Modelling surface NMR spin-echo experiments in a heterogeneous B1 field

Abstract

SUMMARY Surface nuclear magnetic resonance (NMR) measurements show great promise for characterization of subsurface water content, pore-sizes and permeability. The link between surface NMR and pore-size/permeability is founded in the connection between the NMR signal's time dependence and the geometry of the pore-space. To strengthen links between the NMR signal and pore-geometry multipulse surface NMR sequences have been developed to estimate the parameter T2, which carries a strong link to pore-geometry and has formed the basis for NMR-based permeability estimation in the petroleum industry for decades. Producing reliable subsurface characterizations from multipulse surface NMR measurements that measure T2 requires that the forward model is able to accurately predict the transverse magnetization at the time when the measurement occurs. Traditional surface NMR T2 forward models employ an analytic expression for the transverse magnetization, an expression developed in the context of laboratory NMR experiments conducted under conditions significantly different from surface NMR and which require several assumptions to simplify the underlying Bloch equation. To investigate the reliability of this analytic expression under surface NMR conditions, a synthetic comparison is performed where the analytic expression is contrasted against the transverse magnetization predicted from a solution of the full-Bloch equation without the same simplifying assumptions and which can appropriately weight heterogeneity in the applied and background magnetic fields. The comparison shows that the analytic expression breaks down in a range of conditions typical to surface NMR measurements.