A Method for Constructing Relative Permeability Curves Using Nuclear Magnetic Resonance Logging Data: A Case Study of Low-Permeability Sandstone Formations in Xihu Sag

Summary Relative permeability is a crucial petrophysical parameter that influences multiphase fluid flow in hydrocarbon reservoirs. Traditional laboratory measurements are costly and prone to data gaps. In this study, we present a novel approach using nuclear magnetic resonance (NMR) logging to generate continuous in-situ relative permeability curves. Analysis of 16 core samples led to the development of a three-step predictive model: (1) deriving water-phase permeability based on capillary bundle theory and Poiseuille’s law, (2) correlating resistivity index with relative permeability using Archie’s equation, and (3) linking NMR transverse relaxation time (T2) to relative permeability using a fractal geometry model. A crucial adjustment was made to the Brooks-Corey model to correct for overestimation of gas-phase relative permeability, incorporating corrections for residual gas saturation based on pore structure classification. Validation against core data showed strong agreement and accurate prediction of isotonic points. Application of the model to the Huagang formation in the Xihu Sag confirmed its effectiveness in constructing relative permeability curves. In formations with NMR logging data, the isotonic point and corresponding relative permeability curves aligned well with experimental results. This approach reduces reliance on laboratory testing and enables real-time generation of relative permeability curves during drilling operations. The findings offer a practical tool for reservoir engineers to optimize waterflooding strategies and improve oil recovery in challenging reservoirs.