Multi-physics approach to modelling near-miscible CO2-WAG process

Abstract

CO2 Water-Alternating-Gas injection (CO2-WAG) under near-miscible conditions entails interactions between multi-physical processes at different length scales, which is not fully understood mechanistically. This research contributes to our fundamental understanding of the fluid behaviour driven by major physical mechanisms including compositional effects (denoted as MCE), interfacial tension effects (denoted as MIFT), capillary forces and gas trapping. The total system size was of order 1 large-scale “grid block” (~50m), which allowed us to capture the multi-scale behaviour from the cm to the 50m scale. Using fine-scale 2D areal simulations, this study identifies the separate and combined contributions of mechanisms to the recovery of bypassed oil induced by viscous fingering. Mechanisms MCE and MIFT work in tandem to improve the oil recovery through stripping oil components and enhancing viscous crossflow. However, the magnitude of such benefits is highly dependent on the ancillary effects of gas trapping and the capillary forces (system wettability). In a water-wet system, gas trapping modestly constricts actions of MIFT whereas capillary forces significantly degrade the sweep efficiency and lead to multiple isolated oil zones. We also found that the negative impacts of capillary pressure on the oil recovery can be much reduced if CO2 is injected prior to any water injection. On the other hand, oil-wet capillary forces hardly decrease the ultimate oil recovery in our cases. For the first time, flow trajectories, phase occupancies, oil compositions and interfacial tension as a function of permeability are explicitly depicted for each WAG cycle, which effectively unpicks the complexity of near-miscible CO2-WAG process.