Item

Abstract

Induced seismicity has been a major challenge for engineering subsurface reservoirs. Particularly, hydraulic stimulation of Enhanced Geothermal Systems (EGS) is typically accompanied by seismic events that may be felt by public and even include damaging earthquakes. Massive fluid injection reactivates the pre-existing critically stressed faults and releases the strain energy by seismic slip. In densely fractured rocks, induced seismic sequence is jointly controlled by the fracture network geometry and the activated coupled hydromechanical processes. However, the physical linkage between geometrical attributes of fracture network spatiotemporal and magnitude distribution of induced seismicity is not fully understood. A physical understanding may define constraints over the maximum induced earthquake magnitude, which is a crucial parameter for seismic risk and hazard assessment.Here, we used stochastically generated discrete fracture networks (DFN) that represent the natural fractures as (poly)lines in a two-dimensional cross section. Various DFNs were generated over two orders of magnitude in length scales (1-100 m) to cover the observed length distribution in outcrop analogues. Then, we utilized a hydromechanical model to resolve the activated processes by fluid injection. The occurrence of seismic and aseismic slip was obtained by post-processing of the resulting displacement field. The preliminary investigations revealed the significant role of connectivity on the spatial distribution of seismic events. However, a sensitivity study is required to clarify the impact of unknown fracture network attributes (length and spatial clustering) on the characteristics of seismic sequence (b-value). The results of the analysis might have significant implications injection-related activities such as enhanced geothermal systems.