Accounting for relaxation processes during the pulse in surface NMR data

Abstract

Under favorable conditions, the surface nuclear-magnetic-resonance (NMR) technique can provide direct quantitative estimates of subsurface water variations and indirect information on pore size and hydraulic conductivity. The technique is based on various exponential relaxation processes that become measurable after the intrinsic spin magnetic moments of groundwater protons have been rotated out of equilibrium by a pulse of alternating electromagnetic (EM) field generated at the surface. An implicit assumption in previous surface NMR studies is that relaxation processes need to be considered only after the EM pulse has been extinguished. Although this approximation is valid for short EM pulses, neglecting relaxation during the pulse (RDP) can result in significant errors for the generally long pulses used in surface NMR investigations. Because the influence of RDP cannot be isolated and quantified using field-scale approaches, this study is based on sample-scale NMR experiments and numerical simulations (Bloch equations) that mimic field practices and conditions. The results demonstrate that standard surface NMR methods that ignore the effects of RDP may yield significantly erroneous estimates of water volume (RDP-related errors of 25% are possible) and the key relaxation parameter (RDP-related errors of 50% are possible) that supplies information on pore size and hydraulic conductivity. Fortunately, the study also demonstrates that relatively simple interpretational approaches can reduce the RDP-related errors to less than 2%.