Exploration on the characteristics of 3D crack network expansion induced by hydraulic fracturing: A hybrid approach combining experiments and algorithms

Abstract

To efficiently develop and utilize unconventional energy, it is essential to investigate the characteristics of 3D crack network expansion during sandstone hydraulic fracturing (HF). This paper presents the results of physical simulation experiments of sandstone HF under varying shear stress levels, where CT image data acquisition was used to enable 3D reconstruction and establishment of identification algorithms for 3D crack network. The graph theory representation method of crack network was improved using the intersection and extension of crack branch elements as precursors for two vertex sets U and V of a bipartite graph G=(U,V,E), extending the topological structure representation method to 3D crack network. Using topology and fractal theory, a quantitative characterization method of crack network was developed, which comprehensively considered topological structure parameters and fractal dimension. The findings showed that the fractal dimension of the crack increases linearly with the increase of the shear stress level, and the parameters of the crack topological structure increased first and then decreased. While the parameters of the crack topological structure first increased and then decreased.