The Role of Background Stress State in Fluid‐Induced Aseismic Slip and Dynamic Rupture on a 3‐m Laboratory Fault

Abstract

The injection of fluids into the Earth-be it for CO 2 sequestration, enhanced geothermal systems, or oil and gas operations-is known to induce earthquakes (Ellsworth, 2013;Keranen et al., 2013;Raleigh et al., 1976). Minimizing induced seismicity requires an understanding of what causes a fault to begin to slip, the mechanisms driving the transition from aseismic to seismic slip (i.e., initiation of dynamic rupture), and how large the resulting seismic event will grow (i.e., how far dynamic rupture is sustained). These factors help inform the maximum event magnitude and potential for runaway ruptures. This study explores how background stress levels affect the initiation and termination of fluid-induced ruptures using a 3 m rock experiment.Fluid injection field experiments on the decameter scale highlight the important role of induced aseismic slip in the initiation of induced seismicity. Results from show that fluid injection primarily induced aseismic slip. They observed microseismicity as a by-product of aseismic slip rather than directly

Creep fronts and complexity in laboratory earthquake sequences illuminate delayed earthquake triggering

“…The simulations show that small changes in τ 0 (50 kPa or <1%) cause two orders of magnitude variation in the average propagation speed of the creep fronts. Fronts propagate faster at higher τ 0 , consistent with other recent studies 10 – 12 , 58 , 59 . Furthermore, as the fault fabric develops, and a-b and D c decrease and R u increases 46 , creep front behavior and the isolated events at A1 and A2 become increasingly sensitive to small variations in stress levels (Fig.…”

Single‐Well Pore Pressure Preconditioning for Enhanced Geothermal System Stimulation

“…As the shear stress is assumed to be unaffected by the change in pore pressure, an increase in effective normal stress means that the residual shear stress is increasing, reducing the stress drop, and therefore energy, available to the rupturing earthquake, an idea consistent with experiments (Cebry et al, 2022). Dynamic crack propagation being largely sensitive to stress drop (Ampuero et al, 2006;Bayart et al, 2018;Cebry et al, 2022;Garagash & Germanovich, 2012;Ke et al, 2018;Viesca & Rice, 2012), this effect has the ability to halt these ruptures by reducing the amount of energy flowing to the crack tip. Indeed, it has previously been shown numerically that fault sections that were once prone to large seismic ruptures may tend to later halt rupture propagation after a change in stress (Lapusta et al, 2000).…”

“…This criterion is equivalent to saying that the stress criticality must become less than the residual friction, $$\tilde{\tau }< {\tilde{f}}_{\mathrm{r}}$$, as found by Garagash and Germanovich (2012) and shown experimentally by Cebry et al. (2022). In this case, the crack will still be able to propagate due to the increase of pore pressure near the wellbore, but, like for the criterion derived by Garagash and Germanovich (2012), the crack will be ultimately stable as long as it remains in the preconditioned zone.…”

“…McLaskey, 2018) faults as well as predicted theoretically (Ampuero et al, 2006;Husseini et al, 1975). This arrest occurs because propagation of a seismic rupture depends on the local fracture energy, itself a function of the stress along the fault plane, and the global stress drop profile along the fault which provides energy to the propagating rupture tip (Bayart et al, 2016(Bayart et al, , 2018Cebry et al, 2022;Freund, 1998;Galis et al, 2017;Gvirtzman & Fineberg, 2021;Husseini et al, 1975;Kammer et al, 2015;Paglialunga et al, 2022).…”

“…Due to the simple nature of the model used here, the center of any nucleated dynamic event is at the wellbore. Note, however, that while induced-seismic events do not necessarily nucleate at the wellbore, and have even been seen to propagate back toward the wellbore (Folesky et al, 2016), laboratory studies have indicated that dynamic rupture seems to initiate within an aseismically slipping patch caused by pore pressure increase (e.g., Cebry et al (2022)). As sufficient pore-pressure preconditioning causes the aseismically slipping front to remain behind the pore-pressure barrier, a dynamic event would have to nucleate behind this barrier.…”