The effect of pore size and magnetic susceptibility on the surface NMR relaxation parameter T 2 *

Surface nuclear magnetic resonance (NMR) is a non-invasive geophysical method that can provide valuable information about aquifer properties related to groundwater flow and storage. Our ability to extract such information from surface NMR data, however, is limited by an insufficient understanding of the relaxation parameter T2 * governing the decay rate of the surface NMR signal in Earth’s magnetic field. In this study, we use a combination of numerical and laboratory experiments to systematically explore the effect of two key geologic properties, pore size and magnetic susceptibility, on the T2 * relaxation process. A one-dimensional numerical model is developed and parametrized to simulate the surface NMR response for a wide range of geologic materials. These simulations illuminate the processes controlling T2 * relaxation and identify conditions under which T2 * exhibits varied sensitivity to pore size. For materials with low magnetic susceptibility, T2 * is highly sensitive to pore size; however, as susceptibility increases, this sensitivity diminishes and T2 * becomes dominated by complex dephasing effects, particularly when pores are large. Laboratory Earth’s field NMR experiments complement the numerical simulations. Measurements on watersaturated quartz sands show that for weakly magnetic materials, T2 * can be sensitive to pore size and thus could provide useful information about aquifer properties. NMR have been explored (Legchenko et al. 2002; Hertrich 2008), however, the signal measured by surface NMR is fundamentally controlled by T2 * and it is this relaxation parameter that can be most robustly determined as a function of depth. The T2 * relaxation process measured by surface NMR is similar to T1 and T2 relaxation, in that it is partially controlled by the pore-scale environment of the groundwater under investigation. However, there is not a simple scaling relationship between T2 * and T1 or T2, and the link between T2 * and pore-scale properties is not well-understood. This gap in understanding hampers our ability to extract useful information about aquifer properties from surface NMR data. Further, it limits the extent to which we can predict the performance and reliability of surface NMR data in varied geologic environments. In contrast to the extensive history of T1 and T2 research, there have been very few studies of the relationship between T2 * and pore-scale properties. Several studies have attempted to substitute T2 * in the petrophysical relationships originally developed for T1 and T2, with mixed success (e.g., Legchenko et al. 2002; Lubczynski and Roy 2003). Only one study to date has directly explored the relationship between these various relaxation times using laboratory measurements (Müller et al. 2005). Further systematic studies of this type are needed to improve our fundamental understanding of the T2 * response in groundwater aquifers and our ability to effectively interpret these measurements. INTRODUCTION Proton nuclear magnetic resonance (NMR) has been used in the Earths sciences since the 1960s, in both laboratory and borehole measurements, as a means of characterizing the pore-scale properties of fluid-saturated geologic media. NMR is ideally suited for this purpose due to a strong link between the measured relaxation parameters T1 and T2 and the physio-chemical environment of hydrogen nuclei in the pore fluid. Numerous theoretical and empirical studies have demonstrated robust relationships between these relaxation parameters and geologic properties governing fluid flow and storage, such as pore size (e.g., Gallegos et al. 1987; Kenyon et al. 1989) and permeability (e.g., Timur 1969; Kenyon et al. 1988; Straley et al. 1997). Hence, NMR has proven to be a particularly useful well-logging tool in oil and gas applications. Over the last two decades, new surface-based NMR systems have been developed to extend NMR methods to shallow hydrogeologic applications (Semenov 1987; Legchenko and Valla 2002; Walsh 2008). These systems yield valuable non-invasive NMR relaxation measurements of groundwater to depths of 100 m. Unlike advanced borehole or laboratory NMR methods, which measure T1 and T2, standard surface NMR methods directly measure a related, but not equivalent, relaxation parameter T2 *. Methodologies for approximating T1 and T2 with surface Near Surface Geophysics, 2011, 9, xxx-xxx doi:10.3997/1873-0604.2010062 E. Grunewald and R. Knight 2 © 2010 European Association of Geoscientists & Engineers, Near Surface Geophysics, 2011, 9, xxx-xxx where the initial decay amplitude, E0, is proportional to the saturated pore volume. For the commonly assumed condition of fast-diffusion (Brownstein and Tarr 1979), the net observed relaxation rate represents two relaxation processes acting in parallel within the bulk water and at the pore surface: