Flow in fractured rock.

In fractured rocks of low permeability, the hydraulic properties of the rock mass are strongly influenced by the connectivity and fracture geometry of the fracture system, the stiffness and deformational properties of fracture surfaces and the geostatic stresses. This dissertation demonstrates through theoretical analysis the sensitivity of fracture connectivity and flow rate to fracture radius, fracture density and measurement scale. Percolation factor and percolation frequency are suggested as indices of connectivity and flow rate. Models of hydromechanical coupling, normal closure and simple geometrical joint systems are proposed to study fracture porosity and fracture permeability and are compared with measured values from published papers. Fracture surface characteristics are expressed as indices of JRC and JCS. The relation between fracture aperture and geostatic stress is also examined. Based on the percolation process, a three dimensional discrete fracture model with statistical parameters of fracture geometry is developed to investigate the relations between the connectivity and flow rate and the percolation factor and percolation frequency. This model has the capability to generate a fracture network and to solve for steady state flow. The flow through each fracture is discretized by the boundary element method. By performing numerical simulation, the percolation threshold was found to be in the range of 0.9 to 2.4 for orthogonal joint sets. There is a rapid increase in flow rate with increasing fracture density or fracture length as the percolation factor reaches the percolation threshold. The percolation factor and percolation frequency are scale-dependent. A fracture network with large fractures and a high fracture density has a high percolation frequency and a high percolation factor. A network with high percolation frequency and percolation factor has a high flow rate. A constitutive model linking the initial conducting aperture, mechanical conducting aperture, JRC, JCS, geostatic stress and fracture geometries can be used to predict fracture porosity and fracture permeability in terms of depth. Preliminary comparison with field data shows that models comprising three orthogonal sets and disc-type fracture models can be used to simulate observed behavior. Fracture porosity and fracture permeability based on a model of three orthogonal sets can be used to predict flow through volcanic rocks. For sedimentary rocks a model incorporating a model of three orthogonal sets and one parallel set can be used with varying levels of confidence