Induced Seismicity Driven by Fluid Diffusion Revealed by a Near‐Field Hydraulic Stimulation Monitoring Array in the Montney Basin, British Columbia

Abstract

This study presents observations using new data from a deployment of eight broadband seismometers surrounding a horizontal well pad at distances of~1-3 km for the period before, during, and after a hydraulic fracturing treatment in the Montney Basin, British Columbia, Canada. We use a multistation-matched filter detection and double-difference earthquake relocation to develop a catalog of 350 events associated with hydraulic fracturing stimulation, with magnitudes ranging from −2.8 to 1.8 and estimated catalog completeness of approximately −0.2. The seismicity distribution suggests a statistically significant association with injection, and event migration can be described by a hydraulic diffusivity of~0.2 m 2 /s. A comparison between daily seismicity rate and analytical stress evolution inferred from daily injection volumes implies that pore pressure diffusion largely controls earthquake nucleation at distances less than 1 km, whereas poroelastic stress transfer likely dominates at intermediate distances of~1-4 km at time scales shorter than diffusion. Both mechanisms likely have a limited effect on stress perturbation at distances over 5 km.

Fluid-injection induced earthquakes characterized by hybrid-frequency waveforms manifest the transition from aseismic to seismic slip

“…As shown in Fig. 2B, the migration of EHWs mainly follows a fluid diffusivity value of ~0.2 m 2 /s (see in Methods), which is consistent with the estimation based on ordinary induced events (13). A few EHWs located near the station MG08 are compatible with a higher diffusivity value of ~1.5 m 2 /s, suggesting the existence of a fluid conduit that connects to the well.…”

“…We propose that EHWs could be related to the injection targeting the lower Middle Montney formation, which only occurs during daytime hours to limit the number of induced earthquakes felt by local residents (see Text S2). The planar structure implied by the southern EHW cluster also intersects a group of ordinary induced events (13) (Fig. S2) that do not show any comparable waveform similarity or similar temporal pattern.…”

“…Previous studies suggest many M3+ events are generated on pre-existing seismogenic faults near HF wellbores that are critically stressed prior to injection, where injection facilitates rupture by shifting the stress state towards failure through pore pressure increase (9,10) or poroelastic stress transfer (11,12). However, the combination of factors that are associated with HF injection, such as the relatively small total injected fluid volumes, the short injection time history, and the often nearly instantaneous onset of seismicity at distances of several kilometers from wellbores, makes it challenging to explain the occurrence of M3+ earthquakes with the classical concept of pore pressure and/or poroelastic stress change (13,14).…”

Hydraulic Fracturing Induced Seismicity in the Southern Sichuan Basin Due to Fluid Diffusion Inferred From Seismic and Injection Data Analysis

“…Two triggering mechanisms have been widely adopted to explain the occurrence of earthquakes induced by hydraulic fracturing. First, the preexisting faults are directly connected to the hydraulic fractures and increased pore pressure on the faults due to fluid diffusion triggers the faults to slip (e.g., Bao & Eaton, 2016;Goebel & Brodsky, 2018;Schultz et al, 2017;Yu et al, 2019). Alternatively, the hydraulic fractures The background earthquake data are obtained from the CEDC catalog.…”

“…In general, there are two mechanisms that are applied to explain the induced seismicity due to hydraulic fracturing. The first mechanism is the activation of preexisting faults caused by fluid diffusion (e.g., Bao & Eaton, 2016;Goebel & Brodsky, 2018;Schultz et al, 2017;Yu et al, 2019), and the second one is due to the poroelastic effect of rock matrix (e.g., Bao & Eaton, 2016;Deng et al, 2016;Goebel & Brodsky, 2018;Segall & Lu, 2015). Compared to the previous studies, we used the seismic data to image the velocity structures around the earthquake hypocenters in this study.…”