Numerical simulation of foam-assisted immiscible gas injection for enhanced oil recovery

The efficiency of foam displacement in gas enhanced oil recovery (EOR) in low-permeability (1-10 mD), heterogeneous carbonate formations depends on the stability of the foam in the presence of oil. This study presents a new approach to determine foam model parameters using the Computer Modelling Group (CMG) GEM™ Foam simulator and foam core-flood experiments in the absence and presence of oil. Laboratory data was obtained from coreflood experiments including water flooding and co-injection of alkyl polyglucoside (APG) surfactant solution and hydrocarbon gas at different rates to generate foam in-situ in carbonate rocks. A series of numerical simulations was conducted to construct the relative permeability functions for gas, formation brine, and crude oil in oil-water carbonate cores. These functions were obtained by history matching water flooding data including the measured pressure drop and oil recovery using the CMG CMOST™. Next, these relative permeability functions were used to simulate foam propagation in carbonate cores using an empirical foam model implemented in CMG-GEM. The main focus of this simulation work is to develop a method for determining foam model parameters based on foam core flooding experiments. Then, the obtained endpoint relative permeabilities were used to construct relative permeability models. Second, pressure drop measured during foam flooding in a water-wet core at different surfactant concentrations was used to determine part of the foam model parameters that represent surfactant effect on foam behavior. These parameters were then used to history-match pressure drops at different injection rates and foam qualities from the same core to identify other foam model parameters that represents the effect of foam dry-out, mobility reduction, and capillary pressure. Finally, the foam model parameters representing the effect of oil on foam propagation were determined by history matching pressure drops obtained from foam corefloods in an oil-wet core. An excellent agreement between the foam simulations and experiments was obtained. The study indicated that in order to obtain the foam parameters that can be used for the field development plan, it is beneficial to designated experimental data into stages which can reduce the requirement for extensive simulations to select parameters for foam- assisted enhanced oil recovery. Additionally, rather than concurrent matching, when parameters from previous simulations were kept constant while parameters in subsequent simulations were modified, history-matching quality improved.Petroleum and Geosystems Engineerin