(564c) Application of Real-Component Based Molecular Characterization on Petroleum Refinery Simulation | AIChE

(564c) Application of Real-Component Based Molecular Characterization on Petroleum Refinery Simulation


Wang, M. - Presenter, Texas Tech University
Chen, C. C., Texas Tech University

As light to medium sweet crude oils keep depleting while demand of oil products are increasing, heavier and sourer crude has become a growing component to global oil market. Accompany with that, extreme bulk properties (including high molecular weights, high boiling points, large densities, high viscosities, etc.) are expected. However, current experimental approaches are still limited in characterizing heavy crudes in two aspects: 1) chemical structures of hydrocarbons with carbon number up to merely 12 can be accurately analyzed; 2) distillation cuts can only be obtained up to 750 K of normal boiling point, even with deep vacuum fractionation apparatus1. With limited experimental data available, less than 50 wt. % of individual heavy oil or bitumen can be analyzed, which renders simple model extrapolation over the whole crude highly unreliable.

The present work compares performance of petroleum refining simulation between two different methods: 1) the traditional API correlations commonly used in industry, 2) a real-component based molecular characterization method2derived from thermodynamically-consistent correlation of available crude oil data on chemical and physical properties. Compared with the former method, the later one not only gives reasonable predictions for properties of crude oil including heavy cuts, but also provides chemical structures and compositions of model constituent components over the whole crude oil, which serves as a cornerstone of high fidelity refining simulation.


1. Sánchez-Lemus, M. C., et al. (2016). "Physical properties of heavy oil distillation cuts." Fuel 180: 457-472.

2. Chen, C.-C. and H. Que (2013). "Method of characterizing chemical composition of crude oil for petroleum refinery processing". US Patent Application 20130185044 A1.