Deep geothermal energy (DGE) represents an opportunity for a sustainable and carbon-free energy supply. One of the main concerns of DGE is induced seismicity that may produce damaging earthquakes, challenging its widespread exploitation. It is widely believed that the seismicity risk can be controlled by using doublet systems circulating water to minimize the injection-induced pressure changes. However, cold water reinjection may also give rise to thermal stresses within and beyond the cooled region, whose potential impacts on fault reactivation are less well understood. Here, we investigate by coupled thermo-hydro-mechanical modeling the processes that may lead to fault reactivation in a hot sedimentary aquifer (HSA) in which water is circulated through a doublet. We show that thermal stresses are transmitted much ahead of the cooled region and are likely to destabilize faults located far away from the doublet. Meanwhile, the fault permeability mainly controls the fault reactivation timing, which entails the importance of employing appropriate characterization methods. This investigation is crucial for understanding the mechanisms controlling induced seismicity associated with DGE in a HSA and allows the success of future DGE projects.

Fluid injection into geological formations for energy resource development frequently induces (micro)seismicity. Moderate- to large-magnitude induced earthquakes may cause injuries and/or economic loss, with the consequence of jeopardizing the operation and future development of these geo-energy projects. To achieve an improved understanding of the mechanisms of induced seismicity, develop forecasting tools and manage the associated risks, it is necessary to carefully examine seismic data from reported cases of induced seismicity and the parameters controlling them. However, these data are challenging to gather together and are time-consuming to collate as they come from different disciplines and sources. Here, we present a publicly available, multi-physical database of injection-induced seismicity (Kivi et al., 2022a; https://doi.org/10.20350/digitalCSIC/14813), sourced from an extensive review of published documents. Currently, it contains 158 datasets of induced seismicity caused by various subsurface energy-related applications worldwide. Each dataset covers a wide range of variables, delineating general site information, host rock properties, in situ geologic and tectonic conditions, fault characteristics, conducted field operations, and recorded seismic activities. We publish the database in flat-file formats (i.e., .xls and .csv tables) to facilitate its dissemination and utilization by geoscientists while keeping it directly readable by computer codes for convenient data manipulation. The multi-disciplinary content of this database adds unique value to databases focusing only on seismicity data. In particular, the collected data aim at facilitating the understanding of the spatiotemporal occurrence of induced earthquakes, the diagnosis of potential triggering mechanisms, and the development of scaling relations of maximum possible earthquake magnitudes and operational parameters. The database will boost research in seismic hazard forecasting and mitigation, paving the way for increasing contributions of geo-energy resources to meeting net-zero carbon emissions.