Microstructure-informed simulation of wave propagation in fractured rock

Seismic surveys allow remote subsurface imaging of the subsurface inclusive of reservoirs and features such as fault damage zones. Seismic imaging is especially well-suited for locating fractures and fracture networks, but its utility in characterizing fracture morphology has been limited. Considering the key role of fracture aperture and connectivity in controlling the hydraulic properties of fractured rock, there is a need for the quantitative characterization of fractures for their hydromechanical properties. This necessitates establishing a quantitative and scalable mapping between fracture geometrical features and seismic wave signatures. We report on methods developed to connect seismic wave attributes to the hydromechanical characteristics of the fractures. We have conducted a series of coupled operando experiments with synchrotron X-ray computed tomography (CT) and ultrasonic imaging on stressed fractured rock (Berea sandstone). We analyze the X-ray CT images to estimate in-situ fracture aperture, true contact area, and the contact size distributions as they evolve with stress. The ultrasonic data are analyzed to obtain wave speed, amplitude and center frequency. Further, we use synchrotron X-ray CT images of fractures under stress to create high-fidelity numerical models ('digital fractured rock') of in-situ fractures and run numerical simulations (using the finite element method) of ultrasonic wave propagation in the fractured samples. The model parameters are chosen such that the numerical ultrasonic response is in good agreement with the experimental ultrasonic data. Here, we present the details of the numerical model and the comparison of numerical and experimental results. The resulting ultrasonic signatures are analyzed in relation to changes in fracture aperture and contact inferred from the time-lapse X-ray CT images. Such simulations can help upscale our laboratory-scale experimental results.