Analysis and modelling of induced seismicity in fluid-driven settings and on laboratory faults

Journal Title
Journal ISSN
Volume Title
Seismicity, whether natural from tectonic plate motion, induced by industrial activities, or observed on laboratory faults, is complex due to nonlinearities and long-range spatiotemporal correlations in the Earth's crust. Previous studies revealed spatiotemporal clustering and aftershocks, reminiscent of tectonic earthquakes, in fluid-induced seismicity and seismicity on laboratory faults. Fluid-induced seismicity arises from industrial activities like hydraulic fracturing (e.g., a 4.6Mw in Alberta, Canada) and enhanced geothermal systems (e.g., a 5.5Mw in Pohang, South Korea), where high-pressure fluids are injected into the Earth's crust to improve hydrocarbon extraction or geothermal reservoir permeability. This seismicity raises concerns for industry, regulators, and residents, highlighting the challenge of identifying underlying processes and estimating fluid-induced seismic hazards. Further, ongoing debate persists on the presence or absence of magnitude clustering in seismicity, with implications for improved short-term earthquake magnitude forecasting. Therefore, to address these challenges, we investigate the underlying causes of field observations from a modelling perspective and utilize laboratory stick-slip experiments to enhance our understanding of seismicity. We introduce a novel model of fluid-induced seismicity combining viscoelasticity with fluid diffusion and invasion percolation to simulate crustal rheology and stress interactions in porous media. This model accurately elucidates the spatial footprint of fluid-induced seismicity and reproduces aftershocks akin to tectonic loading. Moreover, at high injection rates, recovering aftershock statistical properties requires direct access to the internal stress dynamic, potentially explaining the absence of reported aftershock triggering in some studies (e.g., Soultz geothermal sites, France). In a reported case of significant aftershock triggering in the Kiskatinaw area, BC, Canada, we find that smaller magnitude, more frequent triggers primarily drive this triggering, consistent with past instances in fluid-induced settings. Additionally, we observe localized spatiotemporal clustering, characterized by a rapid spatial decay in aftershocks beyond the rupture length, suggesting the dominant influence of fluid migration in this region. Finally, our analysis of seismicity on laboratory faults contributes to the debate over magnitude clustering in seismicity. Our study reveals magnitude clustering occurring exclusively during slip events on laboratory faults, resulting from variations in frequency-magnitude distributions. This thesis offers valuable insights into complex seismicity dynamics and enhances efforts in (fluid-induced) seismic hazard assessments.
Khajehdehi, O. (2024). Analysis and modelling of induced seismicity in fluid-driven settings and on laboratory faults (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from