Mechanistic controls on hydrogen migration, trapping, and recovery in porous reservoirs

Abstract

Underground hydrogen storage (UHS) in geological formations is increasingly recognized as a cornerstone of low-carbon energy systems, offering large-scale, long-duration buffering to stabilize variable energy supply. Its feasibility in porous reservoirs such as saline aquifers and depleted hydrocarbon fields depends on coupled multiphase flow and biogeochemical processes, yet substantial uncertainty remains regarding the ability of porous media to sustain cyclic injection-withdrawal while preserving storage integrity, injectivity, and gas purity. Hydrogen displacement is controlled by relative permeability, residual trapping, gas redistribution, and pore-scale instabilities such as snap-off and capillary fingering, and advances in high-pressure microfluidics, micro-computed tomography, and interfacial characterization now enable direct observation of these processes. Compared with CO2 and CH4, hydrogen shows distinct interfacial behaviour, with endpoint kr,H2≈0.003–0.083 (about an order of magnitude below CO2-brine baselines) and fitted Brooks-Corey gas-phase exponents ng≈3–10 against ≈1.6–1.8 for supercritical CO2, and with storage performance further dependent on wettability, mineralogy, viscosity, and flow history. Hydrogen-rock wettability is not uniformly strongly water-wet but varies with mineralogy, organic content, and microbial activity, and salinity effects on contact angle are reported but not universal. Microbially mediated wettability alteration, together with mineral dissolution and secondary precipitation, can further modify porosity, permeability, and capillary behaviour under cyclic loading. A significant gap nonetheless remains between homogeneous laboratory conditions and the heterogeneity of real reservoirs, where the transition to rate-sensitive regimes critically affects gas connectivity and trapping, and where current models often neglect rate-dependent, microbially mediated, and pore-scale redistribution processes that evolve dynamically under cyclic operation. This review connects previously siloed treatments of diffusion, multiphase flow, geochemistry, and microbiology through a cross-scale, process-coupled framing of cyclic injection-withdrawal, and identifies standardized side-by-side benchmarking on identical lithologies as the methodological step needed to resolve the inter-study contradictions and to support multiscale, rate-sensitive predictive frameworks for safe, efficient, and scalable UHS deployment.