MD-CFD simulation on the miscible displacement process of hydrocarbon gas flooding under deep reservoir conditions

In this study, the microscopic miscible mechanism and the macroscopic flow characteristics of hydrocarbon gas and crude oil are investigated by combining molecular dynamic (MD) simulation and computational fluid dynamics (CFD) simulation. Under the deep reservoir conditions, the MD results manifest that compared with water flooding, hydrocarbon gas flooding could be more miscible with oil molecules, due to their stronger interactions. Based on the calculated density, viscosity and diffusion coefficient by the MD method, the CFD model is constructed to study the miscible displacement process of hydrocarbon gas flooding at 413 K and 50 MPa. The CFD results illustrate that the ability of oil displacement of different injection mediums follows the order of methane > ethane > propane > water. Because of the oil-gas miscibility, hydrocarbon gas could weaken the negative effect of the porosity and the wettability, leading to the decrease of the residual oil (RO) content. Moreover, the multi-component injection method is proposed for multi-scale porous media including nano-pore and micron-pore with small pore size. Notably, the variable velocity injection method is firstly designed for enhanced oil recovery. Compared with the constant low-velocity method, the RO content and the injection time of this method are effectively reduced by 12.4% and 66.7%, respectively.