Abstract
Surfactant flooding is a common enhanced oil recovery technique that has been researched for its scientific potential, but inhibited by its economics. This method relies on greatly reducing interfacial tensions between phases, allowing the displacement of previously immobile oil. For this process to succeed on an industry scale the high chemical cost involved must be outweighed by the increased oil production. One way to decrease the cost of surfactant chemicals, is to ensure the slug sized used is no larger than necessary. Accounting for the physical mixing in a compositional simulator plays a large role in determining the optimum slug size and success of the project.
Mixing occurs when the surface area between concentration gradients expands, increasing diffusion. In reservoir simulators, the amount of numerical dispersion can alter greatly depending on the model structure. To achieve a model reflecting a physical case, the mixing influencing surfactant concentrations must match the mixing occurring in the reservoir. The more heterogeneous a reservoir is, the more physical mixing will occur as the area of contact increases significantly with length traveled. Because dispersion is scale dependent, matching the physical amount of dispersion by varying the gridblock size can be challenging. As a numerical model is upscaled, the physical mixing decreases while the numerical dispersion associated with the grid increases.
To show the effects of the changing gridblock scale and heterogeneities, two dimensionless parameters were used to test various models for their optimum surfactant-polymer slug size. Both one and two-dimensional models were created in UTCHEM to simulate a surfactant injection process by altering one of the dimensionless parameters, and observing the optimal slug size. A matured water flood was followed by chemical flooding. The optimal was found using the incremental efficiency of the surfactant, taking the incremental oil recovered divided by the mass of chemical injected. The incremental efficiency curves display a single optimum for each case, which closely tie to the economic potential of the process. With knowledge from core flood experimentation, when looking to apply at field scale, it is vital to account for the difference in the mixing effects. This work could result in a better understanding of the optimal surfactant volume needed, greatly reducing chemical costs.
