Using Minkowski Functionals to Improve the Accuracy of Residual Gas Saturation (compressed air, CH4, H2) from CT Core Flooding Experiments

Residual saturation is a key input parameter for modelling subsurface processes such as hydrocarbon (HC) gas production, gas injection, Water Alternating Gas (WAG) injection, as well as CCS and H2 storage. Since pore scale gas dynamics is subject to a wider range of physical effects than immiscible and insoluble liquids, such measurements cannot be performed using a model fluid system but require a representative gas-liquid system. In low-rate unsteady-state (USS) core flooding, the method mostly used in the industry, factors such as gas compressibility and ripening-induced gas/brine dissolution can lead to a high degree of uncertainty in estimating relative permeabilities and residual saturation (Sgr), with negative impact on field development decisions.
To address this problem, in-situ high resolution X-ray micro-CT (CT) imaging was used to monitor the pore-scale processes and Sgr during the flow experiments where gas-saturated brine displaced gas (compressed air/N2, CH4 and H2). The experiments confirmed the impact of gas/brine dissolution, i.e., a continuous decrease in gas saturation the more brine was injected, beyond the expected values of Sgr. This occurred even though the injected brine was fully equilibrated with the gas, and continued even after the brine injection was stopped, because of an intrinsic effect in the porous medium related to Ostwald ripening. In gas-liquid systems, capillary equilibrium involves diffusive transport between disconnected gas clusters through a super-saturated liquid phase. Therefore, injecting (non-supersaturated) gas-saturated brine leads to dissolution of gas in the pore space, making it very difficult to determine Sgr.
A new approach is proposed in this paper, where pore scale flow regimes are characterized by fluid topology, i.e., Minkowski functionals computed from the in-situ pore scale imaging. The transition from displacement to trapping dominated regime is characterized by the Euler characteristic reaching a stable value and the onset of the dissolution regime is characterized by the moment the interfacial area starting to decrease.
In the dissolution-dominated regime, the dissolution rate is gas-specific, and it scales with Henry’s constant times the diffusion coefficient. In the final trapping state, there are systematic differences in Sgr between the gases studied, with CH4 and H2 showing higher Sgr than compressed air/ N2. This has potentially significant implications: 1) as most of the Sgr and gas/water relative permeabilities are measured in the laboratory using compressed air/N2 and therefore it may imply that at least some of those experiments and results may need to be revisited; 2) as Sgr is the key parameter on recovery and/or storage in modelling.