Dual-parameter Regression Models for Assessing the Risk of Induced Seismicity in Enhanced Geothermal Systems (EGS)

With the global transition to sustainable energy, Enhanced Geothermal Systems (EGS) have garnered increasing attention as a promising source of clean and renewable energy. EGS utilizes hydraulic fracturing to enhance reservoir permeability by initiating and expanding fractures, thereby enabling efficient exploitation of hot dry rock resources. However, hydraulic fracturing is often accompanied by microseismic events or small-scale earthquakes, posing challenges to project safety and economic feasibility. This problem has become a focal point of recent research. Significant case studies have been conducted in regions such as Pohang, South Korea; the Cooper Basin, Australia; Basel, Switzerland; Insheim, Germany; Fenton Hill, USA; as well as in China's geothermal fields, including Qabuqa in Qinghai Province, Huashadong in South China, and Matouying in North China. Despite variations in injection strategies and operational conditions, all these projects have encountered risks of induced seismicity. The mechanism of these seismic events varies depending on the lithology and tectonic setting of the region and deserves further exploration.

Case studies reveal that larger-magnitude-induced earthquakes are predominantly associated with granitic reservoirs and commonly linked to well-developed strike-slip faults. These events are primarily triggered by the activation of pre-existing faults due to increased pore pressure from fluid injection. Specifically, elevated injection pressure reduces fault friction, leading to slip instability. These findings provide valuable insights for the prediction and mitigation of induced seismicity during artificial fracturing. However, sedimentary reservoirs exhibit lower rates and magnitudes of induced seismicity, though their triggering mechanisms warrant further investigation.

In this study, regression analysis was employed to identify key factors influencing the maximum magnitude of induced seismic events, and the data are mainly derived from the study areas of typical hot dry rocks around the world. The analysis focused on parameters such as fault length, maximum injection pressure, maximum injection rate, total injected fluid volume, and fracturing depth. The results show that there are significant differences in the influence of different factors on the maximum magnitude under different lithological conditions. Dual-parameter regression models reveal that the combination of fault length and total injected fluid volume shows higher correlation with seismicity in sedimentary reservoirs, whereas the combination of maximum injection pressure and injection rate is more relevant for magmatic rock. A comprehensive analysis of both sedimentary and magmatic reservoirs demonstrates that injection pressure and fault length are the primary parameters controlling the maximum magnitude of induced seismicity. Dual-parameter models exhibit superior predictive capabilities across lithological conditions, offering robust theoretical support for future seismic risk assessments.

Acknowledgment

This study was supported by the Natural Science Foundation of China (42430808).