Genetic units averages of kappa for capillary pressure estimation from NMR transversal T2 distributions (Niger Delta as field case study)

Abstract

Laboratory-based nuclear magnetic resonance (NMR) measurements from cores and the distribution of transversal T2 relaxations are usually employed for modelling petrophysical properties of reservoir rocks, due to their sensitivity to pore-size distribution, which invariably controls capillary pressures and permeabilities. Several methods have been proposed to derive synthetic drainage capillary pressures directly from NMR data. However, most approaches lack geologic calibration of the dataset and empirical correction factors are usually introduced in the scaling factor to improve the capillary pressure prediction at low wetting phase saturation. Volokitin et al. (A practical approach to obtain primary drainage capillary pressure curves from NMR core and log data, SCA-9924, 2001) proposed an approach for NMR T2 capillary pressure modelling based on a relationship between the NMR T2 and pore-throat distribution embedded within a single universal proportionality parameter termed kappa, κ; where: they concluded that the universal scaling factor, κ = 3 psi Hg s seems to work for all clastic reservoirs. In this present work, analysis aimed at validating the existence of the proposed relationship between measured capillary pressures and inverse NMR T2 distribution after Volokitin et al. indicate a nonlinear relationship. Pore throat distribution and accessible pore volumes analysis using a full-bore core mercury injection capillary pressure data from the Niger Delta Deep Water canyon reservoir demonstrates a single kappa factor is insufficient in characterizing pore throat distribution for reservoir with complex lithologies. Measured capillary pressures were compared with NMR derived ones for calibration with emphasis on various genetic reservoir units (lithofacies association). It suffices then to conclude that there exists no direct linear relationship between the NMR T2 distributions and pore throat distributions. In addition, a constant scaling factor for kappa is insufficient in modelling subsurface capillary pressures/pore throat distribution from NMR T2 data. An improved approach for NMR capillary pressure modelling can be obtained through local calibration, possibly allowing the kappa, κ scaling factor to vary as a function of rock types (pore size/grain sorting) and influence of clay diagenesis within the clastic system. The method has been applied successfully to a number of core samples from the clastic turbidites with a wide range of permeabilities. The NMR derived and primary drainage mercury injection capillary pressure curves show a very good agreement.