Abstract
ABSTRACT: Previous studies have revealed that hydraulic fracturing behavior in crystalline rock (i.e., granite) is highly dependent on mineral content, grain size, and heterogeneity and is therefore quite different from the behavior observed in sedimentary rock (i.e., sandstone). We investigated hydraulic fracture paths generated via continuous and cyclic injections in Gonghe granite through uniaxial testing. Hydraulic fracturing breakdown pressure was reduced by 20% in cyclic injection. In particularly, we analyzed the role of biotite grains and pre-existing defects - such as natural fractures - in controlling hydraulic fracture propagation. We found that hydraulic fracture crossed natural fracture either without or with an offset in continuous injection, while hydraulic fracture was arrested by natural fracture in cyclic injection, and NF is more likely to be activated during high-cycle injection. This can be beneficial if hydro-shearing of pre-existing flaws is anticipated. 1. INTRODUCTION The grain-scale heterogeneity of rock has been found to greatly affect crack growth behavior in mechanical fracturing tests. Previous studies have revealed that hydraulic fracturing in crystalline rock for geothermal purposes (i.e., granite) is highly dependent on mineral present, grain size and heterogeneity (Ishida et al., 2000; Zhuang and Zang, 2021). A growing number of work has employed grain-based model coupled with fluid flow in pore network to investigate hydraulic fracturing behavior in crystalline rock, and has reported unique results in terms of asymmetric fracture growth, zipper-mode crack initiation, and the partial re-closure of previous cracks (Kong et al., 2021; 2022), and influence of spatial distribution of minerals (Li et al., 2021). Cracks can be induced in mineral grains (intragranular or transgranular cracking) or along grain interfaces (grain-boundary or intergranular cracks). An in-depth analysis of the hydraulic fracture path can shed some light on the fluid-driven fracture nucleation and fracture propagation process (Chen et al., 2015; Ishida et al., 2016; Zhuang et al., 2020).
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have