Dynamic Reservoir Rock Typing for Supercritical CO2-Brine System in Sandstone

Abstract

Abstract Dynamic reservoir rock typing is a critical yet infrequently explored aspect of CO2 storage, essential for evaluating flow characterization in dynamic reservoir modeling within aquifer reservoirs. This study introduces a new insight into the establishment of dynamic reservoir rock types for the supercritical CO2-brine system, leveraging relative permeability data and implementing it into 3D numerical reservoir simulation. Our research draws on 22 sandstone core plugs obtained from potential CO2 storage aquifers in Alberta, Canada, encompassing measurements of relative permeabilities during primary drainage and secondary imbibition cycles. The rock typing methodology employed incorporates pore geometry and pore structure (PGS), in conjunction with the True Effective Mobility (TEM) function, to comprehensively characterize multi-phase fluid flow properties in rocks. Subsequently, we visualize the outcomes of the rock typing process through 1D and 3D model representations, including the simulation of flow characteristics through compositional numerical modeling for geological CO2 storage. As a result, four rock groups were established based on pore geometry and pore structure relationships in the studied samples. The critical findings are that the obtained results demonstrate clear groupings of similar TEM-function curves based on relative permeabilities of both brine and CO2, observed in both drainage and imbibition experiments. Averaged relative permeability curves were derived from the TEM-function and subsequently converted them into conventional relative permeability values for each rock type. Notably, 3D numerical simulations of flow dynamics unveiled unique and contrasting multi-phase fluid behavior within each rock group, particularly evident in saturation profiles over time. Furthermore, we evaluated the correlation between residual CO2 trapping and irreducible water saturation within the rock samples. Our findings suggest an inversely proportional relationship, indicating that higher irreducible water saturation leads to lower residual CO2 trapping. As a novelty, combining PGS rock typing and TEM-function analysis facilitated the effective and efficient grouping of capillary pressure and relative permeability data, ensuring high consistency and minimized overlap in each rock type. Moreover, this approach offers an alternative solution for averaging relative permeability data within each rock type that can greatly reduce the uncertainty of defining relative permeability input and accelerate the process of dynamic reservoir modeling.