Gas clogging in porous media driven by dissolved-gas disequilibrium: a mechanistic study based on experiments and process-based modeling.

Gas clogging can reduce the efficiency of managed aquifer recharge (MAR), yet its controlling mechanisms remain insufficiently quantified. Using sand-column experiments combined with a process-based model, we investigated whether hydraulic-conductivity decline in clean sand is primarily associated with externally supplied micro-nano bubbles or with dissolved-gas transfer and growth of pre-existing trapped gas. The measured influent micro-nano bubble population was dominated by submicron sizes and represented a very small detectable gas volume, making direct pore filling by this resolved fraction alone unlikely to account for the observed hydraulic decline. Inlet-outlet dissolved oxygen measurements indicate non-conservative dissolved-gas behavior within the sand column, consistent with gas exchange between the flowing water and an immobile trapped gas phase. Across the six experiments, relative hydraulic conductivity decreased by 53.7-85.3% during micro-nano bubble-water recharge, whereas substantially smaller changes occurred in the pure water controls. Clogging intensity increased with recharge rate and higher initial DO conditions, consistent with coupled effects of dissolved-gas condition and advective gas supply. The coupled model reproduced both hydraulic and DO responses, with K/K0 RMSEs of 0.070 and 0.045 and outlet-DO RMSEs of 0.145 and 0.126 mg L-1 for E4 and E2, respectively. Overall, the results suggest that gas clogging under clean-sand conditions is more consistent with dissolved-gas transfer and growth of pre-existing trapped bubbles than with direct pore blockage by the measured influent bubble population.