas saturations prior to the start of steam injection. Uneven thermal energy distribution and limited steam zone temperatures minimized the effectiveness of the steam distillation recovery mechanism.
The Shallow Oil Zone (SOZ) in the Elk Hills oil field, Kern County, California, is still largely under primary recovery and is thought to represent a potential 100 million barrel [15.9 million m3] enhanced oil recovery target. A light oil steamflood (LOSF) pilot project1 was judged to have potential to enhance oil recovery from this reservoir based on the technical success of several earlier California light oil steamflood projects.2,3,4 This paper presents the results of black oil and thermal simulation models for a large sector inclusive of the LOSF pilot area. The computer simulation study was performed to assist the U.S. Department of Energy in assessing recovery efficiencies and operational planning.
The principal objective for this simulation project was to history match a defined sector of the LOSF Pilot project for the purpose of developing a useful geological and engineering model. This model would be used to screen other prospective SOZ steamflood areas and to evaluate the economics for expansion of the pilot project. Two prior compositional steamflood models had been developed which simulated the response of symmetrical portions (118, 114, and 112 patterns) of the enclosed LOSF pilot pattern. Both models gave optimistic projections of oil recovery and reservoir response to thermal stimulation.
A large factor relating to the inability of these models to match pilot performance was the significant out-of-pattern reponse. As a result, it was determined that a larger sector model was required to adequately simulate the project, taking into account the large out-of-pattern response. It was also suspected that the presence and influence of faulting, the magnitude of areal and vertical permeability variations, and the distribution of oil and gas saturations at the start of steamflooding were affecting the observed performance of the pilot project.
The principal objective for this simulation project was to history match a defined sector of the LOSF Pilot project for the purpose of developing a useful geological and engineering model. This model would be used to screen other prospective SOZ steamflood areas and to evaluate the economics for expansion of the pilot project. Two prior compositional steamflood models had been developed which simulated the response of symmetrical portions (118, 114, and 112 patterns) of the enclosed LOSF pilot pattern. Both models gave optimistic projections of oil recovery and reservoir response to thermal stimulation.
A large factor relating to the inability of these models to match pilot performance was the significant out-of-pattern reponse. As a result, it was determined that a larger sector model was required to adequately simulate the project, taking into account the large out-of-pattern response. It was also suspected that the presence and influence of faulting, the magnitude of areal and vertical permeability variations, and the distribution of oil and gas saturations at the start of steamflooding were affecting the observed performance of the pilot project.
