Quantitative monitoring of dynamic CO₂ saturation at the CaMI.FRS

Xiaohui Cai, Qi Hu, Tianze Zhang, Kristopher A. Innanen

Time-lapse seismic monitoring is essential for verifying the behavior of subsurface CO₂ storage, as it provides large-scale, high-resolution sensitivity to fluid migration. However,most field applications remain qualitative, and quantitative estimation of volumetric CO₂ saturation from seismic data remains challenging and relatively underexplored. Here we develop and field-test a fully integrated workflow that links time-lapse VSP data, elastic full-waveform inversion (FWI), and Bayesian rock-physics inversion to recover CO₂ saturation at meter-scale resolution. Based on high-fidelity 4D repeatability conditioning, elastic FWI yields robust time-lapse perturbations in V_P, V_S, and ρ, consistent with well-log constraints and with the expected elastic signatures of CO₂ injection. Bayesian inversion translates these elastic changes into porosity, mineral composition, and CO₂ saturation while quantifying uncertainty. The resulting saturation field reveals a coherent CO₂ plume aligned with reservoir expectations. This study provides one of the first field demonstrations that quantitative CO₂ saturation can be recovered from a single-well VSP through a fully physics-based inversion, enabling geothermal-scale resolution of subsurface changes.The workflow also establishes a critical foundation for closing the loop between seismic monitoring and reservoir simulation, offering a powerful new tool for reliable monitoring and verification of geological carbon storage.