Transient Foam Displacement in the Presence of Residual Oil: Experiment and Simulation Using a Population-Balance Model
The population balance model of Kovscek, Patzek, and Radke is extended to model the flow of foam in porous media containing residual oil. A mechanistic rate of coalescence due to oil, based on a recently proposed pinch-off mechanism, is included in the model. New foam-flow experiments in the absence and presence of residual oil are conducted in 60-cm long one-dimensional Berea sandstone cores of 0.23 and 1.2 mm2 permeability. Results are obtained for total superficial velocities near 0.3 m/day and 97 % quality foam with a 0.5 w/o sodium dodedcyl sulfate, 0.5 w/o NaCl surfactant solution and a decane oil phase. Water saturation profiles are measured by microwave attenuation and pressure profiles are measured with pressure taps distributed along the length of the cores.
The displacement experiments show an entrance region as long as 15 cm where the gas phase evolves from a high mobility continuous gas phase to low mobility foam.
When foam is destabilized by oil, a capillary end effect is observed that causes development of a large pressure gradient that propagates from the exit of the core to the entrance of the core over tens of injected pore volumes. These entrance and end effects are believed to dominate overall pressure drops measured in previous foam-flow experiments in the presence of oil, especially those in cores less than about 20-cm long.
During the transient displacement mode, excellent agreement is seen between
experiment and theoretical predictions of the population-balance model over varying gas and surfactant-solution injection-velocities and at differing residual oil saturations. Also in agreement with experiment, the proposed population-balance model predicts lower foam mobility in the 1.2 mm2 core than in the 0.23 mm2 core both in the presence and in the absence of oil. This latter result is confirmed by new foam displacement experiments conducted in a heterogeneous Berea sandstone core
Thus, our findings suggest that in an oil reservoir foam will flow preferentially through low permeability oil-containing regions of the reservoir and partially block high permeability oil-depleted regions as compared to a Newtonian drive fluid. The permeability contrast leveling by foam is important because it confirms that foam can be used to block previously swept high-permeability regions of an oil reservoir and preferentially direct flow to previously bypassed low-permeability regions that still contain oil. Our extended population-balance model predicts that the effect is augmented when foam-destabilizing oil is present at higher saturation in the low permeability region than in the high permeability region.
This work significantly extends our earlier population-balance foam simulator and presents a rather complete mechanistic foam displacement model including the effects gas and liquid flow, absolute permeability, surfactant concentration, foam sensitivity to oil, and oil saturation.