Leakage detection remains one of the most challenging aspects of monitoring in geological carbon capture and storage (CCS) projects. While time-lapse seismic methods have demonstrated success in imaging reservoir-scale CO₂ plumes, showing that small leaks in shallow formations can be reliably detected has been far less established. To address this gap, the Otway Shallow Release Project simulated a shallow leakage scenario by injecting 16.5 tons of gaseous CO₂ into a fault zone at depths of 77–87 m.

To build an image of the carbon dioxide plume’s evolution, daily reverse vertical seismic profiling (RVSP) surveys were conducted using a high-frequency borehole sparker source and a dense surface geophone array. Direct wave arrival times were picked for each of vintages and inverted through travel-time tomography. The difference of the velocity models relative to the baseline revealed a measurable CO₂-induced decrease in P-wave velocity. Within hours of injection start, an anomaly was observed at the injection depth, followed by upward migration along the fault zone. Over subsequent days, the anomaly split into two distinct low-velocity regions, one at the injection depth and another one at depths of 20–40 m and eventually extending to the near-surface.

These results provide some of the first field-scale evidence that 4D traveltime tomography of RVSP data can detect and monitor small-scale, fast-evolving CO₂ plume in shallow formations. The ability to acquire daily 3D vintages within hours demonstrates the applicability of this approach for real-time leakage monitoring. For CCS projects, this work highlights that high-resolution seismic methods can provide early warning of the gas migration along faults, supporting both containment assurance and the design of monitoring strategies before impacts on groundwater or the atmosphere occur.