Prediction of crude oil physical properties based on molecular compositional modeling approach

Abstract In this study, the Structural Units‐Bonding Matrix will be introduced into the crude oil characteristics analysis, specifically engineered to depict the molecular intricacies of petroleum substances. This methodology synergizes the bonding matrix with structural units, assigning the bonding matrix to the foundational layer, while the structural units adorn the surface layer. Through meticulous analysis of crude oil data, this study identified the core and branched molecular constituents of crude oil, facilitating the digital articulation of these molecules. Leveraging this foundational work, this study developed composition models for five quintessential types of crude oil, enabling the precise prognostication of their macroscopic physical properties. Further, by evaluating the viscosity‐temperature characteristics of these selected crude oils, this study established a predictive model for the viscosity–temperature relationship of crude oils, grounded in the quantitative structure–property relationship method. This approach not only augments this study understanding of crude oil molecular composition but also enhances the predictive accuracy of their physical properties, heralding a significant advancement in the field of petroleum molecular science.