Faults represent a critical factor for the long-term security of Carbon Capture and Storage (CCS) projects, as they can act as potential pathways for CO₂ leakage from storage reservoirs to the near-surface. Understanding the behaviour of such faults under injection scenarios is therefore essential for assessing storage integrity. While numerous modelling studies have investigated fault reactivation and its implications for CO₂ migration, experimental validation under field conditions remains scarce.

In April 2024, a controlled CO₂ injection experiment targeting the shallow Brumbys Fault was conducted at the CO2CRC Otway International Test Centre in Victoria, Australia. To monitor the injection process, high-resolution cross-well seismic data were collected using two wells adjacent to the fault: one equipped with a fibre-optic cable connected to a Distributed Acoustic Sensing (DAS) system serving as the receiver array, and the other hosting a high-frequency (~1 kHz) source. Cross-well seismic tomography was applied to invert the times of first arrivals to create images of the P-wave velocity distribution within the subsurface cross-section. During injection, tomographic images revealed a low-velocity anomaly zone near the injection interval, and subsequently similar anomalies at shallower depths in post-injection monitoring indicating the upward migration of carbon dioxide. Later surveys indicated that this anomaly diminished over time, possibly due to CO₂ dissolution or gradual release. Furthermore, the suitability of this method for long-term monitoring was evaluated through several additional surveys. Time-lapse changes observed in the later monitors confirmed the effectiveness of this setup for extended seismic monitoring.

This experiment demonstrates the effective application of DAS with a high-frequency source for cross-well seismic monitoring. The resulting high-resolution tomographic images highlight the method’s capability to detect and track velocity changes associated with shallow CO₂ migration. These results suggest that the approach is well-suited for long-term monitoring of shallow faults and for assessing potential CO₂ leakage pathways.