Elucidating geochemical response of shallow heterogeneous aquifers to CO2 leakage using high-performance computing: Implications for monitoring of CO2 sequestration

Abstract Predicting and quantifying impacts of potential carbon dioxide (CO 2 ) leakage into shallow aquifers that overlie geologic CO 2 storage formations is an important part of developing reliable carbon storage techniques. Leakage of CO 2 through fractures, faults or faulty wellbores can reduce groundwater pH, inducing geochemical reactions that release solutes into the groundwater and pose a risk of degrading groundwater quality. In order to help quantify this risk, predictions of metal concentrations are needed during geologic storage of CO 2 . Here, we present regional-scale reactive transport simulations, at relatively fine-scale, of CO 2 leakage into shallow aquifers run on the PFLOTRAN platform using high-performance computing. Multiple realizations of heterogeneous permeability distributions were generated using standard geostatistical methods. Increased statistical anisotropy of the permeability field resulted in more lateral and vertical spreading of the plume of impacted water, leading to increased Pb 2+ (lead) concentrations and lower pH at a well down gradient of the CO 2 leak. Pb 2+ concentrations were higher in simulations where calcite was the source of Pb 2+ compared to galena. The low solubility of galena effectively buffered the Pb 2+ concentrations as galena reached saturation under reducing conditions along the flow path. In all cases, Pb 2+ concentrations remained below the maximum contaminant level set by the EPA. Results from this study, compared to natural variability observed in aquifers, suggest that bicarbonate (HCO 3 − ) concentrations may be a better geochemical indicator of a CO 2 leak under the conditions simulated here.