NMR relaxation of clay/brine mixtures

Effective interpretation of nuclear magnetic resonance (NMR) logs in shaly sands requires an understanding of the NMR contribution of clays. Of particular importance is the role of clays in the rapidly relaxing part of the NMR signal. In this study we measured the transverse relaxation time spectrum (T-2) of brine mixed with four clays (illite, smectite, kaolinite and glauconite) as a function of compaction. The Larmor frequency was 2 MHz and the echo spacing 0.16 ms. Mild compaction was achieved by centrifuging the clay slurry at three successive pressures ranging from 1 to 125 psi. Highly compacted samples were produced in a uniaxial press at six sequential pressures ranging from 500 to 16,000 psi. Each clay/brine slurry and its associated compacted sample showed a single peak in the T-2 distribution spectrum. A second peak, which could be interpreted as the "clay-bound water," was never observed. The T-2 peak position shifted to faster relaxation times as compaction increased, in proportion to the pore volume-to-surface ratio, V-p/A(s). The single peak and V-p/A(s) proportionality are consistent with fast diffusion between the pore water and the monolayer of water on the clay surface. Surface relaxivity varied among the four clay minerals; glauconite, the clay with the highest magnetic susceptibility and iron content had the largest surface relaxivity. These results have important implications for the interpretation of NMR logs in shaly sands. Because of the effects of compaction and to a lesser extent the iron content on a clay's T-2 peak position, it is not possible to independently determine clay type from some characteristic relaxation time. These data also imply that it is not feasible to estimate the cation exchange capacity from a single time cutoff of the T-2 distribution without additional information such as laboratory measurements or other log data.