Characterizing a decametre-scale granitic reservoir using GPR and seismic methods – A case study for preparing hydraulic stimulations

Abstract

Abstract. Ground-penetrating radar (GPR) and seismic imaging have proven to be important tools for the characterization of rock volumes. Both methods provide information about the physical rock mass properties and geology structures away from boreholes or tunnel walls. Here, we present the results from a geophysical characterization campaign that was conducted in preparation for a decametre-scale hydraulic stimulation experiment in the crystalline rock volume at the Grimsel Test Site (Central Switzerland). For this characterization experiment, we used tunnel based GPR reflection imaging as well as seismic traveltime tomography to investigate the volumes between several tunnels and boreholes. The interpretation of the GPR data with respect to geological structures is based on the unmigrated and migrated images. For the tomographic analysis of the seismic first-arrival traveltime data, we inverted for an anisotropic velocity model described by the Thomsen parameters v0, ϵ and δ to account for the rock mass foliation. Subsequently, the GPR and seismic images were interpreted in combination with the geological model of the test volume and the known in-situ stress states. We found that the ductile shear zones are clearly imaged by GPR and show an increase in seismic anisotropy due to a stronger foliation, while they are not visible in the P-wave (v0) velocity model. Regions of decreased seismic p-wave velocity, however, correlate with regions of high fracture density. For geophysical characterization of potential deep geothermal reservoirs, our results imply that wireline compatible borehole GPR should be considered for shear zone characterization, and that seismic anisotropy and velocity information are desirable to acquire in order to gain information about ductile shear zones and fracture density, respectively.

Estimates of Individual Fracture Compliances Along Boreholes From Full‐Waveform Sonic Log Data

“…Here, the lack of correlation between high fracture density and low velocity zones in the 3D seismic volume obtained by Doetsch et al. (2020), could be explained by the low compliance exhibited by the fractures close to the S1.3 shear zone.…”

“…Doetsch et al. (2020) also pointed out that the elevated fracture density at ∼43 m depth in INJ1 does not manifest itself with a clear decrease in seismic velocity. Here, the lack of correlation between high fracture density and low velocity zones in the 3D seismic volume obtained by Doetsch et al.…”

“…The two boreholes considered here were dedicated to stimulation‐type injections and are referred to as INJ1 and INJ2 (Figure 6). They penetrate moderately fractured crystalline rocks dissected by major shear zones, which are classified as being of S1‐ and S3‐type in Figure 6 (Brixel et al., 2020; Doetsch et al., 2020; Krietsch et al., 2018; Wenning et al., 2018). The S3‐type shear zones are associated with lamprophyre dikes.…”

“…This further suggests that compliance estimates from FWS data can provide valuable insight prior to hydraulic stimulation experiments. Doetsch et al (2020) also pointed out that the elevated fracture density at ∼43 m depth in INJ1 does not manifest itself with a clear decrease in seismic velocity. Here, the lack of correlation between high fracture density and low velocity zones in the 3D seismic volume obtained by Doetsch et al (2020), could be explained by the low compliance exhibited by the fractures close to the S1.3 shear zone.…”

“…Our compliance estimates are consistent with this conceptual model in both boreholes. Doetsch et al (2020) provided a seismic characterization of the same rock volume prior to hydraulic stimulations and found that the S3-type shear zones are characterized by reduced seismic wave velocities with respect to the host rock. They argue that a higher fracture density could be the controlling factor associated with the velocity decrease.…”