Assessment of the Relative Diffusion Characteristics of Hydrogen and Methane System for Underground Hydrogen Storage in Depleted Gas Reservoirs: An Investigation Utilizing Molecular Simulation and Pore Network Modeling

Abstract

Abstract Depleted gas reservoirs are regarded as ideal subterranean sites for hydrogen storage; however, the intricate geochemical reactions within the reservoirs, particularly concerning the hydrogen-consuming methane production reaction occurring within the strata, present challenges for accurately assessing the diffusion characteristics of this binary system. To address this, we develop a multiscale workflow that couples molecular simulations with a pore-network model. Intrinsic, confinement-aware binary diffusivities for H2–CH4 are obtained from MD/GCMC at representative throat sizes and operating conditions. These size- and pressure-dependent diffusivities are then upscaled into a cubic pore-network parameterized by a truncated-Weibull throat-size distribution. At the pore scale, binary diffusion is formulated using the Maxwell–Stefan framework, while a stoichiometrically consistent hydrogen-consuming source term represents methanation within a prescribed reaction zone. By solving the governing equations of the reaction-transport coupled model, a quantitative relationship between the relative diffusion coefficient of the hydrogen-methane system and hydrogen concentration was established, providing a critical foundation for evaluating the diffusion effects during hydrogen storage. The findings indicate that the effective diffusion coefficient of the H2-CH4 system is governed by multiple factors, including hydrogen concentration, reservoir pressure, and chemical reactions. A significant negative correlation was observed between the diffusion coefficient and pressure; for instance, as pressure increased from 5 MPa to 30 MPa, the effective diffusion coefficient of H2 decreased from 4.48×10-7 m2/s to 1.31×10-7 m2/s. Conversely, a positive correlation exists between the diffusion coefficient and H2 concentration. Under conditions of 40°C and 10 MPa, when the molar fraction of H2 increased from 0.4 to 0.6, the effective diffusion coefficient rose from approximately 5×10-7 m2/s to 4×10-6 m2/s. The chemical reaction consumes H2 and produces CH4, leading to a reduction in H2 concentration, which subsequently inhibits the effective diffusion of H2. In summary, the pore network model proposed in this study, which couples reaction and transport processes, provides an effective tool for quantitatively assessing the diffusion kinetics of hydrogen storage in depleted gas reservoirs, offering essential insights for the optimization of storage solutions and the long-term safety evaluation thereof.