Gas trapping is an important mechanism in both Water or Surfactant Alternating Gas (WAG/SAG) and foam injection processes in porous media. Foams for enhanced oil recovery (EOR) can increase sweep efficiency as they decrease the gas relative permeability, and this is mainly due to gas trapping. However, gas trapping mechanisms are poorly understood. Some studies have been performed during corefloods, but little work has been carried out to describe the bubble trapping behaviour at the pore scale. We have carried out foam flow tests in a micromodel etched with an irregular hexagonal pattern. Image analysis of the foam flow allowed the bubble centres to be tracked and local velocities to be obtained. It was found that the flow in the micromodel is dominated by intermittency and localized zones of trapped gas. The quantity of trapped gas was measured both by considering the fraction of bubbles that were trapped (via velocity thresholding) and by measuring the area fraction containing immobile gas (via image analysis). A decrease in the quantity of trapped gas was observed for both increasing total velocity and increasing foam quality. Calculations of the gas relative permeability were made with the Brooks Corey equation, using the measured trapped gas saturations. The results showed a decrease in gas relative permeabilities, and gas mobility, for increasing fractions of trapped gas. It is suggested that the shear thinning behaviour of foam could be coupled to the saturation of trapped gas.

Foam structure evolves with time due to gas diffusion between bubbles (coarsening). In a bulk foam, coarsening behaviour is well defined, but there is less understanding of coarsening in confined geometries such as porous media. Previous predictions suggest that coarsening will cause foam lamellae to move to low energy configurations in the pore throats, resulting in greater capillary resistance when restarting flow. Foam coarsening experiments were conducted in both a model-porous-media micromodel and in a sandstone core. In both cases, foam was generated by coinjecting surfactant solution and nitrogen. Once steady state flow had been achieved, the injection was stopped and the system sealed off. In the micromodel, the foam coarsening was recorded using time-lapse photography. In the core flood, the additional driving pressure required to reinitiate flow after coarsening was measured. In the micromodel the bubbles coarsened rapidly to the pore size. At the completion of coarsening the lamellae were located in minimum energy configurations in the pore throats. The wall effect meant that the coarsening did not conform to the unconstricted growth laws. The coreflood tests also showed coarsening to be a rapid process. The additional driving pressure to restart flow reached a maximum after just 2 minutes.