| Description |
A critical aspect of geologic carbon storage, a carbon-emissions reduction method under extensive review and testing, is the ability to simulate multiphase CO2 flow and transport. Relative permeability is a flow parameter particularly critical for accurate forecasting of multiphase behavior of CO2 in the subsurface. Specifically, for clastic formations, small-scale (cm) bedding planes can have a significant impact on multiphase CO2-brine fluid flow, depending on the relative permeability relationship assumed. Such small-scale differences in permeability attributable to individual bedding planes may also have a substantial impact on predicted CO2 storage capacity and long-term plume migration behavior. A major goal of this study was to evaluate and calibrate relative permeability models against experimental data to improve simulation capability. We analyzed previouslypublished laboratory-scale measurements of relative permeability of Berea sandstone, and developed a corresponding three dimensional simulation model of those measurements. The simulation model was created in the TOUGHREACT reactive transport simulator, and we elucidated best-fit relative permeability formulations to match the experimental data. Among several functions evaluated, best-fits between simulation results and experimental observations were achieved with a calibrated van Genuchten-Mualem formulation. To extend the analysis to a more heterogeneous medium, we applied the best-fit relative permeability formulations to a new model of a small-scale Navajo Sandstone reservoir. The model was one cubic meter in size, with eight individual lithofacies of differing permeability, gridded to mimic small-scale bedding planes. For this model we iv assumed that each lithofacies exhibits a random permeability field, resulting in a model with heterogeneous lithofacies. We then evaluated four different relative permeability functions to quantify their impact on flow results for each model, with all other parameters maintained constant. Results of this analysis suggest that CO2 plume movement and behavior are significantly dependent on the specific relative permeability formulation assigned, including the assumed irreducible saturation values of CO2 and brine. More specifically, different relative permeability formulations translate to significant differences in CO2 plume behavior. |
