A probabilistic, empirical model for permeability of mudstone

Argilaceous rock is being considered for geologic isolation of spent nuclear fuel in various countries. While permeability plays a central role in determination of transport and fate of radioactive material in earth's crust, most existing models for permeability are deterministic, and relate permeability to Terzaghi effective pressure. Laboratory and field data, however, (i) indicate that other parameters affect permeability of fluid-saturated porous rock and (ii) show significant variability in reported permeability measurements. For example, porosity, pore geometry, and clay size fraction, and the inherent variability of such parameters are important in successful prediction of permeability. This paper first presents a database of permeability measurements determined using laboratory and field tests that includes various rock parameters that affect permeability. A study of relationship of these parameters and permeability is presented. A framework is then presented for development of a probabilistic, empirical model for prediction of permeability of mudstone. The model is unbiased, accounts for the relevant uncertainties, and considers the important parameters affecting permeability.