An experimental study on the effect of magnetic field strength and internal gradient on NMR-Derived petrophysical properties of sandstones

Abstract

Nuclear Magnetic Resonance (NMR) is being widely deployed as a powerful tool for estimating petrophysical properties. Compared to conventional carbonate reservoirs, sandstone reservoirs contain a high amount of paramagnetic ions (such as iron, nickel, or manganese) usually found in clays. The interpretation of NMR-derived petrophysical properties for sandstone formation can be complicated due to the paramagnetic ions which cause inhomogeneity in the magnetic field called internal gradient. Earlier studies focused on investigating the impact of clays and field strength on internal gradient, but implications for estimating NMR-derived petrophysical properties are not well understood yet. This study aims to investigate the impact of the internal gradient and magnetic field strength on the T2 relaxation times and the NMR-derived porosity, T2 logarithmic mean value (T2,LM), and permeability. NMR T2 measurements at several echo times were performed to evaluate the internal gradient for six sandstone samples characterized by variable porosity, permeability, and clay contents/mineralogy. The analysis was executed at two different operating Larmor frequencies (2 and 12 MHz), which are the most common frequencies for rock core analysis. The results show that the internal gradient magnitude increases with the increase of the paramagnetic-rich clays (Chlorite and Illite in the studied samples) while no correlation is observed with Kaolinite content. Furthermore, the internal gradient's magnitude and effect are consistently more pronounced at the 12 MHz measurements than those performed at 2 MHz. Porosity estimates from NMR are found to be independent of the internal gradient and the magnetic field strength as long as sufficiently low echo time (0.2 ms) is used. Unlike porosity, noticeable discrepancies in T2,LM were reported between 2 MHz and 12 MHz measurements, with the later showing consistently lower T2,LM values. The discrepancies in T2,LM values result in significant differences (up to 300%) when calculating NMR-derived permeability using the same formulations for both 2 and 12 MHz data. That is, calibration done using the T2 data from 12 MHz equipment cannot be applied for permeability estimation from lower field NMR data (like logging NMR). The discrepancies between 2 and 12 MHz in terms of T2,LM and derived-permeability increase exponentially as the internal gradient increases but such discrepancies diminish for tight rocks (permeability < 1 mD). This behavior of the tight sandstone samples is attributed to the relaxation regime, which is impacted by pore sizes. If pores are small enough, such as the case occasionally for shale or tight sandstone, relaxations move from a short time regime to the motionally averaged. In such a regime, the internal gradient is averaged by diffusion over the length scale, meaning that relaxation time has no dependence on echo time.