Addressing the Challenges of Simulating EOR and CO2 Storage Projects at Reservoir Scale: Accurate Estimation of Trapped Gas Saturation

Abstract

Abstract One of the first and foremost steps in the feasibility analysis and site selection of geological CO2 storage projects is estimating the storage capacity of the appointed aquifer or depleted reservoir. It has been established that the volume of CO2 stored due to capillary trapping is significantly higher than other active mechanisms. Therefore, an accurate method is required to determine the trapped gas saturation in the system. This method is also of significant importance for simulating any other process involving cyclic injection of fluids in subsurface reservoirs as the hysteresis in relative permeability is a direct function of trapped gas saturation. Examples of the cyclic process in subsurface reservoir engineering are gas storage projects and reservoirs undergoing Enhanced Oil Recovery (EOR) injections. In this study, we present a detailed study of the reservoir scale simulation results using commercial software, and we discuss the challenges observed in calculating the trapped gas saturation. The first challenge is that Land's formulation relates the initial and residual non-wetting saturations measured during an imbibition cycle. However, in many reservoir blocks, the volume and rate of displacing fluid are insufficient to ensure reaching the residual values. Accurate determination of saturation histories in various reservoir grid blocks is also challenging as small oscillations make it hard to identify the flow reversal points. A significant amount of error is introduced in compositional simulations when the composition of trapped gas saturation enters mass transfer calculations, and the trapped gas is dissolved again in the oil phase. Whereas physically, it should be isolated and shielded by the water phase. Finally, an inaccurate definition of saturation-dependent functions can increase the error associated with calculating relative permeability data using trapped gas saturation. In this study, we present a new workflow for calculating the trapped gas saturation, addressing all the abovementioned issues. The backbone of this workflow is an efficient algorithm which removes any oscillation misidentified as a flow reversal point. The results discussed in this paper indicate that the available formulation in the literature should be deployed carefully (considering the active mechanism in the system) to decrease the uncertainties. As a result, the feasibility of EOR methods, the site selection for CO2 storage projects and the decision-making process can be based on more reliable data.