Sequestration of carbon dioxide in deep underground reservoirs has been discussed for the reduction of atmospheric greenhouse gas emissions in the short- to medium-term until more sustainable technologies are available. Cost and long-term stability are major factors in adoption, so techniques to improve the storage efficiency and trapping security are essential. Such improvements require modeling of the porous geological formations involved in the sequestration process, and comparison to both lab- and field-based experimental studies. To this end, we are developing a comprehensive, large-scale pore-network model to describe multi-phase flow in porous media, including the structural, dissolution, and mineral trapping regimes. To explore the optimal operating parameters for mineralization trapping, we describe a two-phase pore-network model of brine-saturated aquifers and model the invasion of supercritical carbon dioxide (CO2) into the pore structure. Regularly-aligned 2D and 3D pore networks are constructed, and rules-based transport models are used to characterize the saturation behavior over a range of viscosity and capillary parameters, and coordination numbers. Finally, saturation patterns are presented for model caprock and sandstone reservoir conditions, taking into account different contact angles for CO2 on mica and quartz at supercritical conditions. These saturation patterns demonstrate the importance of surface heterogeneities in pore-scale modeling of deep saline aquifers.
Applying ANN, ANFIS, and LSSVM Models for Estimation of Acid Solvent Solubility in Supercritical CO2
