The growing installation of industrial facilities for subsurface exploration worldwide requires continuous refinements in understanding both the mechanisms by which seismicity is induced by field operations and the related seismic hazard. Particularly in proximity of densely populated areas, induced low‐to‐moderate magnitude seismicity characterized by high‐frequency content can be clearly felt by the surrounding inhabitants and, in some cases, may produce damage. In this respect we propose a technique for time‐dependent probabilistic seismic‐hazard analysis to be used in geothermal fields as a monitoring tool for the effects of on‐going field operations. The technique integrates the observed features of the seismicity induced by fluid injection and extraction with a local ground‐motion prediction equation. The result of the analysis is the time‐evolving probability of exceedance of peak ground acceleration (PGA), which can be compared with selected critical values to manage field operations.
To evaluate the reliability of the proposed technique, we applied it to data collected in The Geysers geothermal field in northern California between 1 September 2007 and 15 November 2010. We show that the period considered the seismic hazard at The Geysers was variable in time and space, which is a consequence of the field operations and the variation of both seismicity rate and b‐value. We conclude that, for the exposure period taken into account (i.e., two months), as a conservative limit, PGA values corresponding to the lowest probability of exceedance (e.g., 30%) must not be exceeded to ensure safe field operations. We suggest testing the proposed technique at other geothermal areas or in regions where seismicity is induced, for example, by hydrocarbon exploitation or carbon dioxide storage.