Other than commonly assumed the relaxation times observed in the electrical low-frequency range (1 mHz – 40 kHz) of natural porous media like sandstones and tuff stones cannot be directly related to the dominant (modal) pore throat sizes, measured (e.g.) with mercury intrusion porosimetry (MIP). Working with a great variety of sandstones from very different origins and featuring great variations in textural and chemical compositions as well as in geometrical pore space properties, it was observed that particularly samples with narrow pore throats were characterized by long (low-frequency) relaxations. These, however, can (following the current theories) be rather explained by long “characteristic length scales” in these media or low diffusion coefficients along the electrical double layer. However, there is no straightforward way (or single approved method) of getting reliable numbers for properties such as the lengths of pore throats, the diameter and length of the wide pores and their respective distributions. Consequently we follow a multi-methodical approach and combine the benefits of MIP, micro-computed tomography (μ-CT) and nuclear magnetic resonance (NMR) to achieve much deeper insight due to the different resolutions and sensitivities to either pore constrictions (throats) or wide pores. This helps us to understand, whether the observed electrical relaxation phenomena actually depend on geometric length scales or rather on other properties such as chemical composition, clay content, clay type or cation exchange capacity. In this paper, we showcase selected results of a systematic study of a total of 16 sandstones and three tuffs. Findings and the particular advantage of the used method combination are discussed and shown in detail for a representative sample selection.
Nuclear magnetic resonance (NMR) is a well established laboratory / borehole method to characterize the storage and transport properties of rocks due to its direct sensitivity to the corresponding pore fluid saturation (water or oil) and pore sizes.
For petrophysical applications there are several different NMR laboratory devices commercially available varying over a wide range of e.g. magnetic field strength / frequency (2 MHz to 30 MHz), applicable measurement protocols (T1, T2, T1-T2, T2-D, etc.) and sample sizes (2.5 cm to 10 cm in diameter). In this work we present NMR measurements, layed out in a round robin like manner, on a set of 20 sandstone samples. We use three different NMR devices containing two standard setups with homogenous magnetic fields (LIAG and RWTH) and one single-sided setup with gradient field (BGR) to measure T1 and T2 relaxation data. In our evaluation we especially focus on the comparison of the individually inverted relaxation time distributions to quantify the differences arising from different laboratory setups. Diverging results can be deduced on the one hand to the inherit differences between homogeneous and gradient fields but on the other hand also due to quality differences between the two homogeneous setups. Additionally, we also examine the influence of the individually chosen inversion parameters (signal processing, distribution sampling points, error weighting, regularization, etc.) to establish a general standardized best practice recommendation for future petrophysical NMR laboratory measurements.
Because of its mobility and ability to investigate exposed surfaces, single-sided (SiS) nuclear magnetic resonance (NMR) technology enables new application fields in geosciences. To test and assess its corresponding potential, we compare longitudinal (T1) and transverse (T2) data measured by SiS NMR with those of conventional geoscientific laboratory NMR. We use reference sandstone samples covering a broad range of pore sizes. Our study demonstrates that the lower signal-to-noise ratio of SiS NMR data generally tends to slightly overestimated widths of relaxation time distributions and consequently pore size distributions. While SiS and conventional NMR produce very similar T1 relaxation data, unbiased SiS NMR results for T2 measurements can only be expected for fine material, i.e. clayey or silty sediments and soils with main relaxation times below 0.05s. This limit is given by the diffusion relaxation rate due to the gradient in the primary magnetic field associated with the SiS NMR. Above that limit, i.e. for coarse material, the relaxation data is strongly attenuated. If considering the diffusion relaxation time of 0.2 s in the numerical data inversion process, the information content >0.2s is blurred over a range larger than that of conventional NMR. However, our results show that principle range and magnitudes of the relaxation time distributions are reconstructed to some extent. Regarding these findings, SiS NMR can be helpful to solve geoscientific issues, e.g. to assess the hydro-mechanical properties of the walls of underground facilities or to provide local soil moisture data sets for calibrating indirect remote techniques on the regional scale. The greatest opportunity provided by the SiS NMR technology is the acquisition of profile relaxation data for rocks with significant bedding structures at the µm scale. With this unique feature, SiS NMR can support the understanding and modeling of hydraulic and diffusional anisotropy behavior of sedimentary rocks.