Integrating Non-matrix Features into Permeability Modeling Using a Core-based 3D Probability Approach for Heterogeneous Carbonate Reservoirs, Offshore Abu Dhabi

Accurate permeability modeling is critical for reliable production forecasting in carbonate reservoirs, where conventional matrix-based models often significantly underestimate the actual flow capacity observed in well tests. This study introduces a novel approach that integrates non-matrix geological features—specifically dissolution vugs—into a three-dimensional, probability-based permeability model to improve the accuracy of dynamic reservoir simulations. The methodology was applied to a complex offshore carbonate reservoir located in Abu Dhabi, United Arab Emirates. Detailed core analyses identified the spatial distribution of dissolution vugs and their correlation with key reservoir properties such as rock types and porosity. These non-matrix features were found to be concentrated in high-energy depositional environments, which were effectively captured using Static Rock Typing (SRT) as a proxy. A probabilistic model was then developed using SRT logs from over fifty-one wells across the field, enabling the enhancement of the permeability distribution by incorporating non-matrix elements. This integrated approach provides a geologically consistent permeability model that better aligns with well test data, enhances the accuracy of history matching, and supports improved reservoir performance forecasting and optimization. It offers a robust framework for addressing lateral heterogeneity and improving fluid flow predictions in complex carbonate systems.