Potential of ambient noise techniques to monitor reservoir dynamics at the St. Gallen geothermal site

In the last decade two large geothermal energy projects were launched in Switzerland (Basel 2006, St Gallen 2013). Both of them were stopped after the occurrence of strongly felt earthquakes (Ml3.4 and 3.5, respectively). This illustrates that one of the key challenges for the use of deep geothermal energy remains to control the risk of inducing felt and potentially hazardous seismic events during the development and operation of an underground heat exchangers.

Current monitoring techniques of induced seismicity, e.g. traffic light systems, attempt to forecast seismic hazard during and after stimulation based on observed seismicity and hydraulic data. A limitation of these techniques is their focus on seismic processes.

We demonstrate the potential of ambient seismic noise correlation techniques to monitor aseismic reservoir dynamics related to the 2013 geothermal project in St. Gallen. In St. Gallen, reservoir characterization tests lead to an unexpected leakage of methane gas into the well. Well-head pressure rose rapidly and operators decided to prevent a possible well blow-out with counter-pressure. The result was an immediate increase of induced seismicity with a maximum event of Ml3.5. While the reservoir characterization was not accompanied by any significant induced seismicity that could have given an indication for the ongoing processes in the reservoir, ambient noise cross-correlations reveal a significant aseismic perturbation in the system that can be clearly linked to the stimulation tests. These additional constraints may help to better understand reservoir dynamics. We also discuss the future role of noise correlation based techniques for monitoring/mitigation purposes.