Study on three-dimensional fracture network connectivity path of rock mass and seepage characteristics based on equivalent pipe network

Nuclear waste repositories have extremely stringent requirements for geological environment. However, natural fractures in rock mass can be potential channels for nuclide migration, therefore, the influence of fractures on the permeability of rock mass must be assessed. In this paper, a well research was conducted on well-exposed granite outcrops in the Xinchang site (the Chinese high-level radioactive waste repository). The high-precision three-dimensional model of a typical outcrop is built to obtain fracture information combined with field measurement, and then the three-dimensional fracture network model is generated using the relevant parameters by Monte Carlo method. To obtain more comprehensive fracture connectivity while avoiding the traditional method of searching the connectivity path in the complicated 3D fracture model taking up a lot of storage space and costing a lot of time, this paper presents an approach using MATLAB cell array instead of traditional adjacency matrix to search and store fracture network connectivity paths. In DFN model, the fracture disc with certain thickness is equivalent to three-dimensional pipe network model (EPNM) with variable diameter, and the equivalent path permeability coefficient (EPC) is proposed to objectively study the permeability characteristics of the seepage path in fractured rock mass based on that. Especially noteworthy is that some fractures in a certain strike range belong to open type, while those in another range belong to cemented closed fractures, when fresh fractures were exposed by cutting off the surface rock to a certain depth. The calculation of EPC under different conditions shows that the order of magnitude of EPC mean value is 1e−7m/s and 1e−3m/s, respectively, when fractures are cemented and not partly. On this basis, the size of the representative elementary volume (REV) of the fractured rock mass in the study area is determined to be about 25 m. By rotating the matrix in model, the spatial permeability tensor of the region (including permeability principal value and main direction) is obtained, which is within the range of borehole data. The predicted results may provide some reference for the related projects in the future.