Effects of asphaltene and resin contents on crude oil viscosity: Experimental and modeling studies

• A novel model predicts dead oil viscosity using temperature, API gravity, and asphaltene and resin contents. • Asphaltene and resin contents significantly influence dead oil viscosity. • The model achieved AAREs of 18.53% (training) and 20.83% (validation). • The model outperformed 19 existing correlations with R 2 values of 0.97 and 0.96. Dead oil viscosity is critical for modeling crude oil flow in porous media and pipelines, designing production facilities, and enhancing oil recovery. However, most dead oil viscosity models fail to consider the effects of asphaltene and resin content. This study introduces a novel nonlinear regression-based model that predicts dead oil viscosity as a function of temperature, API gravity, asphaltene content, and resin contents using a dataset of 357 experimental measurements obtained from three heavy oil samples and their reconstituted derivatives tested between 77–176 °F at atmospheric pressure. The saturates, aromatics, resins and asphaltenes (SARA) fractions of these oil samples were determined using thin-layer chromatography and validated with both automated SARA high-performance liquid chromatography and absorption spectroscopy following the Japan Petroleum Institute (JPI-5S-45–95) standards. The results confirm that variations in asphaltene and resin contents significantly influence oil viscosity. A new viscosity correlation was developed using 250 datasets for model training and validated with the remaining 107 datasets from the total dataset. By incorporating asphaltene and resin contents, the model demonstrates superior accuracy and reliability compared to 19 existing viscosity correlations. It achieved the lowest average absolute relative errors of 18.53 % and 20.83 % for the training and validation datasets, respectively, along with the highest coefficients of determination (R 2 = 0.97 and 0.96). These findings indicate that the proposed model offers a more robust and accurate alternative for predicting dead oil viscosity, particularly for heavy oils where polar components play a critical role. This improved predictive capability can significantly enhance the accuracy of flow modeling, production system design, and enhanced oil recovery planning, especially in fields dealing with complex heavy oil systems.