Integrated column experiments and reactive transport modeling for identifying chemical indicators of CO₂ leakage in a shallow granitic/gneissic aquifer

Leakage of CO₂ from carbon capture and storage (CCS) sites into shallow aquifers can alter groundwater chemistry, affecting parameters such as pH, electrical conductivity (EC), alkalinity, and cation concentrations. Monitoring these parameters is essential for early detection of CO₂ leakage. This study employed multicomponent reactive transport modeling to analyze geochemical changes observed in laboratory column experiments (push-and-pull and natural gradient tests) using soil and groundwater from the Environmental Impact Test (EIT) facility in Korea. The site mainly consists of a granitic/gneissic aquifer system. The model incorporated primary alumino-silicate minerals (quartz, microcline, anorthite, albite, and biotite) identified through X-ray diffraction (XRD) analysis and accounted for the precipitation of secondary minerals (gibbsite, calcite, and dolomite). Simulations successfully reproduced observed trends in pH, alkalinity, and major cations (Ca, Mg, Na, and K) and Fe, indicating that mineral dissolution, particularly of anorthite and biotite, was the primary buffering mechanism under acidic conditions. Gibbsite precipitation effectively limited Al mobility. Predictive simulations showed that geochemical responses were largely confined within 2.5 meters of the CO₂ source during the 30 days following injection, with delayed breakthrough at greater distances. These findings underscore the utility of reactive transport modeling for quantifying spatiotemporal geochemical evolution following CO₂ leakage. The approach supports improved monitoring well placement, sampling strategies, and site-specific risk assessments, and can be broadly applied to geochemical investigations in CCS and other groundwater systems.