Inference of the preferred direction of fractures in the universal elliptical disc model based on multiple borehole information

Abstract

Obtaining the preferred direction of fractures is helpful for reconstructing the stress state of rock mass and assessing its engineering geological properties. However, most fractures are buried within the rock mass, and we could not directly obtain the preferred direction. Instead, we can infer it based on fracture information related to the preferred direction that is exposed on rock outcrops or within boreholes. To describe the preferred direction of fractures, the universal elliptical disc (UED) model is introduced, characterized by its orientation and rotation angle. In this study, we infer the preferred direction of fractures in the UED model based on the fracture information intersecting with the borehole. Among them, the inference of the orientation is relatively mature and can be interpreted through the trace formed by intersecting with the borehole wall, while the inference of the rotation angle is the focus of this paper. To infer the rotation angle, the mathematical relationship between the geometric parameters of the UED model and the number of borehole intersections is first established. Subsequently, an optimization function is constructed using Monte Carlo simulation integration, and a genetic algorithm-based method for inferring the mean rotation angle is proposed. A hypothetical case is designed to validate the effectiveness of the proposed method. The proposed method is applied to determine the preferred direction of the mining-induced fractures in the Ji14-31050 mining face of Pingdingshan Coal Mine No. 10 in China. Moreover, the successful inference of rotation angle based on borehole information also provides a reference for inferring other parameters of the UED model and further summarizes the combination of borehole and outcrop numbers required to infer all geometric parameters of the UED model.