Method of Determining Unconventional Reservoir Saturation with NMR Logging

Abstract

Abstract Nuclear magnetic resonance (NMR) T1-T2 logs from several west Texas wells over various formation layers were acquired to assess the values of NMR logging for fluid typing and saturation quantification in tight mudstone reservoirs. To quantify fluids in those low-porosity formations, high-density NMR data with adequate signal-to-noise ratio (SNR) are essential, and an inversion algorithm capable of resolving fluids on T1-T2 map is needed. These requirements are achieved with multifrequency data acquisition having a short interecho spacing (TE) for all frequencies, which yields a total of 15,400 echoes per depth in a single NMR pulse sequence cycle. This large data density enables less vertical averaging to be needed in both 1D and 2D processing, which is essential for capturing the porosity and fluid volume variation in these layered formations. The use of the inversion-forward modeling-inversion (IFMI) method improves the spectral resolution of the relaxation time distribution and T1-T2 map resolution because the method adequately addresses the effect of gradient differences in NMR responses in multifrequency data acquisition. 2D T1-T2 map fluid resolution of downhole stations and moving logs are compared to determine the minimal stacking required for fluid discrimination. The porosity and water volume are compared with those of conventional and mineralogy logs.