Your abstract should contain less than 200 words using Times New Roman in font 11 with single spacing. The factors that influence on the fluid flow are hydraulic and geometric characteristics of fractures. So two-dimensional (2D) and three-dimensional (3D) discrete fracture network (DFN) approach has widely been used in order investigate the hydraulic behavior of fractured rock mass. 2D DFN approach can have a limit of analyzing 3D rock mass structures. But there were little studies on the relationship between 2D and 3D hydraulic behaviors of a fractured rock mass by a discrete fracture network.
This paper focuses on comparison of 2D and 3D hydraulic behaviors of a fractured rock mass using DFN. First, we developed the program that can generate the 2D and 3D DFN and analyze the connectivity of the generated fractures and then estimate the equivalent hydraulic conductivity tensors. And Numerical simulations were performed to compare 2D hydraulic behavior of fractured rock mass with 3D behavior using same hydraulic and geometric characteristics of fractures. The analysis results show that 2D and 3D results could be different. Hydraulic conductivity tensors Kxx, Kyy, Kzz by 2D analysis were larger than them by 3D. And a representative element volume (REV) size for the hydraulic behavior of a fractured rock mass in 2D analysis was larger than it by 3D. 2D DFN seemed to overestimate hydraulic conductivity tensors comparing with 3D. And Ratios of hydraulic conductivity tensors, Kxx /Kyy, Kyy /Kzz and Kzz /Kxx in 2D analysis were larger than 3D. The results can give reference to engineering projects in a fractured rock mass.
In discrete fracture network modeling, the distribution functions of orientation, length, and aperture of fractures in rock mass are assumed and a non-unique fracture network is generated by random number generation. Then the model is implemented into numerical analysis. This process assumes that the most important factor for the flow analysis in a crystalline rock mass is fractures. This approach was adopted in 1980s and the geometric characteristics of fractures in rock mass were intensively studied. But this approach has limitation in analyzing a large volume due to the computer capacity.
Billaux et al (1989), Kulatilake et al (1993) and others conducted researches on three-dimensional discrete fracture network model based on the existing assumed distribution functions of length, orientation, aperture, and position of fractures. Wang et al (2002) estimated a REV size and three-dimensional hydraulic conductivity tensors for a fractured rock mass through a single well packer test and discrete fracture fluid flow modeling. Kulatilake et al(2000) investigate an effect of block size and fracture geometry on fractured rock hydraulics and REV in two-dimensional. Baghbanan and Jing (2007) studied hydraulic properties of fractured rock masses with correlated fracture length and aperture using a two-dimensional discrete fracture network. In Korea, Song (1993), Chang (1996), and Ko (2000) studied on two- and three-dimensional network model and Chang (1998), Park (2000) used the commercial program NAPSAC and NAMMU for the analyses.
