Quantifying the impact of geological and construction factors on hydraulic fracture dynamics in heterogeneous rock layers using the phase field method

Hydraulic fracturing, a pivotal practice for optimizing development in tight oil reservoirs, encounters challenges in accurately predicting propagation paths within heterogeneous rock formations. Existing phase models often overlook the influence of complex formations, especially in porous media with heterogeneous characteristics. Addressing this gap, we present a novel phase field method aligned with on-site practices. This innovative approach considers cracks, bedding planes, and faults as heterogeneous structures, integrating directional physical and mechanical properties. The key innovation lies in constructing assignment functions to characterize mechanical and seepage parameters of the rock matrix. Anchored in pore elasticity theory and the principle of energy minimization, we establish and validate a hydraulic coupled phase field model through plaster experiments. This model not only corrects the oversight of formation heterogeneity but also provides a comprehensive analysis of the impact of physical and construction parameters on crack propagation. The study extends to numerical simulation studies on three typical heterogeneous rock layers-natural fractures, bedding planes, and faults. Utilizing indicators like fracture pressure, crack extension length, and damage area, a random forest algorithm quantifies the degree of influence of each factor on fracturing. Our results underscore varying impacts from geological factors (natural fractures, bedding angle, stress difference, elastic modulus, matrix porosity, and permeability) and construction factors (cluster spacing, perforation length, fracturing fluid viscosity, displacement, well spacing, and perforation opening) on the fracturing effect of heterogeneous rock layers. In conclusion, our proposed phase field framework demonstrates predictive capability across diverse heterogeneous situations. This innovative model not only rectifies past oversights but also offers insights unattainable through traditional experimental testing, significantly advancing our understanding of hydraulic fracture dynamics.