Analyzing the correlation between stochastic fracture networks geometrical properties and stress variability: a rock and fracture parameters study

Abstract

AbstractIn this study, the correlation between geometric properties of the fracture network and stress variability in a fractured rock was studied. Initially, discrete fracture networks were generated using a stochastic approach, then, considering the tensorial nature of stress, the stress field under various tectonic stress conditions was determined using finite-difference method. Ultimately, stress data were analyzed using tensor-based mathematical relations. Subsequently, the effects of four parameters including rock tensile strength, rock cohesion, fracture normal stiffness and fracture dilation angle on the stress perturbation distribution were evaluated. The obtained results indicated that stress perturbation and dispersion are directly related to fracture density, which is expressed as the number of fractures per unit area utilizing the window sampling approach. It was also demonstrated that they are inversely related to power-law length exponent which represents the length of fracture. It was observed that stress distribution, among the evaluated parameters, is more sensitive to the fracture normal stiffness and the effects of rock parameters on stress distribution are negligible. It was concluded that the highest stress distribution is created when the fracture network is dense with fractures having high length and low normal stiffness value.