Estimation of water saturation from nuclear magnetic resonance (NMR) and conventional logs in low permeability sandstone reservoirs

Abstract

Abstract It is difficult to obtain rock resistivity parameters by using the cross plots of porosity vs. formation factor and water saturation vs. resistivity index to calculate reservoir water saturation in low permeability sandstones. The cementation and saturation exponents ( m and n separately) are divergent, and no fixed values can be obtained due to the complicated pore structure. This leads to a problem in water saturation calculation. To investigate the main factors that heavily affect the cementation and saturation exponents, 36 core samples, which were drilled from low permeability sands of Xujiahe Formation, Sichuan basin, southwest China, are chosen for laboratory resistivity and nuclear magnetic resonance (NMR) measurements, 20 of them for mercury injection capillary pressure (MICP) measurements and 10 of them for casting thin-section analysis. The results show that these two parameters are associated with rock pore structure. For rocks with good pore structure, the proportion of macropore components is dominant, high cementation exponents and low saturation exponents can be obtained, and on the contrary, rocks with poor pore structure will be dominated by the proportion of small pore components, and they will contain low cementation exponents and high saturation exponents. To quantitatively acquire reliable cementation and saturation exponents for water saturation estimation, a logarithmic function is established to calculate cementation exponent from porosity. Irreducible water saturation ( S wi ), which is estimated from NMR logs by using the optimal T 2cutoff , is presented to characterize the proportion of small pore components. A technique of calculating saturation exponent by combining with S wi , (1− S wi ) and the logarithmic mean of NMR T 2 spectrum ( T 2 lm ) is proposed, and the corresponding model is established. The credibility of these techniques is confirmed by comparing the predicted cementation and saturation exponents with the core analyzed results. The absolute errors between the predicted cementation exponents and the experimental results are lower than 0.08, and the absolute errors between the predicted saturation exponents and the experimental results are lower than 0.2. These techniques proposed in this study are extended to several low permeability sands for field applications; the field examples illustrate that cementation and saturation exponents can be accurately estimated in the intervals with which NMR logs were acquired. By using the variable rock resistivity parameters, precisely water saturation can be calculated for low permeability sandstones evaluation.