A predictive model correlating permeability to two-dimensional fracture network parameters
Richeng Liu,Tantan Zhu,Yujing Jiang,Bo Li,Liyuan Yu,Yan Du,Yingchao Wang
DOI: https://doi.org/10.1007/s10064-018-1231-8
IF: 4.2
2018-01-29
Bulletin of Engineering Geology and the Environment
Abstract:This study presents a predictive model of permeability with correlation to parameters of two-dimensional fracture networks, including mass density of fractures (dm), average number of intersections per meter at the inlet and outlet boundaries (din), and connectivity (Cr). The fracture networks were constructed by considering the influence of fracture number density, fracture length, orientation, and variance of fracture length. A total of 86 discrete fracture network (DFN) models were established using the Monte Carlo technique, and the relationships between permeability and dm, din, and Cr were analyzed. By fitting these calculated results, a multi-variable regression function was proposed for predicting permeability. The results show that the number density of fractures plays a more significant role in permeability than fracture length. Fracture orientation can change the connectivity of fracture networks robustly, and thereafter influence the permeability. Although variance of fracture length can affect the pattern of cumulative frequency – fracture length curves, the variance has negligible influence on the permeability of fracture networks. A necessary condition to form connected flow paths from inlet boundary to outlet boundary is: dm > 0.27 m/m2, din > 0.19 /m, and Cr > 1.20. The proposed regression function can predict permeability with a correlation coefficient larger than 0.87, and its validity is verified by comparisons with results reported in the literature. Finally, the potential future works that can facilitate the predictive models of permeability are pointed out as open questions.
engineering, environmental,geosciences, multidisciplinary, geological