Shallow Faults Reactivated by Hydraulic Fracturing: The 2019 Weiyuan Earthquake Sequences in Sichuan, China

Abstract

Human activity-induced earthquakes are emerging as a global issue, and revealing its underlying mechanisms is essential for earthquake hazard mitigation and energy development. We investigated the relationship between the seismotectonic model and seismic sequences from moderate Mw 4.3 and Mw 5.2 earthquakes that occurred in February and September 2019, respectively, in the Weiyuan anticline of Sichuan basin, China. We found that the Mw 5.2 earthquake ruptured a back thrust of structural wedges and released most aftershocks near the wedge tip. However, the two foreshocks of the Mw 4.3 earthquake sequence occurred in hydrofractured Silurian shale at depth of 2.5–3 km, and the mainshock ruptured the overlying oblique tear fault at a depth of ∼1 km. Hydraulic fracturing in the sedimentary cover of this block may induce earthquakes through fluid pressure diffusion in the Silurian shale and through poroelastic effects on back thrusts within structural wedges, respectively. We assessed the hazard potential of four seismic sources in the Weiyuan block and suggest it is critical to conduct a coupled flow-geomechanics assessment and management on induced seismicity and related cascading effects in the densely inhabited and seismically active Sichuan basin.

Study on the Impact of Hydraulic Fracturing on Surrounding Ancillary Buildings considering SSI

“…On the basis of these two blocks, demonstration area construction and expansion of mining scope have been carried out. At the same time, the seismic frequency in this area has also increased sharply [3][4][5]. For example [6][7][8][9], the Junlian earthquake with magnitude 4.9, Xingwen Earthquake with magnitude 5.7, Weiyuan earthquake with magnitude 5.4 occurring in Sichuan Basin, and Red Dee earthquake with magnitude 4.2 occurring in Western Canada Basin have been proved to have a time and space correlation with the fracturing operation in the nearby horizontal well.…”

“…Numerical results show that SSI had significant influence on structural response and cannot be ignored. In [12], considering soilstructure interaction, pushover method, natural frequency, vertex limit displacement, and plastic hinge expansion were used to analyze the frame structure of a 6-story office 4 Geofluids building under the assumption of rigid foundation. The results show that considering soil-structure interaction, the natural vibration period of RC frame structure became longer, and more hinges appeared at the bottom of column with the increase of vertex displacement.…”

Machine-learning-facilitated earthquake and anthropogenic source detections near the Weiyuan Shale Gas Blocks, Sichuan, China

“…PhaseNet (Zhu WQ and Beroza, 2019) is a deep‐neural‐network‐based arrival‐time picking method. Although it was trained on northern California earthquakes, it has since demonstrated success in picking accurate arrival times of both P and S waves in diverse regions and signal environments, which has made it an attractive choice for routine earthquake detection (Zhang M et al, 2019; Liu M et al, 2020; Park et al, 2020; Wang MM et al, 2020; Tan YJ et al, 2021).…”

“…In this study, we investigate earthquakes that occurred in the region surrounding the Weiyuan Shale Gas Blocks (SGB), Sichuan, China, between November 2015 and November 2016, using a temporary network of 50 stations (Figure 1). The deployment occurred during a time when small‐magnitude earthquakes were increasing in number, but three years before the deadly M w 4−5 class earthquakes that occurred in 2019 (Sheng MH et al, 2020; Wang MM et al, 2020; Yang HF et al, 2020). Here we demonstrate the feasibility of rapidly generating a catalog using a dense network and machine‐learning algorithms.…”

Spatial-temporal Evolution of Injection Induced Earthquakes in Weiyuan Area by Machine-Learning Phase Picker and Waveform Cross-correlation

“…According to the analysis, the magnitude five earthquakes that occurred in early September show depth variation from two to five kilometers (Wang M. et al, 2020;Lei & Wang, 2020;Sheng et al, 2020). Although location disputes and differences are common in earthquake studies as geophysical inversion with different observations and model parameters may have divergent results, the hypocenter locations in induced earthquake study are pivotal in determining the corresponding inducing mechanism.…”