Upscaling gas–water relative permeability in hydrate-bearing sediments: A discrete element method and lattice Boltzmann method study based on micro-computed tomography

Accurately characterizing gas–water multiphase flow behavior (e.g., gas–water relative permeability) in porous media is essential for simulating and optimizing subsurface fluid transport. While relative permeability is conventionally obtained through coreflooding experiments, the inherent complexity of hydrate-bearing sediments (HBSs) makes laboratory measurements highly challenging. Moreover, stress-induced pore deformation in highly stress-sensitive HBS alters the pore network topology, significantly impacting macroscopic flow behavior. This study develops a robust upscaling framework that sequentially couples the discrete element method (DEM) with the lattice Boltzmann method (LBM), integrating micro-computed tomography (micro-CT) imaging to predict Darcy-scale gas–water relative permeability in HBS. Digital samples reconstructed from micro-CT images are represented using multi-sphere clumps to capture intrinsic pore-scale heterogeneity. The DEM simulates stress-induced structural deformation, and the Shan–Chen LBM subsequently evaluates gas–water two-phase flow behavior within the deformed pore structure. Crucially, the upscaled results are validated by their good agreement with established experimental data for both stress-dependent absolute and relative permeabilities. Results reveal that higher confining pressure increases flow channel heterogeneity and local capillary resistance. Consequently, water mobility is restricted due to enhanced capillary trapping, while gas relative permeability exhibits a minor increase. The demonstrated stress-seepage coupling, where severe absolute permeability degradation dominates flow behavior and hinders gas migration, highlights the critical need to incorporate this coupled geomechanical and gas–water flow behavior in reservoir simulators for accurate field-scale forecasting.