Nanoconfined Water Switch in Porous Media

In unsaturated soils and rocks, nanoconfined water films commonly coat mineral surfaces, thereby reducing fluid flow and transport rates. However, the underlying mechanisms by which these nanoconfined water films regulate flow and transport remain poorly understood. In this study, molecular dynamics (MD) simulations were employed to investigate the structural characteristics and transport effects of nanoconfined water films with varying thicknesses under different temperatures, pore sizes, gaseous environments, pressures, and salinity. The simulation data were further utilized to train a machine learning model. The results reveal the existence of a critical thickness for nanoconfined water films. When the film thickness is below this threshold, the film remains stable and exerts a negligible effect on gas transport. Conversely, when the thickness exceeds the critical value, the film becomes highly unstable, ruptures rapidly to form a water bridge, and significantly impedes gas transport. This behavior gives rise to a distinctive switch-like phenomenon in gas transport, which we term the Nanoconfined Water Switch. To broaden the applicability of this mechanism, we further developed a machine learning model capable of predicting the operational state of nanoconfined water switches in caprock nanopores. This approach provides valuable insight into natural gas reservoir exploration and the identification of secure underground gas storage sites.