Foam flooding, as an enhanced oil recovery (EOR) process, has become of great interest to the oil industry. However, the mechanism of oil displacement by foam has not been fully understood. This thesis describes the visualization of foam on a pore-scale level using microfluidic devices. A micro model patterned with a heterogeneous porous media was designed to mimic reservoir with permeability contrast. The micro model is made with Norland Optical Adhesive 81 (NOA81), an optical transparent material with high organic resistance available for crude oil tests. Foam was injected into the reservoir micro model to displace trapped crude oil. The smart rheology of foam can help mitigate reservoir heterogeneity by separating into dry- and wet-phases such that the flow resistance in high-perm region is much higher than that in low-perm region. Bubbles could block the high-perm region and divert more fluids to displace the oil in low-perm part. One hypothesis is that the phase separation of foam is due to the difference in capillary entry pressure. Thusly foam shows a better sweep efficiency compared to gas and water flooding.

This thesis also compared different injection foam quality for different surfactant solutions. Results showed that higher foam quality has better sweep efficiency, especially in low-perm region. This is because high foam quality can create denser lamellae that increase flow resistance, which will thusly divert more liquid phase into low-perm region, improving oil displacement. Foam made with mixture of Lauryl Betaine (LB) and C14-16 Alpha Olefin Sulfonate (AOS) showed a better sweep compared to foam consisting of only AOS. The reason is that AOS foam tended to collapse more easily, changing to very coarse foam.