Finite element simulation and stimulated reservoir volume optimization of hydraulic fracture propagation in heterogeneous formations based on flow-stress-damage coupling

Microseismic analysis reveals that fracturing in multicluster horizontal wells can generate complex hydraulic fracture networks in heterogeneous formations. Nevertheless, precisely characterizing the three-dimensional (3D) configuration of hydraulic fracture networks and stimulated reservoir volume (SRV) remains a challenge. A coupled flow-stress-damage model based on the finite element method was developed to simulate the 3D nonplanar propagation of hydraulic fractures, integrating rock mechanical heterogeneity, and natural fracture distributions. Conceptually, the damaged element was represented as a rock element containing small cracks, and the aperture of the hydraulic fracture was determined by the aperture of the crack. By monitoring microseismic activity and the cumulative volume of damaged elements, SRV during hydraulic fracturing was simulated. Laboratory-scale simulation demonstrated the model's feasibility in replicating the stress shadow and multiple hydraulic fracture interference process. Additionally, field-scale simulations revealed the mechanism of multicluster fracturing in naturally fractured formations and proposed a novel optimization method based on maximizing SRV and effective proppant addition for horizontal well cluster spacing. Results indicate that multiple hydraulic fractures can induce transverse fractures, enhancing hydraulic fracture complexity and SRV. The optimal conditions for creating complex hydraulic fracture networks include high injection rates, large fluid volumes, low fluid viscosity, and minimum horizontal stress differences. A cluster spacing of 7 m is most conducive to achieving a complex hydraulic fracture network with maximum SRV and appropriate crack apertures for proppant addition. This study provides a reliable tool for hydraulic fracturing simulation and insights into fracturing mechanisms and cluster spacing optimization methods.