The relative permeability of CO2-oil in sandstone plays a key role in characterizing fluid transport and migration within reservoirs and is also significant for the development of CO2 injection models for oil extraction. Due to the complex pore structure of sandstone reservoirs and the miscibility between oil and CO2, conventional two-phase relative permeability models do not effectively meet the computational requirements. Therefore, this study proposes a new theoretical model to determine the gas relative permeability (GRP) and oil relative permeability (ORP) at different oil saturation, which is based on the well-established gas–water relative permeability models, utilizing nuclear magnetic resonance (NMR) testing to partition the pore space. It innovatively combines the oil occurrence state in pores of sandstone with fractal theory and introduces the mole fraction of oil (Mvo) in the mixed-phase region to characterize the degree of CO2-oil miscibility. The model's validity and accuracy are verified through comparison with experimental data and classic models. In the subsequent sensitivity analysis, the study investigates the effects of oil occurrence state, critical pore diameter, fractal dimension, oil viscosity, and mole fraction of oil on relative permeability, yielding the following innovative results: neglecting pores containing only bound oil or movable oil can lead to an underestimation of GRP by nearly 70%, and the error in ORP can be as high as four times. Furthermore, the critical pore diameter and fractal dimension also have significant impacts on gas–oil relative permeability, while the effects of oil viscosity and oil mole fraction are relatively minor.
Qi et al. (Tue,) studied this question.