Multiscale Coupling between Molecular Simulations and Reservoir Simulator: Geochemical Reactions for Low Salinity Water Injection in Carbonates

Abstract

This work, is based on the multiscale coupling between molecular simulations and reservoir simulators, to explore the brine composition for enhanced oil recovery via the low salinity water injection (LSWI) processes. To achieve this goal, molecular simulations were performed, providing physical-chemistry parameters to reservoir simulators and validate the proposed brine compositional model. The key data required within reservoir simulators are related to the chemical reactions, which are occurring due to the LSWI process, such as their free energies, kinetic constants, ionic strengths, chemical activities, and activation energies. To improve the accuracy of this input dataset, the main aqueous phase geochemical reactions were mapped, adsorption energies of hydrocarbons and brine ions on calcite surface were determined and ions-bearing calcium carbonate were evaluated. The calculations were based on the density functional theory (DFT) and classical molecular dynamics (MD) using Quantum-ESPRESSO and LAMMPS codes, respectively. The geochemical reactions that take place at mineral dissolution and ionic release, related to the LWSI process (MgSO4, CaSO4, BaSO4, Na2CO3, and CaCO3), were also determined. The obtained chemical equilibrium showed that the MgSO4 dissolution reaction was favored, while other minerals did not show a similar trend. Adsorption studies of organic the molecules naphthalene and anthracene over different surface sites were performed. The adsorption energies were similar for both molecules, where the most favorable configuration has the rings oriented parallel to the mineral surface. The potential of mean force obtained for brine ion adsorption suggested that there were no barriers for adsorbing Ca2+ and CO32- brine ions on calcite surface. In contrast, the other ions adsorption (Na+ and Cl-) have presented higher estimated activation energies. The energetic difference showed that the SO42- incorporation in calcite is more favorable than Mg2+. The Ba2+ showed unfavorable incorporation energy. The thermodynamic properties (free energies, entropies, and heat capacities) were calculated from the vibrational properties. Obtaining such input data by molecular simulations can significantly reduce uncertainties, by increasing the reservoir simulators predictive power, facilitating the optimization and understanding of the processes involved in the injection of low salinity fluids. From these results, the obtained equilibrium constants, free energies and adsorption energies can be used as input data in further reservoir simulators. In addition, it would allow the validation of the proposed model from the understanding of the physical processes underlying LSWI.