No abstract

Summary The use of nuclear magnetic resonance (NMR) techniques allows in-situ characterisation of geophysical properties such as moisture content, permeability and wettability. However, the accuracy and applicability of such measurements is limited by internal magnetic field gradients which are a consequence of magnetic susceptibility differences at solid-fluid interfaces. Such effects are particularly prominent in iron ore rock samples which contain ferrimagnetic and ferromagnetic mineralogy leading to high magnetic susceptibility. Multiple echo time Carr-Purcell-Meiboom-Gill (CPMG) NMR pulse sequences are commonly used to capture the influence of internal gradients, with the intention of deconvoluting diffusion in effective internal gradients (geff) from true transvers relaxation (T2). The interpretation of such measurements is complicated by the presence of multiple diffusive regimes: the short-time, motionally averaged and localisation regimes respectively. We introduce a new model for diffusive NMR signal attenuation, called the multi-regime model which is intended to better capture diffusive behaviour across the three regimes. The multi-regime model is compared against previous methods for quantifying diffusive decay (the short-time only and generalised inversion models). Multi-echo measurements of iron ore samples are fit with each model in order to quantify 2D T2-geff distributions. The resulting distributions demonstrate how the multi-regime model can provide insight into the relative influence of the different diffusive regimes in a given sample. This assists in understanding the influence of diffusive decay on measurement accuracy, for example the increased measurement error with increasing prevalence of the localisation regime. The multi-regime model provides a key step in accurately segregating surface relaxation and diffusive relaxation, which is crucial for accurately estimating pore size distributions, permeability and wettability in high magnetic susceptibility samples using NMR.