Quantifying Wettability Index Using NMR T2 Distributions of Water Phases Extracted with Non-Negative Matrix Factorization

Abstract

The wettability of formation is a critically important factor for optimizing the hydrocarbon reservoir production. It is a complicated and challenging issue to quantify the wettability of rock at in-situ conditions. Because of its importance, various methods have been developed to quantify the rock wettability. However, each method has its strength and limitations. This paper presents a new wettability quantification method based on featured T2 distributions of water. We define a new wettability index to quantify the wettability using NMR T2 distributions of water in wetting and non-wetting phases. We apply the principal component analysis (PCA) method to determine the number of components of the T2 distributions inverted from NMR measurements at different pore pressures. The non-negative matrix factorization is conducted on the matrix of T2 distributions with the determined number of components. A window technique is applied to extract NMR T2 distributions of water in different phases to quantify the rock wettability with the newly defined wettability index. NMR measurements were conducted on sandstone samples of variant wettability to validate this proposed method. The samples were prepared at different saturation conditions that were achieved using core flooding techniques. The NMR measurements acquired at different saturated conditions of core plug samples were inverted to obtain T2 distributions, whose peak amplitudes and positions indicate sensitivities to saturations and wettability. The wettability indices for the flooded cores were computed in the new wettability quantification method and compared with those from the Amott-Harvey method. The validation and comparison results show that: Compared with the Amott-Harvey method, the newly defined wettability index can be used to more effectively quantify the altered wettability of a surfactant flooded coreThe PCA identifies that fluid displacement processes (water displacing oil or oil displacing water) result in the shape changes of the T2 distributions inverted from NMR T2 measurements for flooded core plug samples.The non-negative matrix factorization can be applied to extract the feature components of T2 distributions of a flooded core, which are related to the wettability of the core.The T2 distributions of the mobile waters in water-wet pores and oil-wet pores can be extracted from identified components of T2 distributions.