Lawrence Berkeley National Laboratory

The ability to characterize time-varying reservoir properties, such as the state of stress, has fundamental implications in subsurface engineering, relevant to geologic sequestration of CO2. Stress variation, here in the form of changes in pore fluid pressure, is one factor known to affect seismic velocity. Induced variations in velocity have been used in seismic studies to determine and monitor changes in the stress state. Previous studies conducted to determine velocity-stress sensitivity at reservoir conditions rely primarily on laboratory measurements of core samples or theoretical relationships. We have developed a novel field-scale experiment designed to study the in situ relationship between pore-fluid pressure and seismic velocity using a crosswell continuous active-source seismic monitoring (CASSM) system. At the Cranfield, Mississippi, CO2 sequestration field site, we actively monitored seismic response for five days with a temporal resolution of 5 min; the target was a 26 m thick injection zone at approximately 3.2 km depth in a fluvial sandstone formation (lower Tuscaloosa Formation). The variation of pore fluid pressure was obtained during discrete events of fluid withdrawal from one of the two wells and monitored with downhole pressure sensors. The results indicate a correlation between decreasing CASSM time delay (i.e., velocity change for a raypath in the reservoir) and periods of reduced fluid pore pressure. The correlation is interpreted as the velocity-stress sensitivity measured in the reservoir. This observation is consistent with published laboratory studies documenting a velocity (V) increase with an effective stress increase. A traveltime change (dt) of 0.036 ms is measured as the consequence of a change in pressure of approximately 2.55 MPa (dPe). For T 1/4 13 ms total traveltime, the velocity-stress sensitivity is dV/V/dPe 1/4 dt/T/dPe 1/4 10.9 × 10-4/MPa. The overall results suggest that CASSM measurements represent a valid technique for in situ determination of velocity-stress sensitivity in field-scale monitoring studies.