During CO2 fluid injection into oil reservoirs or saline aquifers, CO2-water-rock interactions can alter porous media properties, thereby influencing the CO2 microscale sweep efficiency, primarily due to capillary effects. Laboratory experiments and micro/nanoscale numerical simulations often struggle to isolate the specific contributions of individual pore properties, limiting targeted injection optimization for maximizing geological storage potential.
To investigate the dynamic evolution of pore structures and properties during multi-mineral competitive dissolution-precipitation reactions under CO2 injection, we developed a lattice Boltzmann method (LBM). This method made it possible to simulate CO2-water-rock interactions in shale oil reservoirs and to analyze pore properties (e.g., average wettability, roughness, porosity) and CO2 microscale sweep efficiency. The LBM simulations generated a dataset covering various pore property scenarios to support causal machine learning. Using a double machine learning framework with a random forest algorithm, a causal inference prediction model was built for CO2 microscale sweep efficiency, treating reaction time as a continuous treatment variable.
This model quantified the relative importance of key pore parameters—porosity, wettability, and mean pore diameter—on sweep efficiency within the pore network. Its results indicate that reservoirs with higher proportions of carbonate minerals (calcite) exhibit greater CO2 microscale sweep efficiency. The CO2-water-rock reaction triggers calcite dissolution, forming preferential flow paths, while the secondary precipitation of oil-wet calcite induces localized wettability alteration. This dynamic "dissolution–secondary precipitation" process modifies capillary forces by altering the structure and physical properties of pore-throats, thereby influencing the microscale sweep range of CO2 fluids. However, under identical mineral proportions, CO2 sweep efficiency varies among samples, with higher calcite proportions correlating with broader variation in sweep performance.
These findings underscore the crucial role of physical pore properties, beyond mineral composition alone, in governing sweep efficiency. Causal learning identified key pore-throat parameters that control CO2 microscale sweep behavior, with wettability emerging as the most influential factor. Neutrally water-wet pore-throats exhibited the highest CO2 sweep efficiency. [Significance] By constructing a Lattice Boltzmann model for CO2-water-rock interactions and quantifying the impact of key physical parameters, this study provides a reference and guidance for the targeted adjustment of CO2 injection strategies and the enhancement of the geological CO2 storage effectiveness.
| 科 Family | 属数 Number of genus | 种数 Number of species | 占总种数比例 Percentage of total species (%) | 属 Genus | 种数 Number of species | 占总种数比例 Percentage of total species (%) |
|---|---|---|---|---|---|---|
| 鹅膏菌科Amanitaceae | 2 | 11 | 5.26 | 鹅膏菌属 Amanita | 10 | 4.78 |
| 小菇科 Mycenaceae | 2 | 12 | 5.74 | 丝盖伞属 Inocybe | 5 | 2.39 |
| 多孔菌科 Polyporaceae | 8 | 14 | 6.70 | 蜡蘑属 Laccaria | 5 | 2.39 |
| 红菇科 Russulaceae | 3 | 23 | 11.00 | 小皮伞属 Marasmius | 6 | 2.87 |
| 小菇属 Mycena | 11 | 5.26 | ||||
| 光柄菇属 Pluteus | 5 | 2.39 | ||||
| 红菇属 Russula | 17 | 8.13 | ||||
| 栓菌属 Trametes | 5 | 2.39 |