Influence of Natural-Fracture Connectivity on Hydraulic-Fracture Propagation in Shale Reservoirs

Abstract

Natural-fracture connectivity substantially influences hydraulic-fracture interaction with pre-existing discontinuities, but its quantitative role in fracture-network propagation remains insufficiently constrained. In this study, a coupled LEFM–cohesive-zone hydraulic-fracture propagation model was developed by combining crack-tip deflection criteria, traction-separation damage evolution and fluid–solid coupling. True triaxial hydraulic-fracturing experiments were conducted on artificial fracture networks with I-, V-, Y- and X-shaped connectivity elements to evaluate the model response. The results show that connected natural fractures redirect hydraulic fractures under low horizontal stress differences, producing deflection angles of 30–50 degrees. When the stress difference exceeds 4 MPa, fracture growth becomes more strongly aligned with the maximum principal stress direction. In the true triaxial tests, the total number of connected natural fractures increased from 14 in the I-shaped network to 17 and 21 in the Y- and X-shaped networks, corresponding to increases of 21.4% and 50.0%, respectively. X-shaped networks showed the strongest sensitivity to stress difference and injection rate, while higher elastic modulus reduced fracture width and promoted longer, narrower fractures. Scale-normalized comparisons based on image-derived experimental measurements showed that the predicted propagation length, fracture width and connected-fracture number followed the experimental trend from I-shaped to Y-shaped and X-shaped networks, with relative errors within 7.1% and a mean absolute percentage error of 4.8%. These findings suggest that fracture topology strongly influences pressure transmission and multidirectional activation in the tested models, whereas field-scale extrapolation requires three-dimensional validation and transport analysis.