A pore morphology-based initializer to accelerate lattice Boltzmann simulations of capillary-dominated flow with variable wettability

Direct numerical simulations are fundamental for investigating multiphase flow dynamics in porous media, yet their high computational cost limits application for simulating a wide range of flow conditions. The lattice Boltzmann method (LBM), in particular, becomes numerically challenging in capillary-dominated regimes. To address these limitations, we introduce a pore morphology-based initializer (PMI) that incorporates wettability to optimize initial fluid distributions and accelerate LBM convergence. Across a range of fluid saturations, wettability conditions, and capillary numbers (Ca), the PMI reduced the time step to steady state by up to 60% at Ca=O(10−4), and by two orders of magnitude at Ca=O(10−5). These gains arise from initializing simulations close to equilibrium, enabling stable continuation into regimes governed solely by capillary forces, processes that standard LBM rarely reaches due to numerical instability and computational cost. Overall, this approach provides a computationally efficient pathway to probe capillary-dominated fluid dynamics, offering critical insights for the prediction and optimization of secure CO2 and H2 storage in reservoir rocks.