Assessment of Geochemical Reactivity of Potential Geological Formations Characterized by Diverse Lithologies

Abstract

Global warming is driven by greenhouse gases, which trap heat in the atmosphere through the greenhouse effect. According to recent projections, rising greenhouse gas levels may drive global temperatures above 1.5ºC as early as 2040. Excess atmospheric CO₂ is typically removed through the natural carbon cycle; however, this cycle is becoming insufficient due to the burning of fossil fuels and human activities. Geologic carbon sequestration, the process of injecting CO₂ into deep geological formations, is a promising technology to mitigate the adverse effects of greenhouse gases. Accurate predictions of subsurface interactions are essential for evaluating the effectiveness of this long-term storage strategy. Reactive transport modeling (RTM) is widely used to simulate geochemical reactions and transport phenomena in subsurface environments. These simulations incorporate formation mineralogy, kinetic data, mineral surface areas, and site-specific temperature and pressure conditions. Therefore, applying accurate and representative parameters is crucial to obtaining meaningful results.

This dissertation presents four studies that advance the understanding of carbon sequestration in diverse geological settings. The first three chapters use RTM to evaluate geochemical interactions between injected CO₂ and formations in the southeastern United States. In the Washita-Fredericksburg formation, simulations showed muscovite dissolution and minor kaolinite precipitation, with limited evidence of secondary mineral trapping. The second study, focused on the Conasauga Group in northwest Georgia, revealed significant calcite and dolomite dissolution and a corresponding increase in porosity. The third study investigated the role of mineralogical heterogeneity in the Tuscaloosa Group. Despite compositional variability, simulation trends were consistent, showing minimal porosity change and sustained acidity, indicating limited buffering capacity. The final chapter develops a microanalytical method using µXRD and µXRF data to improve mineral phase identification in mafic and ultramafic samples. This method enhances the reliability of mineralogical inputs essential for RTM. Together, these projects underscore the importance of detailed mineral characterization and accounting for geological variability to improve predictions of long-term CO₂ storage. Understanding site-specific geochemical reactions is crucial to preventing risks and ensuring permanent storage.