High-Precision Seismic Prediction Technology for Micro-Faults to Support Shale Gas Horizontal Well Drilling

Three-dimensional high-precision fault prediction provides critical support for shale gas horizontal drilling operations. This paper addresses the challenges of traditional micro-fault prediction methods, including insufficient prediction accuracy and significant result ambiguity, by proposing a high-precision prediction technology based on amplifying seismic fault responses and multi-dimensional data fusion. This approach enhances the accuracy of micro-fault prediction in oil and gas reservoir evaluation, providing more reliable evidence for exploration and development. In this study, we first applied TTI anisotropic depth migration technology and Gabor wavelet transform-based wavenumber spectrum analysis to obtain high-quality seismic depth domain data, extracting comprehensive seismic information and highlighting micro-fault characteristics. Building on this foundation, we integrated convolutional neural network-based intelligent fault prediction technology to achieve complementary feature information, establishing a multi-scale seismic attribute fusion framework. Finally, we validated the approach through a case study in the southern Sichuan shale gas block. Results demonstrate that this method effectively enhances the identifiability of weak fault signals and reduces uncertainties in fault prediction outcomes. Within a spatial range of 60 m, the identification rate reaches 100% for faults with displacements of 20–40 m, 75% for small faults with displacements of 10–20 m, and 71% for micro-faults with displacements less than 10 m. These findings indicate that the technology significantly improves micro-fault prediction in complex geological environments, providing a scientific basis for oil and gas reservoir evaluation.