Induced seismicity associated with the geothermal power plant at Insheim (Germany) has been continuously analyzed using a combination of permanent and temporary stations since 2012 over a time span of more than 4 yrs. More than 600 induced events have been detected and located. The largest event had a magnitude of ML≈2.4⁠. Accurate locations are crucial for understanding the behavior of the reservoir. An initial velocity model derived from vertical seismic profiling (VSP) turned out not to be appropriate for the location of induced events. Location uncertainty was reduced by probabilistic earthquake locations using an optimized minimum 1D velocity model. Resolution was further increased by relative relocation based on differential travel times obtained from absolute P‐ and S‐phase onset readings, as well as from waveform cross correlations. The initial absolute location errors decreased from 500 m horizontally and ≈1000  m vertically to ≈300 and ≈500  m⁠, respectively. Clusters of similar events were detected by waveform cross correlation. The average difference in the relative locations between highly similar events amounts to about 50 m, yielding independent support for the relative location error estimate. Most of the events occur along a fault zone in the basement connecting the injection and the production wells. Interestingly, one cluster including the largest event occurred close to the production well. In addition, locations of induced events hint at the activation of formerly unknown conjugated fault zones. Derived fault‐plane solutions of selected induced events confirm this interpretation and agree well with the regional stress regime. The seismically activated volume increased within 1 yr to its final volume of ≈1.65×109  m3 and remained more or less constant. Seismic activity correlates strongly with injection rates and increases in case of unexpected shut‐ins.

You do not have access to this content, please speak to your institutional administrator if you feel you should have access.