Impact of humid supercritical CO2 flow and CO2 solvation-induced cement porewater expulsion on reactive self-sealing processes along wellbore microannuli

Abstract

Ensuring geological containment of injected CO2 is a key requirement in the planning and execution of Carbon Capture and Storage (CCS) projects. This includes a proper analysis, and if needed remediation, of the wellbore cement sealing integrity. Defects like casing-cement microannuli can impair this sealing integrity, especially if they expose the cement to flowing CO2-rich fluids. Chemical reaction between the cement and CO2 could in specific cases potentially enlarge seepage pathways, while in others reaction may seal off pathways via carbonate precipitation. Previous studies focused on these reactive transport processes during flow of CO2-rich brine, demonstrating small cement defects possess some definite capacity to self-seal. While this earlier work produced valuable insight, water-based flow experiments are not fully representative for CO2 seepage along real CCS wells. Actual seepage flows will likely be buoyancy-driven and consist mainly of humid CO2 (in its gaseous, liquid, or supercritical phase) rather than carbonated brine. Humid CO2 and carbonated brine have very different fluid properties, e.g., in terms of mineral solubility, which could profoundly change dissolution-precipitation processes and their impact on CCS well integrity. This study presents the first reactive flow-through experiments performed using humid CO2 as the flowing medium, allowing us to assess how the effective permeability of 2–20 μm wide casing-cement microannuli evolves under multiphase CO2 reactive transport at 60 °C and 7–9 MPa pressure. After reference measurements using humid N2, the results show that exposure to humid CO2 can produce large reductions in defect effective permeability, with all samples reaching down to intact cement values. However, the overall reaction extent and amount of precipitation observed were limited compared to earlier flow experiments using carbonated brine. It is inferred that part of the observed reductions in flowrate likely occurred due to a shift in multiphase flow dynamics upon switching from humid N2 to CO2-based flow. Despite these differences, previously established self-sealing criteria for cement defects under carbonate brine flow seem to hold for humid CO2 flow to within current experimental uncertainty.