Geochemical monitoring for carbon capture and storage projects: Quantifying risk through statistical methods

Carbon capture and storage (CCS) offers a practical approach for reducing greenhouse gas emissions. Subsurface or geologic storage of CO 2 , however, presents inherent risks of leakage with potential adverse environmental impacts, particularly to underground sources of drinking water (USDWs). Pre-injection monitoring of overlying aquifers can establish a geochemical baseline for analytes sensitive to potential CO 2 leakage into aquifers, enabling detection of changes to aquifer chemistry during or after CO 2 injection. Despite the importance of natural baselines in this monitoring approach, they are often established with limited pre-injection data collection, for example quarterly samples over a single year. This short-term approach limits evaluation of future aquifer chemistry samples within the context of standard statistical metrics (e.g. mean, standard deviation) for leak detection. Long-term records of aquifer chemistry can be analyzed to 1) evaluate the robustness of short-term monitoring to establish a natural baseline and 2) provide proxy records for areas of similar lithology where other data do not exist. Here, we analyze publicly available, long-term (more than 3 years) geochemical datasets, focusing on alkalinity, to evaluate statistical methods for assessing geochemical anomalies. We use simple, lithology-specific statistical models to establish a framework for interpreting post-injection geochemical data and differentiate between natural fluctuations and potential CO 2 leakage impacts. Ultimately, this study aims to contribute to the development of more effective geochemical monitoring strategies for CCS projects by providing a quantitative approach for the interpretation of geochemical datapoints that deviate from baseline measurements.