(413f) Molecular-Level Kinetic Modeling of Hydrocracking Process and Automatic Simplification of Reaction Network | AIChE

(413f) Molecular-Level Kinetic Modeling of Hydrocracking Process and Automatic Simplification of Reaction Network


Zhang, C. - Presenter, University of Illinois at Urbana-Champaign
Qiu, T., Tsinghua University
Hydrocracking is an important secondary processing technology in the petroleum refining. It converts low-value heavy oil fractions (such as vacuum gas oil) into lighter and more valuable products (such as naphtha and diesel). As the quality of crude oil deteriorates, hydrocracking technology, which can handle heavy fractions, has gradually become an important process for fuel production. With the pursuit of higher economic benefits by refining enterprises, stricter environmental supervision by the government, and China's commitment to achieving carbon neutrality by 2060, there is an increasingly strong demand for more detailed descriptions of reaction processes. The ultimate goal is to achieve molecular management in the refining process. Molecular-level kinetic modeling of the reaction process is an important part of molecular management. By providing detailed reaction data through molecular-level kinetic models, the optimal reaction parameters and conditions can be found, and the balance of improving economic benefits and reducing pollutant emissions can be achieved(Van de Vijver et al., 2016).

In this paper, a molecular-level modeling method for hydrocracking process based on structure-oriented lumping (SOL) is proposed, and a reaction network simplification method based on node classification and network flow analysis is proposed for the constructed complex reaction network. The SOL method constructs molecules through structural increments to achieve the molecular-level lumping, reducing complexity while achieving molecular-level description(Quann and Jaffe, 1996, 1992). Based on the SOL method for vector representation of molecules and the concept of similar reactions among homologues, reaction rules can be applied to automatically generate reaction networks. The kinetic parameters of the reaction were estimated based on a three-parameter model(Ghosh et al., 2009), taking into account the reaction rules, homologous categories of reacting molecules and the complexity of molecular structures. Partial parameters were adjusted through optimization. After parameter optimization, the constructed molecular-level kinetic model matches well with actual production data, achieving accurate modeling of the hydrocracking process.

The constructed reaction network contains a large number of molecules and reactions. In order to further improve the calculation speed, the reaction network is simplified. By representing the reaction network in the form of a Petri-net(Koch, 2010), both molecules and reactions are included in one network. Based on the classification of reaction rules, the network flow analysis method sorts the importance of all nodes in the reaction network and automatically discards unimportant molecules and reactions through an iterative process(Bi et al., 2023, 2020; Fang et al., 2016). Finally, while keeping the calculation results almost unchanged, the reaction network size was significantly reduced (by about 50%), which provides more possibilities for future industrial applications of molecular-level kinetic models.


Bi, K., Chen, T., Qiu, T., Ji, X., Dai, Y., 2023. Reaction Network Simplification and Key Routes Extraction for Steam Cracking Process. https://doi.org/10.2139/ssrn.4393676

Bi, K., Zhang, C., Qiu, T., 2020. An ingenious characterization of reaction network using sub-network reconstruction. Computers & Chemical Engineering 134, 106695. https://doi.org/10.1016/j.compchemeng.2019.106695

Fang, Z., Qiu, T., Chen, B., 2016. Improvement of ethylene cracking reaction network with network flow analysis algorithm. Computers & Chemical Engineering, 182–194. https://doi.org/10.1016/j.compchemeng.2016.04.020

Ghosh, P., Andrews, A.T., Quann, R.J., Halbert, T.R., 2009. Detailed Kinetic Model for the Hydro-desulfurization of FCC Naphtha. Energy Fuels 23, 5743–5759. https://doi.org/10.1021/ef900632v

Koch, I., 2010. Petri Nets – A Mathematical Formalism to Analyze Chemical Reaction Networks. Molecular Informatics 29, 838–843. https://doi.org/10.1002/minf.201000086

Quann, R.J., Jaffe, S.B., 1996. Building useful models of complex reaction systems in petroleum refining. Chemical Engineering Science 51, 1615–1635. https://doi.org/10.1016/0009-2509(96)00023-1

Quann, R.J., Jaffe, S.B., 1992. Structure-oriented lumping: describing the chemistry of complex hydrocarbon mixtures. Ind. Eng. Chem. Res. 31, 2483–2497. https://doi.org/10.1021/ie00011a013

Van de Vijver, R., Devocht, B.R., Van Geem, K.M., Thybaut, J.W., Marin, G.B., 2016. Challenges and opportunities for molecule-based management of chemical processes. Current Opinion in Chemical Engineering 13, 142–149. https://doi.org/10.1016/j.coche.2016.09.006