Advancements in electroresistivity processing and data inversion for monitoring CO2 leakage: Insights from a field experiment and empirical data analysis using the pyGIMLi library

Abstract

Monitoring CO₂ sequestration is critical for assessing the environmental integrity of Carbon Capture and Storage (CCS) projects. This study presents an innovative approach for monitoring CO₂ leakage in a shallow aquifer system in Viamão, Brazil, using electroresistivity tomography (ERT). During the experiment, 5–12.5 kg of CO₂ were injected daily into the subsurface, with geoelectrical sensors deployed to measure resistivity variations over time. Data processing and inversion were carried out using the pyGIMLi library, with a focus on optimizing the Lam regularization parameter to enhance model fitting. Results indicate that the Lam parameter plays a significant role in data quality, with increased discretization improving model accuracy. However, factors such as soil type, rainfall, and aquifer recharge rates also influenced the resistivity readings. Key findings include the importance of repeated data collection under stable climatic conditions, challenges posed by clay soils and heavy rainfall, and the necessity of high CO₂ injection rates for accurate detection of CO₂ migration. This study highlights the effectiveness of using open-source tools like pyGIMLi for CO₂ leakage monitoring, advancing the application of electroresistivity methods in environmental studies.