Pore-Scale Simulation of WAG Floods in Mixed-Wet Micromodels

Abstract

Summary In this paper, the simulation is described of Water-Alternating-Gas injection (WAG) flood cycles in 2D etched glass mixed-wet micromodels, using a 3D pore-scale network model for three-phase immiscible flow in porous media of arbitrary wettability. Although most network model input parameters can be explicitly derived from the experiments, the precise wettability parameters are not directly available. Therefore, a sensitivity study was conducted, using the network model in 2D mode, to obtain the wettability characteristics (i.e., the contact angle values and distribution) and the fraction of water-wet pores. Good qualitative and quantitative agreement was found between the experimental and simulated recoveries over the various WAG cycles, and the final residuals were well reproduced (as well as some observed "random recovery jumps"). The simulated displacement statistics showed many so-called multiple-displacement chains involving oil, up to approximately the third WAG cycle. The experimental and simulated fluid distributions were generally in good agreement in that: (a) different gas fingers were observed during various gasfloods, (b) oil movement was observed mainly during the first WAG cycles, and (c) during waterfloods, significant amounts of gas were displaced. Additionally, previously described simulations of water-wet and oil-wet experiments are compared with the present mixed-wet simulations. There are close similarities between the mixed-wet and oil-wet cases, which both maintain some continuity of oil through wetting films, but these cases are quite different from the water-wet case, which has continuity of water through wetting films in all pores. This paper further validates the pore-scale mechanisms incorporated in a network model capable of predicting three-phase relative permeabilities and capillary pressures for complicated processes, such as WAG.