Direct Multiphase Numerical Simulation on Mixed-Wet Reservoir Carbonates

To better understand local displacement efficiency, direct numerical simulations of water-flooding in a mixed-wet rock from a producing reservoir were performed using the multiphase Lattice Boltzmann (LB) method. Experimentally measured contact angles (AlRatrout et al., 2017) were incorporated into the simulation models using our previously reported wetting boundary condition for the LB method (Akai et al., 2018b). The simulation model was calibrated by comparing pore occupancy and fluid conductivity with results from an experimental water-flooding study where the fluid configurations were imaged at a resolution of a few microns (Alhammadi et al., 2017, 2018). Furthermore, to investigate the impact of several enhanced oil recovery (EOR) schemes on recovery, the calibrated simulation model was also used for a sensitivity study. Taking the calibrated model as a base case, three EOR cases were investigated; low salinity water-flooding, surfactant flooding and polymer flooding. For low salinity water-flooding, the wettability of pore walls was changed to be more water-wet than that of the base case. For surfactant flooding, the interfacial tension was reduced. For polymer flooding, the viscosity of injection water was increased. A significant change in oil recovery factor was observed in these cases. These results make it possible to better understand the impact of EOR schemes on microscopic recovery. We demonstrate the predictive power of our direct numerical simulation by presenting comparisons of the fluid distribution at the pore-scale between the experiment and simulation. Then, we show how direct numerical simulation helps understand EOR schemes. This work provides a comprehensive workflow for pore-scale modeling from experiments to modeling.