Field-scale simulation of CO2 enhanced oil recovery and storage through SWAG injection using laboratory estimated relative permeabilities

Numerical simulations are widely used to investigate the performance of simultaneous water and gas (SWAG) injection in an oil field. However, they usually use two-phase water and gas relative permeability functions (k(rg)) and one of the known correlations (e.g., Stone, Baker) to calculate oil relative permeability in a three-phase displacement. Moreover, the same k(rg) is used for all SWAG injections, irrespective of miscibility conditions or fraction of gas injected (FGI). Recent experiments have shown that gas relative permeability functions in three-phase SWAG are significantly different from the conventionally used two-phase functions. This paper investigates the impact of using experimentally estimated three-phase gas relative permeability functions on co-optimization study of CO2 storage and enhanced oil recovery (EOR). A three-dimensional, layered reservoir model initially saturated with oil and connate water is used to examine different injection schemes for co-optimizing oil recovery and CO2 storage efficiency. The oil phase consists of a mixture of 0.65 hexane and 0.35 decane by molar fraction. Numerical simulations are run at 70 degrees C and at three pressures (1300, 1700, and 2100 psi) to represent immiscible, near-miscible, and miscible displacements, respectively. SWAG displacements are run at an FGI of 0.5 for immiscible, near-miscible and miscible conditions. Then the effect of FGI dependent relative permeability on co-optimization of CO2 storage and EOR is investigated in the near-miscible condition for FGI values 0.25, 0.5, 0.75 and 1.0. Experimentally estimated three-phase gas relative permeability results in lower CO2 mobility. This leads to higher CO2 saturation and lower water saturation at the end of simulation. The significant increase in CO2 storage efficiency and hence the co-optimization function are more prominent in near-miscible and miscible displacements than in immiscible. Although, for the homogeneous model, ultimate oil recovery is the same when two-phase or experimentally estimated three-phase relative permeabilities are used, ultimate recovery value is significantly affected in the heterogeneous reservoir model.