Simulation and optimization of unstable dynamic propagation of multiple fractures in the shale formation

Abstract

Multi-cluster hydraulic fracturing of horizontal wells is a well-adopted technique with high efficiency to increase the production of tight and shale formations. However, the stress shadows among clusters pose challenges to the synchronous propagation of hydraulic fractures during multi-cluster fracturing. In order to explore the fracture propagation mechanism and characteristics under the influence of stress shadow in low-spacing staged multi-cluster fracturing, a three-dimensional hydraulic fracturing model was generated using a lattice-based method. This model considered the impact of geological and engineering parameters on the propagation behavior of multiple fractures in shale formation. A variable pumping approach is adopted, where the fracturing fluid is initially injected at a high rate and then transitioned to a lower rate. Afterward, a method was proposed to quantitatively assess the extent of fracturing in a specific area (i.e., the stimulated area), considering the impact of stress shadow within a single stage. The simulation results demonstrated significant differences in the fracture stimulation area due to the influence of each parameter in the case of uncontrollable geological factors and controllable engineering factors. An increase in both Young’s modulus and stress anisotropy of the reservoir results led to a corresponding increase in the total fracture stimulation area. As the principal stress orientation increased, the fracture stimulation area gradually decreased. In terms of operational parameters, the stimulated area of hydraulic fractures gradually decreased as the fracture spacing increased. With increasing injection rate, the stimulated area initially expanded and then decreased, peaking at an injection rate of 0.04 m3/s. These findings can provide valuable insights into the propagation behavior of multi-cluster hydraulic fractures under uncertain parameters, with significant implications for future engineering applications.