Time-Lapse Seismic Imaging of Hydraulic Fracturing at Sanford Underground Research Facility Using Continuous Active Source Seismic Monitoring

Hydraulic fracturing experiments were conducted at Sanford Underground Research Facility as part of a project established to better understand subsurface fracture creation at Enhanced Geothermal System (EGS) sites. A field site was established 4,850 feet below land surface in crystalline metamorphic rock consisting of an array of five nearly-vertical boreholes with 3 m lateral separations, which were designated for hydraulic fracturing and geophysical monitoring. Crosswell geophysical instrumentation included electrical resistivity tomography and active source seismic monitoring using a stationary array of automated piezoelectric sources. The scope of this study was limited to evaluating the utility of continuous active source seismic monitoring (CASSM) for monitoring and characterizing hydraulic fractures deep underground. Data collected before, during, and after hydraulic fracturing were analyzed for P- and S-wave travel-time variations smaller than the original sampling rate of the equipment using seismogram interpolation. Travel-time variations were observed for both P- and S-waves and correlated well with the multiple pressure cycles used during fracture stimulation. Travel-time variation data were used to estimate an apparent fracture diameter of 6 meters and an apparent aperture of 0.38 millimeters. Time-lapse tomographic models delineated the fracture's location and provided insight on its behavior during stimulation. The fracture's signature on tomograms was 1 meter below the stimulation zone, which was unexpected and could result from either rock heterogeneity due to folding or foliation that could affect fracture propagation or to geometric errors during surveying. Tomograms were correlated with interval pressure data from the stimulation zone and displayed some notable trends. The trends were analyzed by plotting travel-time variations with interval pressure data and two fracturing-related processes were inferred. The first was a transition from fracture growth (propagation) to stabilization (open and closing) and the second was possible permanent deformation.