EVALUATION OF MINERAL REACTIONS IN GEOLOGIC CO2 SEQUESTRATION SYSTEMS

Storing CO₂ in geological formations is one of the promising strategies to mitigate the greenhouse gas emission problem. The injected CO₂ will dissolve into the formation brine and ultimately be mineralized through geochemical reactions to achieve permanent storage. Understanding the reactions involved in this process can provide valuable insights on how the formation properties (porosity and permeability) will evolve and can inform injection rate, location of the monitoring well, etc. However, the CO₂-brine-mineral interactions are quite complex and challenging to simulate due to uncertainties in model parameters such as 1-5 orders of magnitude variation in mineral surface area estimation. Imaging has been proven to be a powerful means of quantifying mineral properties including porosity and mineral abundance, accessibility and accessible surface area. However, it requires time and has high computational costs. In addition, the impact of image resolution on these measured properties and ultimately on simulated mineral reactions is not well understood. In this work, the impact of image resolution on estimated mineral properties is first evaluated. Then, the impact of resulting variations in surface area on simulated reactions is evaluated. Rock samples extracted from Paluxy formation at Kemper pilot injection site, Mississippi were imaged using scanning electron microscopy under varying resolutions from 0.34 to 5.71 µm. Porosity and mineral abundance agreed relatively well with changing resolution while the accessibility of mineral phases with small-scale features decreases with decreasing resolution. Variations in mineral accessible surface areas are less than 1 order of magnitude. Image obtained accessible surface areas were then used in CrunchFlow to simulate the CO₂-induced reactions. We found that such variations in mineral accessible surface area resulting from image resolution have a very small impact on the simulated reactions. However, large discrepancies were observed when simulations were carried out using BET surface areas from the literature and geometric surface area. Reactive mineral phases such as carbonates were only impacted by surface area at short times while the more stable phases (e.g. feldspars and clays) were mostly impacted at longer times.