This paper aims to examine the validity of the discrete fracture network (DFN) method in representing a realistic two-dimensional fractured rock in terms of their geomechanical response to in-situ stresses and hydraulic behaviour in a steady state fluid field. First, a real fracture network is extracted from the geological map of an actual rock outcrop, which is termed the analogue fracture network (AFN). Multiple DFN realisations are created using the statistics of the analogue pattern. A conductivity parameter that was found to have a linear relationship with the conductivity of 2D fracture networks is included to further enhance network similarity. A series of numerical experiments are designed with far-field stresses applied at a range of angles to the rock domains and their geomechanical response is modelled using the combined finite-discrete element method (FEMDEM). A geomechanical comparison between the AFN and its DFN equivalents is made based on phenomena such as heterogeneity of fracture-dependent stress contours, sliding between pre-existing fracture walls, coalescence of propagating fractures and variability of aperture distribution. Furthermore, an indirect hydro-mechanical (HM) coupling is applied and the hydraulic behaviour of the porous rock models is investigated using the hybrid finite element-finite volume method (FEFVM). A further comparison is conducted focusing on the hydraulic behaviour of the AFN and DFNs under the effects of geomechanical changes. The results show that although DFNs may represent an AFN quite well for fixed mechanical conditions, such a representation may not be dependable if mechanical changes occur.
Read full abstract