Hydrocarbon saturation in Bakken Petroleum System based on joint inversion of resistivity and dielectric dispersion logs

Abstract

Bakken Petroleum System (BPS) is composed of both conventional and unconventional units, which exhibits significant variations in lithology, rock texture, clay content, total organic carbon, accompanied by high connate water salinity, presence of disseminated pyrite grains, and low values of porosity. These petrophysical attributes of the BPS lead to inconsistency in the oil-in-place estimates for those obtained from Electromagnetic (EM) induction log, Nuclear Magnetic Resonance (NMR) log, dielectric dispersion log measured by Array Dielectric Tool (ADT), and Dean-Stark core measurements. For purposes of improved hydrocarbon saturation estimation and petrophysical characterization in the BPS, a joint-inversion-based interpretation was performed on dispersive electrical conductivity and dielectric permittivity measurements at 4 dielectric-log-acquisition frequencies and 1 induction resistivity acquired at 20 kHz. This analysis was performed across a 350-ft depth interval in one of the science wells intersecting the BPS. Three geo-electromagnetic mixing models, namely Complex Refractive Index model, Stroud-Milton-De model, and Waxman Smits model are integrated and coupled to the inversion scheme to simultaneously estimate water saturation, formation brine conductivity, cementation exponent and saturation exponent in BPS.Water saturation estimates obtained using the proposed interpretation method were compared against those obtained from NMR log, Dean-Stark core measurements and service company’s dielectric inversion. In Middle Bakken from depth XX720 to XX750 ft, our estimates of water saturation are in better agreement with those estimated by service company’s mineral inversion method and service company’s dielectric interpretation as compared to those obtained from NMR interpretation and Dean Stark core measurement. Water saturation and formation brine conductivity estimates in Middle Bakken are in the ranges of 0.5–1 and 25–45 S/m, respectively. Inversion-derived brine conductivity and saturation exponent estimates are most uncertain in Lodgepole and Three Forks 2 formations, which exhibit a wide range of pore size distribution. Average relative errors in matching the 1 induction resistivity and 8 dielectric dispersion logs using the inversion-derived estimates are 33% and 20%, respectively, in the 350-ft depth interval of BPS. The proposed inversion achieves high certainty for the estimates when the formation has low clay content, low electrical anisotropy, and high porosity.