Simulation and modeling of three mechanisms of flow through porous media.

Abstract

The mathematical modeling of two-phase flow in saturated porous media, as well as the modeling of adsorption/retention behavior of surface active materials in a porous medium composed of a complex network of macro, meso, and micropores in 1-D, 2-D and axisymmetric cases and the displacement of two phases in capillary conducts, were studied. Also, a Comparison of two CFD tools is made (COMSOL and ANSYS Fluent). The results were compared with experimental data from the literature. For the saturated porous media study and mathematical modeling of adsorption/retention behavior, COMSOL software was used, while for the capillary conducts displacement, ANSYS Fluent was chosen. In the saturated porous media analysis, different ways to obtain permeabilities and their effects on the flow in saturated porous media were compared. For the mathematical modeling of adsorption/retention behavior of surface active materials in a porous medium different effects were analyzed: the selection of boundary conditions, the size of the tracer and surfactant signals, effects of reversible and irreversible adsorption, the difference between local equilibrium and the rate-limited process. Also, for 2-D and axisymmetric simulations, heterogeneities, blends of surfactant and dispersion/diffusion effects were studied. The proposed mathematical model compares favorably with experimental data from literature when taking macro and mesoporosity into account. This model should be helpful in guiding the design of dynamic adsorption experiments, and to understand how heterogeneities in the rock may influence the interpretation of experimental results. Finally, capillary displacement was analyzed in an axisymmetric system, and the results were compared against a onedimensional model and experimental results obtained through experimentation. This was done to identify the benefits of simulating this type of phenomena, because, sometimes a simplified model can hide vital information for the experiments. The axisymmetric simulations were superior when showing the complete information of the phenomena but at a much higher computational cost than the one-dimensional model