Recovery Correlations for In-Situ Combustion Field Projects and Application to Combustion Pilots

Abstract

This paper presents two correlations developed to predict the fieldwide oil recovery of dry in-situ combustion processes. The correlations are based on a combination of engineering and statistical approaches. This work also discusses extension of the second correlation to pilot projects and concludes that correlation parameters used for fieldwide projects seem reasonable for use in pilot projects. Introduction During the last three decades, the injection of heat-bearing fluid into a reservoir has gained considerable usage as an enhanced oil recovery mechanism. Crude oils with low API gravity and high viscosity do not respond satisfactorily to conventional recovery processes. Since oil viscosity is highly temperature dependent, the performance of these reservoirs can be improved significantly by raising the reservoir temperature through the addition of heat. In-situ combustion is one of the processes that can increase oil recovery through heat addition. Oil recovery predictions for in-situ combustion projects would provide a basis for calculating air-injected/oil-produced ratios (AOR's), oil rates, oil recovery, and economic limits. There is a possibility that such predictions could be achieved by describing the movement of heat and fluids analytically, and such approaches have been tried; however, no analytic model has been developed that completely describes the movement of heat and fluids.1 Numerical modeling has been an alternative way of describing in-situ combustion; however, due to the complexities of the process, the multidimensional geometry, and the cost of running the more complex models, no approach to recovery prediction has been entirely satisfactory. It appeared that the solution to this dilemma was to attempt to correlate field results. Our approach was to base our correlating parameters primarily on those terms that occur in heat and material balances of combustion and only then to add additional variables on a purely statistical correlation basis. Although some screening guides2–6 in the literature attempt to define which projects may be successes or failures, these guides make only a yes/no evaluation or apply only to certain fields. Therefore, there is a clear need to develop a simple model to predict the oil recovery to be expected from in-situ combustion.