Molecular Modeling Of Composition And Separation Process Of Petroleum Fractions

The simulation of petroleum separation process is crucial for process control and optimization.Traditional process simulation software usually uses the pseudo-component method to calculate the product yield and partial properties.The molecular level simulation can obtain the complete molecular composition and physical properties of products,which is the future direction of the separation process simulation.In this thesis,a new hybrid framework,which is named the structural unit and bond-electron matrix(SU-BEM),is developed to reconstruct the molecular composition of petroleum fractions.A thermodynamic calculation method at the complex molecular level is developed and combined with the process simulation software to simulate the molecular orientation and predict the final composition and properties of products.The compositional model of petroleum fraction is first constructed by the SU-BEM framework.The structural units and bond-electron matrix represent the basic structural components and the atom connectivity of petroleum molecules to complete the multilevel description of molecules and construct a representative molecular library of petroleum.The group contribution method is implemented in the framework to predict the physical and chemical properties of petroleum molecules based on their chemical structures.A flexible probability density function sequence module is developed in the sampling procedure.All the parameters are optimized by the genetic algorithm to obtain the relative content of representative molecules.The method is validated for applications involving diesel,vacuum gas oil,and fluid catalytic cracking slurry oil.The detailed information about the molecular distribution and bulk properties of the petroleum fraction can be obtained.The pseudo-component method and molecular-level compositional model are used to simulate the atmospheric distillation process of crude oil in Aspen Plus.We adopt the rigorous reversible lumping method to simplify the thermodynamic model of complex molecular systems,which reduces the complexity of the model.Then the product yield and distillation curve of the molecular-level compositional model are calculated to compare with those of the pseudo-component method.Taking the atmospheric gas oil as an example to predict its distributions of boiling point,carbon number and ring number.Based on the obtained molecular composition,the bulk properties of each product are predicted.The compositional model is also applicable to the simulation of the supercritical extraction separation process of heavy oil.The extraction separation of the catalytic cracking slurry is carried out through the pseudo-component method and the rigorous reversible lumping method,respectively.Comparing the product yields at different extraction temperatures,it is proved that the rigorous reversible lumping method can achieve the same calculation accuracy as Aspen Plus.Besides,the advantage of this method is that the molar distribution of all molecules in each product can be calculated.Therefore,this method can predict bulk properties of the products and analyze the variation of product properties at different extraction temperatures.It is expected to provide guidance for the design and optimization of molecular processes based on supercritical extraction of heavy oil.