Mechanistic modeling of crack propagation in hydraulically fractured reservoirs for predicting inter-well fracture communication during infill well stimulation

Abstract

Inter well fracture communication is a persistent challenge in multi well infill stimulation, often reducing production efficiency and compromising reservoir integrity. This study develops a mechanistic framework based on the Extended Finite Element Method (XFEM) with phantom node enrichment to simulate multi fracture propagation between neighboring horizontal wells. The model couples poroelastic rock deformation, fracture-matrix fluid exchange, pressure dependent leak off, and fracture propagation governed by a traction-separation law, providing a fully integrated representation of hydraulic fracturing processes. Parametric analyses reveal that zero-stagger distance with simultaneous injection promotes complete fracture linking, while larger offsets or scheduled treatments mitigate communication through stress shadow effects. Increasing rock tensile strength enhances fracture repulsion and reduces tip to tip linking. Distinct pressure signatures differentiate linking fractures, which exhibit localized sagging, from non linking fractures with monotonic gradients. The framework was validated against Displacement Discontinuity Method (DDM) benchmarks, Kristianovich-Geertsma-de Klerk (KGD) analytical solutions, and field measured pressure data from the Daqing Oilfield, demonstrating strong goodness of fit and confirming model fidelity. This work finds useful application in designing perforation patterns, optimizing cluster spacing, and scheduling treatments in unconventional shale reservoirs. By enabling accurate prediction of fracture linking, coalescence, and repulsion, the framework provides practical guidance for maximizing stimulated reservoir volume while controlling unintended inter well interference.