Monitoring water content variations from seismic noise in a controlled laboratory experiment

Despite the availability of various geophysical techniques for characterizing the fluids in geological reservoirs, implementing cost-effective, accurate, and efficient methods for long-term monitoring with high spatial resolution remains challenging. Recent studies increasingly suggest that seismological methods based on continuous seismic noise recordings could help address these challenges. This study presents a novel laboratory experiment designed to evaluate the sensitivity of passive seismic interferometry (PII) imaging to controlled variations in water content. To do so, we utilized seismic noise generated by a continuous seismic source to reconstruct ballistic surface Rayleigh waves propagating in the [200-500] Hz frequency range within a 1-meter sandbox. Multiple controlled cycles of water imbibition and drainage at the sandbox base result in significant changes in the seismic wavefield, particularly in the dominant surface waves. The large relative velocity variations (δv/v) measured in Rayleigh waves with fine temporal resolution closely match pressure measurements made within the sandbox. These observations are well explained by an original theoretical model combining the Biot-Gassmann-Wood poroelastic framework, which accounts for effective pressure fluctuations, with the frequency-dependent sensitivity kernels of Rayleigh waves. The results confirm the potential of the PII method for monitoring saturation changes in reservoirs and highlight the substantial impact of effective pressure fluctuations.