Dependency of NMR Relaxation Behavior from Water Saturation Degree in Loose Sediments

Nuclear Magnetic Resonance (NMR) applications in geophysics (in the lab, in boreholes and the field, respectively) provide estimations of the water content and, in the case of full water saturation, the saturated permeability. The NMR relaxation behavior at full saturation contains the structure information of the pore space, i.e. the pore size distribution. In principle, this information should also allow for estimating the unsaturated permeability as a function of the saturation degree. The key to do this is to understand and predict the relationship of the NMR relaxation times and the saturation degree. We introduce a reliable model to quantify this relationship and show its plausibility for various loose materials covering a broad range of grain sizes. We show the similarities of our approach to the Brooks-Corey model that is a standard model in soil sciences to estimate the relative hydraulic conductivity. Furthermore we show that the saturation dependency of the relaxation time distribution can be predicted from a single NMR measurement at full saturation. For such predictions we assume the pore space to consist of capillaries with various cross-sectional areas (circular and triangle-shaped cross-sections).