| dc.description.abstract | In recent years, the rapid decline in the production trend of liquid-rich shale reservoirs calls attention to enhance oil recovery (EOR) methods. One of the most promising EOR methods in unconventional reservoirs is a miscible gas injection with soaking, but this requires a more extensive reservoir fluid characterization than for conventional reservoirs to consider the confined behavior of fluids in nanopores. The Peng-Robinson Equation of State (PR-EOS) is coupled with the Young-Laplace capillary pressure model to include the nanopore confinement effect into the fluid behavior model. The study’s general objective is to predict the performances of different gas-assisted EOR methods in unconventional liquid-rich reservoirs with a more compatible and reliable thermodynamic-based model. We extend the features of an in-house developed pressure-volume-temperature (PVT) simulator to calculate minimum miscibility pressure (MMP), an essential property to design a miscible gas injection, and model a gas injection process by molecular diffusion using the appropriate effective diffusion coefficient in a tight-shale oil reservoir.
Previous work on conventional reservoirs has proven that injecting at or above the MMP can significantly increase recovery. However, the nanopore confinement effect in MMP estimation has not been well-documented, and there is no consensus in the modeling approaches. Numerical MMP estimation is more computationally efficient than molecular simulation and produces faster results than laboratory experiments. In this study, we developed an MMP calculation that includes the nanopore confinement effect. Since phase equilibria calculations have been coupled to the Young-Laplace capillary pressure model, a feasible MMP calculation method for this study is the Multiple-Mixing Cells method that calls for flash calculations for each of the mixing cells to obtain dynamic multiple-contact miscibility. MMP predictions without confinement effect were compared with results from a commercial PVT simulator to test the accuracy. After modifying the MMP calculation, we examined multiple MMP estimations with varying fluid components, pore sizes, and reservoir conditions.
Other than fluid miscibility, the flow mechanism in unconventional ultralow permeability plays dominated by molecular diffusion significantly impacts oil production from the gas-assisted recovery process. To simulate this, we need to select the appropriate effective diffusion coefficient. We solved the diffusion equation with and without the swelling-induced convection term numerically in the PVT simulator. The oil production was calculated based on the excess swollen volume of the solvent-oil mixture. Then, we matched the recovery factor with the existing experimental data by adjusting the effective diffusion coefficient. Thus, this study provides a tool that can describe the recovery process quantitatively and qualitatively. | en |
