| Abstract |
Microseismic monitoring plays a pivotal role in characterizing subsurface dynamics and assessing induced seismicity risks in Enhanced Geothermal Systems (EGS). This study presents a comprehensive analysis of microseismic data from the Cape Modern EGS field in Southwest Utah, where three horizontal wells underwent plug-and-perf hydraulic stimulation. The monitoring infrastructure comprised an integrated network of shallow borehole sensors, surface nodal arrays, deep borehole fiber optic sensors, and three-component passive sensors, capturing over 7,000 events during the February-March 2024 stimulation period. We implemented a multi-faceted analytical approach, beginning with phase arrival prediction and first motion polarity determination utilized deep learning techniques, while focal mechanism estimation employed the SKHASH method, incorporating P-wave polarity and S/P amplitude ratios. This methodology yielded 2,091 focal mechanism solutions, with 1,564 achieving high-quality (A and B) classifications. Subsequent clustering analysis using UMAP-HDBSCAN revealed four distinct structural features, with Structure D exhibiting complex spatial patterns and multiple sub-clusters. Stress field analysis using MSATSI demonstrated significant spatial heterogeneity, transitioning from strike-slip dominated regimes in the western section (SHmax 15-30° from North) to reverse faulting patterns in the eastern section (SHmax 0-15° from North). The stress ratio variations indicate complex mechanical interactions between stimulation operations and local geological structures. These findings provide crucial insights for optimizing EGS operations while highlighting the effectiveness of integrating machine learning techniques with traditional geophysical methodologies for enhanced reservoir characterization. |