Foam is a promising solution for improving the poor sweep efficiency of gas injection for enhanced oil recovery (EOR), in that foam lowers gas mobility significantly by trapping gas bubbles. Most of oils destabilize foam, the effect of which dominates foam strength when foam is in contact with oil and is key to success of foam EOR, but not yet fully understood. This study focuses on understanding oil displacement by foam in lab scale and providing insights for field-scale applications. We carry out the study from two perspectives: a) X-ray CT corefloods that give data on both saturation and pressure response in a foam displacement and b) data fitting to corefloods results.

Experimental studies are conducted with two types of simplified model oil to avoid confusion of the effect of multi-components of crude oil: hexadecane (C16) that is benign to foam and a mixture of C16 and oleic acid that is greatly detrimental to foam stability. CT scanning provides data on oil saturation (So) and distribution along a core during the corefloods. Such key information relates So to foam dynamics as implied by pressure response, in particular foam generation and propagation suggesting efficiency of oil displacement by foam. Specifically, foam is injected in two ways: co-injection of surfactant solution and gas with the intention to generate foam in situ in the presence of oil and injection of pregenerated foam. The co-injection serves to check the effect of oil on foam generation and propagation (anti-foaming effect) in the displacement. The second way investigates de-foaming effect of oil on foam displacement. The results show that foam generation in situ in the presence of oil is possible only when the oil is relatively benign to foam. When oil is very detrimental to foam, the in-situ generation is very difficult even at oil saturation as low as residual oil saturation. The displacement of pregenerated foam with oil consists of two stages of foam propagation due to change in So. Foam drains down So with a weaker strength due to high So in the primary propagation. In the secondary propagation, foam propagates faster with a larger strength, pushing oil forward like a piston.

Fitting coreflood data is then performed using a most widely used local-equilibrium implicit-texture foam model in CMG STARS™ simulator. We propose a fitting guideline for estimating foam simulation parameters based on steady-state foam flow data with oil in a recent study that does not give data on So. The model parameters estimated using this method serve as an initial guess for finding the optimal parameters to fit the coreflood data in simulation. A simulator for foam simulation in STARS is set up for this purpose. A scheme is then developed to optimize the model parameters in fitting the dynamic data. The challenges and difficulties in the fitting are briefly summarized for future work.