Abstract
It is essential to predict or evaluate permeable flow for the safety of structures constructed in a cracked rock mass. The permeable flow analysis, however, is not easy due to the limitation of the in-situ measurement of geological and hydrological conditions. To overcome this difficulty, we propose a pair of analysis methods, the Monte-Carlo simulation of fracture networks to estimate probabilistic distributions of permeability and the bounding media analysis to provide upper and lower bounds for the expectation of the permeable flow. An example problem of an actual rock mass is solved by using these two analysis methods. It is shown that the predicted distribution of the effective permeability becomes log-normal, and that the computed upper and lower bounds actually bound the mean of the permeable flow which is computed from the Monte-Carlo simulation of randomly generating porous media. The validity and usefulness of the proposed analysis methods are discussed.
