Influence of heterogeneous mixed-wettability evolution on different phases CO2 displacement and storage efficiency mechanisms: Based on pore structure characterization

The interplay between complex pore structures and wettability significantly influences CO₂ storage behavior in underground reservoirs. To address the heterogeneity induced by their coupling, a pore-structure-based quantitative characterization method for complex-mixed wettability was proposed. A two-dimensional heterogeneous wettability model was constructed using in-situ microscopic imaging of sandstone core cast thin sections and employed to simulate multiphase CO₂ displacement. Results indicate that, under single wettability, Sc-CO₂ (Supercritical Carbon Dioxide) displays enhanced storage efficiency in strongly wetting pore walls, but gaseous CO₂ is more effective in weakly wetting systems. During the late stage of gaseous CO₂ injection, a transition from non-wetting to wetting behavior was observed. At contact angles of 30° and 60°, capillary imbibition caused partial reabsorption of previously displaced water, producing a peak in the storage efficiency curve—most evident at 30° and absent at 90° Under mixed-wettability, the spatial distribution of contact angles notably altered the multiphase flow behavior. A dual driving mechanism of capillary force and displacement pressure enhanced Sc-CO₂ storage efficiency by 2.70 % and 7.40 %, respectively, relative to gaseous CO₂. Six pore-wall configurations with varying contact angle distributions were designed to represent mixed-wettability states. Comparative analysis revealed that as the proportion of weakly wetting walls increased, Sc-CO₂ efficiency declined, whereas that of gaseous CO₂ improved. Both Sc-CO₂ and gaseous CO₂ exhibited reduced storage efficiency under complex mixed-wettability compared to single or mixed conditions. These findings highlight that neglecting mixed-wettability heterogeneity may result in overestimation of CO₂ storage performance and must be meticulously considered in prediction modeling and field-scale assessments.