The Co2crc Otway Shallow Co2 Controlled Release Experiment: Fault Characterization and Leakage Scenarios

Abstract

CO2CRC has made a significant investment into establishing the feasibility of conducting a CO2 injection experiment into a shallow fault. This world-leading experiment, located at the CO2CRC Otway International Test Centre in Victoria, Australia, would seek to improve our understanding of the conditions necessary for CO2 to move vertically up faults. The work undertaken during Phase 2 of the Otway Fault Project confirms the experiment is technically feasible and can be done safely. Two appraisal wells drilled and cored through the shallow Brumbys Fault indicate the fault extends to the base of the upper 2 m thick Hesse Clay layer, which forms the seal to the underlying Port Campbell Limestone aquifer. The fault does not have a defined core but an approximately 6-10 m wide fault cataclastic zone. Permeability within the Port Campbell Limestone is variable, ranging from tens to thousands of millidarcies. The rock strength is low, however, and it is recommended to conduct the experiment at approximately 80 m depth rather than the 40 m originally proposed. This provides more confining pressure and will ensure that the injection pressure does not exceed the fracture pressure. A deeper injection also provides better spatial and timing conditions for geophysical monitoring and tracking of the CO2 plume. Simulations indicate only a small 10 tonne CO2 injection experiment would be required to monitor CO2 migration using geophysical techniques. In addition to providing an opportunity to demonstrate semi-continuous, near real-time monitoring of CO2 migration up a fault, the planned CO2 injection experiment presents a unique opportunity to obtain field measurements on vertical fault permeability. It also provides an opportunity to predict fluid flow and potential metal remobilisation through comprehensive reactive transport modelling, collection and analysis of core post CO2 injection, and evaluate the effectiveness of the modelling versus field observations.