| Title | Fault Slip and Fracture Growth Revealed Induced Seismicity During a Decameter-scale Hydraulic Stimulation Experiment |
|---|---|
| Authors | Linus VILLIGER, Valentin GISCHIG, Joseph DOETSCH, Hannes KRIETSCH, Mohammadreza JALALI, Florian AMANN, Stefan WIEMER |
| Year | 2020 |
| Conference | World Geothermal Congress |
| Keywords | engineered geothermal systems, seismo-hydro-mechanical processes, hydraulic stimulation mechanism, splay fracture, break in frequency magnitude scaling |
| Abstract | In early 2017, 12 hydraulic stimulation experiments were performed in a 20 x 20 x 20 m foliated, crystalline rock volume intersected by two distinct fault sets at the Grimsel Test Site, Switzerland. The goal of these experiments was to support research and development related to enhanced geothermal systems. The first series of stimulations was intended to permanently enhance hydraulic transmissivity associated with shear dislocation along existing structures (i.e. so-called hydroshearing experiments). Four of these stimulations targeted ductile shear zones (termed S1), which are characterized by a more distinct foliation compared to the host rock and brittle fractures of various orientations. Two stimulations were conducted on two brittle-ductile faults (termed S3.1 and S3.2) each consisting of a fault zone associated with biotite-rich metabasic dykes up to 1 m thick. The lateral distance between these two S3 faults was about 2.5 m and the rock mass between the faults was heavily fractured (Krietsch et al., 2018a). To gain insight on how the variable geology affects the seismo-hydro-mechanical response, the injection protocol was similar for all stimulation experiments. The far field stress state determined in a previous stress characterization campaign (along with an extensive geological characterization) suggested a higher tendency for shear dislocation for injections in the ductile shear zone S1 (Krietsch et al., 2018b). The second series of six stimulations were performed in the same experimental volume in borehole intervals that did not contain any natural fractures using higher injection rates and injection pressures with the intention to initiate hydraulic fractures, which connect the wellbore to the existing fracture network. Multiple types of injection fluids were used to study their propensity to fracture: water and higher viscous fluids (with a dynamic viscosity about 30 times higher than water) were studied. To measure seismicity during the various injections a comprehensive array of 26 highly sensitive acoustic emission sensors (AE, bandwidth: 1 – 100 kHz, highest sensitivity at 70 kHz) were installed. The core of this sensor network consisted of eight AEs diploid in four monitoring boreholes, which allow to reduce the distance between sensors and the injection interval in the range of 5 to 25 m. Results of the first experiment series - the hydroshearing experiments - show higher increases in hydraulic transmissivity for injections targeting S1 shear zones compared to injections into S3 fault zones. This is expected given the higher slip tendency on S1 structures. However, the number of detected and located seismic events is lower compared to stimulations targeting S3 fault zones. The b-values (referring to the size distribution in frequency-magnitude distribution) were lower for stimulations in shear zones S1. In all the stimulations, clustering of seismic events can be observed, highlighting preexisting and new features inside the shear zones. Some of the seismic clouds also indicate new fractures that splay off from the shear zones. The high variability in seismic response from these first series of experiments is related to the different geological characteristics of the fault zones. New fractures induced in intact rock during the second series of experiments form perpendicular to the minimum principal stress and partially connect to the existing fracture network. Interpretations and conclusions of this unique induced seismicity dataset of 12 controlled injection experiments are presented in this contribution. |