Stanford Geothermal WorkshopFebruary 9-11, 2026

Multi-stage hydraulic fracturing is widely used in Enhanced Geothermal Systems (EGS) to create extensive fracture networks, which serve as flow pathways between injector and producer wells. Optimizing well placement requires reliable constraints on fracture geometry, such as height, length, and connectivity, which in EGS remains poorly understood due to limited direct observations. Cross-well distributed fiber-optic strain measurements in offset monitoring wells have proven highly effective for inferring fracture geometry and have reshaped understanding in unconventional reservoirs. In this study, we interpret cross-well fiber data acquired at the FORGE (Frontier Observatory for Research in Geothermal Energy) EGS site and integrate the results with microseismic locations and imaging results to characterize the stimulated fracture geometry. We analyzed cross-well fiber strain data to identify fracture hits in the offset monitoring well. Fracture hits were detected using a multi-attribute approach combining strain rate, strain gradient, and displacement signals. An in-house forward geomechanics model has been used to simulate strain responses under varying conditions. This enables us to interpret the abnormal strain signatures observed in EGS that differ from the typical responses in unconventional reservoirs. We refined our interpretations with microseismic analysis and natural fracture characterization. We have detected 21 fracture hits, which clustered within three depth intervals in the monitoring well—8,771–8,996 ft, 9,432–9,502 ft, and 9,742–9,787 ft—with inferred fracture spacing of ~15–35 ft. Many fractures reopened two to three times during late-stage pumping. On average, newly formed fractures reached the monitoring well after 132 minutes, whereas reopening occurred within 64 minutes. We generated a cross-section view by connecting perforation and fracture-hit location and calculated fracture dip angle, assuming the strike direction coincides with the maximum horizontal stress orientation of 78° relative to the wellbore toe side. The dip angles range from 58° - 102°, with only two fractures dipping toward the toe side. Forward geomechanical modeling using a fracture strike of 78° and dip of 60° reproduces the prominent asymmetric, heart-shaped strain signature observed in Stage 8. The consistency between the local and global dip angles indicates that the hydraulic fracture likely dips 59° from horizontal in stage 8, which is consistent with the conclusion drawn from the in-situ stress analysis. In the cross-sectional view, the inferred fracture propagation direction aligns well with the microseismic event distribution in Stages 3 through 7. However, in Stage 8, the fracture hits occurred first, followed by microseismic activity along a different trajectory. This observation indicates that the hydraulic fracture intersected the 16B well and reactivated a pre-existing natural fracture. The integrated analysis of Stages 8-10 supports the hypothesis that the local Sᵥ is not the principal stress, which explains the hydraulic fracture dipping 59° observed in Stage 8. Our results advance understanding of fracture propagation and geometry in EGS and highlight the value of distributed fiber-optic sensing (DFOS) for geothermal fracture diagnostics. The findings provide key inputs for optimizing injector–producer placement, including both vertical and horizontal well spacing.

Topic: Enhanced Geothermal Systems