Reinforcement learning-based integrated active fault diagnosis and tracking control

Seismic activity near the shale gas fields in the Sichuan Basin of China has risen significantly in the past few years. Due to the lack of industrial stimulation data to the public, it is difficult to directly link the surge in seismicity near the shale gas fields with hydraulic fracturing stimulation. Meanwhile, it is also challenging to better understand the mechanisms inducing/triggering seismicity in these areas. In the absence of stimulation data, understanding subsurface fluid distribution can go a long way toward addressing these problems. Therefore, in this study, we have applied the double-difference seismic tomography method and the focal mechanism tomography method to the monitoring data of a local seismic array in the Changning shale gas field to build high-resolution three-dimensional models of seismic velocity structure and pore pressure field, then used these models to characterize fluid distributions beneath the shale gas field. From the inverted models, the correlation of overpressure regions (>12 MPa) with moderate-to-high Vp/V s (1.75-1.90) is clearly observed. With the aid of the exploration data (including well log, active and passive seismic data), the overpressure regions are interpreted to be created by pore pressure or fluid diffusion along pre-existing faults from the shale gas reservoirs, while variations of the Vp/V s model partly indicate water- and gas-bearing rocks. By jointly analyzing the geophysical model features and earthquake spatio-temporal characteristics, we infer that the earthquakes are primarily caused by increased pore pressures on the seismogenic faults due to the presence of water and/or gas. Besides, the poroelastic stress perturbation and aseismic slip produced by hydraulic fracturing also contribute to the generation of certain earthquake swarms.(c) 2023 Elsevier B.V. All rights reserved.

Automated summary

Seismic activity near shale gas fields in the Sichuan Basin of China has increased significantly, but it is difficult to directly link it with hydraulic fracturing stimulation due to the lack of public industrial stimulation data. Understanding subsurface fluid distribution can help address this issue. This study uses seismic tomography methods to build three-dimensional models and characterizes fluid distributions beneath the shale gas field.