A novel three-dimensional discrete fracture network model for investigating the role of aperture heterogeneity on fluid flow through fractured rock masses
2019
Abstract Effect of anisotropic aperture on the hydraulic properties of single rock fractures has been systematically investigated, yet the aperture variability of individual fractures in 3D discrete fracture networks (DFNs) is commonly negligible by using parallel-plate fractures. The present study proposed a 3D DFN model with fractures having heterogeneous apertures to estimate the influence of fracture variability on fluid flow. In total, a set of 1280 3D models with increasing fracture densities and fracture lengths are generated and the fluid flow through the models is simulated using a developed numerical code. The influences of aperture heterogeneity and network topology on the flow pattern and permeability of 3D DFNs are estimated. The results show that the network topology provides a first-order frame of geometrical connectivity, and the heterogeneous aperture further allows the flow to select some most transmissive channels within these connected fractures. The DFN model with identical apertures generates a large number of medium-flow rate regions whereas the DFN model with fractures having heterogeneous apertures results in extremely low- and high-flow rate regions. The permeability ratio of the two models is widely spread in terms of a small variation in the average mechanical aperture as a result of strong dependence on the aperture distribution. The average permeability ratio increases significantly first and then approaches to 1.0 with increasing the average mechanical aperture. This allows for the definition of a critical mechanical aperture, above which the permeability can be properly predicted using the DFN model with fractures having identical apertures and below which the permeability is much altered by the aperture variability and the DFN model with fractures having heterogeneous apertures should be employed.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
88
References
18
Citations
NaN
KQI