Integrating well logs, 3D seismic, and earthquake data for comprehensive prediction of 3D in-situ stress orientations: A case study from the Weiyuan area in the Sichuan Basin, China

Determining the orientation of in-situ stresses is crucial for various geoscience and engineering applications. Conventional methods for estimating these stress orientations often depend on focal mechanism solutions (FMSs) derived from earthquake data and formation micro-imager (FMI) data from well logs. However, these techniques can be costly, depth-inaccurate, and may lack spatial coverage. To address this issue, we introduce the use of three-dimensional (3D) seismic data (active sources) as a lateral constraint to approximate the 3D stress orientation field. Recognizing that both stress and fracture patterns are closely related to seismic velocity anisotropy, we derive the orientation of azimuthal anisotropy from multi-azimuth 3D seismic data to compensate for the lack of spatial stress orientation information. We apply our proposed workflow to a case study in the Weiyuan area of the Sichuan Basin, China, a region targeted for shale gas production. By integrating diverse datasets, including 3D seismic, earthquakes, and well logs, we develop a comprehensive 3D model of in-situ stress (orientations and magnitudes). Our results demonstrate that the estimated anisotropy orientations from 3D seismic data are consistent with the direction of maximum horizontal principal stress (SHmax) obtained from FMIs. We analyzed 12 earthquakes (magnitude > 3) recorded between 2016 and 2020 for their FMSs and compressional axis (P-axis) orientations. The derived SHmax direction from our 3D stress model is 110 degrees ES (East-South), which shows excellent agreement with the FMSs (within 3.96 degrees). This close alignment validates the reliability and precision of our integrated method for predicting 3D SHmax orientations. (c) 2024 The Authors. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/ 4.0/).