Effect of discrete fractures with or without roughness on seepage characteristics of fractured rocks

This study proposes a new fractal permeability model for fractured rocks that comprehensively accounts for the geometric fracture characteristics and the fluid transport mechanism. Then, the permeability changes of fractured rocks are analyzed using discrete fracture networks (DFNs) with or without roughness and different geometry parameters in the DFN modeling and finite element simulation. The results show that the proposed permeability model well agrees with the experimental data, and the established DFN numerical model more realistically reflects the fracture network in fractured rocks. Fluctuation of tortuous fracture lines (rough fractures) increases the fracture intersection probability, consequently increasing the fracture intersection area or connecting adjacent fractures. Moreover, permeability increases with the fractal dimension Df, porosity ϕ, maximum fracture length lmax, and proportionality coefficient β, and it decreases with increasing fractal dimension DTf of fracture tortuosity. When the fracture proportionality coefficient is 0.001 ≤ β ≤ 0.01, different DFNs yield similar simulation results for permeability. However, with increasing fracture network complexity, the predictive model created using conventional DFN (C-DFN) increasingly overestimates the fractured rock permeability. Thus, building a permeability model for a fractured rock using rough DFN (R-DFN) is more effective than that using C-DFN. Our findings are helpful for real permeability predictions via DFN and analytical modeling.