EOR solvent-oil interaction in clay-hosted pores: Insights from molecular dynamics simulations

H Xiong and D Devegowda and LL Huang, FUEL, 249, 233-251 (2019).

DOI: 10.1016/j.fuel.2019.03.104

Our understanding of EOR processes in shales, especially in organic or kerogen pore systems, has grown rapidly over the last few years. While clays are ubiquitous in shales and may make-up over 50% by volume of the shale matrix, there has been little to no focus on solvent-oil interactions in these pore systems. While hydrocarbon storage is likely to happen predominantly in the organics, the inorganic material might serve as conduits for fluid transfer from the organics to the fracture systems. This work uses molecular dynamics and numerical simulations to investigate the mechanisms governing EOR in clay-hosted pores to address this knowledge gap. Because illite is the most common clay mineral in shales, we use a slit-pore model of illite to simulate oil and oil- solvent mixtures in clay pores to quantify the extent of miscibility, diffusion, viscosity reduction and oil swelling of the oil under illite confinement. Temperature and the presence of water is expected to also control the rate of diffusion, viscosity, oil swelling and the degree of confinement. To account for these effects, our simulations span a wide range of temperatures and concentrations of water. Finally, the findings from molecular dynamics simulations are utilized in reservoir-scale EOR models to demonstrate their impact. Our results confirm that oil swelling, diffusion and viscosity reduction are a few of the dominant pore-scale mechanisms governing EOR. However, our findings also demonstrate that bulk fluid properties can cause overly optimistic predictions of the efficacy of EOR. Confinement in illite is shown to negatively impact miscibility for methane-rich solvents. Secondly the self-diffusion of fluids under illite confinement are only 30% of their values in the bulk. Most significantly, oil-solvent mixture viscosities under illite confinement are about two orders of magnitude higher compared to their bulk values. The presence of water exacerbates all of these effects and in fact compromises the solubility of the injected solvent. Commercially available reservoir simulators accommodate a whole host of reservoir and well architectures, but the transport equations are derived using bulk fluid properties. However, our work underscores the need for rigorous small-scale experiments and modeling studies to properly account for the effects of confinement that can have significantly compromise EOR efficacy in plays where a high percentage of the pore volume is contained in sub-20 nm pores. A model that does not take in to account nonporous confinement can be overly optimistic and compromise project economics, especially if connectivity is driven by the smaller pore throats.

Return to Publications page