472197 Molecular Characterziation and Modeling for Feedstocks Including Heavy Oil and Reactors

Monday, November 14, 2016: 12:52 PM
Taylor (Hilton San Francisco Union Square)
Shu Wang, Suphat Watanasiri and Darin Campbell, Aspen Technology, Inc., Bedford, MA

Simulation of refinery-wide processes helps the petroleum industry to design, optimize and operate the refineries profitably and safely, while meeting all necessary regulatory requirements. The traditional pseudo-component approach developed over 30 years ago is still widely used today to represent the feedstocks and products in these process models. This approach uses correlations and heuristics to provide thermodynamic and other properties required in the simulation. For refinery reactor models, kinetic lumps are used to represent molecular classes that undergo reactions to produce the desired products. These kinetic lumps are determined from either finger-print data of the feed or correlations that are based on other properties (such as PNA or sulfur contents) of the feed. Both the pseudo-component and kinetic lump representations do not contain molecular level information. Due to the increasing demands of environment regulations, changing feedstocks, higher margins, etc., the petroleum industry is very interested in improving the accuracy and understanding of the refinery process models and has been exploring the use molecular information for quite some time. In this presentation, we will describe the molecular approach recently developed to leverage molecular knowledge to increase the accuracy of process simulation. The molecular method was applied to characterize crude oils using traditional laboratory assay data (quantitatively match) as well as analytical data such as Fourier Transform Ion Cyclotron Resonance Mass Spectrometry (FT-ICR-MS) (qualitatively match). The resulting molecular compounds, compositions and associated properties are then used to calculate the properties of the process streams. The molecular information are also used directly to represent kinetic lumps composition used in refinery reactor models. 

Extended Abstract: File Not Uploaded