Abstract: Evaluation of fracture network connectivity is an important part of studying the underground space while percolation analysis is an effective way to examine the connectivity of a fracture network. Percolation threshold of a fracture network is typically determined by indirect characterization parameters of fractures (eg., fractal dimension). However, fracture networks with different connectivity may have the same indirect characterization parameters, leading to decreased prediction reliability. To avoid this problem and to more accurately and quickly characterize fracture network connectivity, we built percolation threshold equations using direct characterization parameters (e.g., fracture length and number) by nonlinear fitting instead of using indirect characterization parameters in the simplified approach. The equations were drawn from the relationships between percolation thresholds and direct characterization parameters of different two-dimensional discrete fracture networks; application of these equations in different scaled fracture networks was discussed and validated. The results show that these equations can efficiently predict percolation thresholds of different scaled fracture networks. Based on this result, we developed criteria for estimating connectivity of fracture networks, which provide certain guidance and reference point for the evaluation of underground space in fractured areas.

Key words: underground space, discrete fracture network, percolation, connectivity, threshold, fracture properties