An improved technique for computing permeability from NMR measurements in mudstones

We develop a technique for extending nuclear magnetic resonance (NMR) permeability estimation to clay-rich sediments. Our technique builds on the Schlumberger-Doll Research (SDR) equation by using porosity, grain size, specific surface, and magnetic susceptibility data to yield more accurate permeability estimation in mudstones with large pore surface areas and complex mineralogies. Based on measurements of natural sediments as well as resedimented laboratory mixtures of silica, bentonite, and kaolinite powders, we find that our method predicts permeability values that match measured values over four orders of magnitude and among lithologies that vary widely in grain size, mineralogy, and surface area. Our results show that the relationship between NMR data and permeability is a function of mineralogy and grain geometry, and that permeability predictions in clay-rich sediments can be improved with insights regarding the nature of the pore system made by the Kozeny theory. This technique extends the utility of NMR measurements beyond typical reservoir-quality rocks to a wide range of lithologies.