"Modelling of large reaction network through stochastic methods: application to oligomerization"
Advancing the chemical engineering fundamentals
Chemical Reaction Engineering: Kinetics & Modelling (T2-2a)
Keywords: sotchastic kinetc, oligomerization, single event
"Modelling of large reaction network through stochastic methods: application to oligomerization"
J. R. Shahrouzi1, D. Guillaume1, P. Galtier1, P. Rouchon2
1Institut Français du Pétrole, BP3 69390, Vernaison, France
2Ecole des mines de Paris, 60 Bd Saint-Michel 75272 Paris cédex 06
Abstract
The kinetic modeling of complex reaction networks, which involve thousands of species, intermediates and elementary steps, is of great industrial importance. Such a complex processes may contain an excessive number of kinetic equations and rate parameters so that there is a common need to identify techniques for simplifying these gigantic networks. The classical modeling approach applies the deterministic solution for a reduced network consisting of a small number of reactions between pseudo-components or lumped species.
Nowadays, the reaction network can be constructed in terms of elementary steps of cation chemistry using a computer algorithm. The single event concept explicits the effect of structure on the entropy contribution of rate coefficient and reduces the rate parameters to tractable numbers [1]. Deterministic solution for single event rate parameters needs a complete construction of network for all possible reactions which is unrealistic due to the storage capacity and memory limitation of computers, specially when the number of carbon atoms rises [2].
In this work, the stochastic simulation approach is applied for oligomerization of light olefins to heavier products instead of usual deterministic approach. The stochastic kinetics is a probabilistic approach for the time evolution of spatially homogenous systems [3]. The interest of this approach for our application is not to generate the complete network of all the possible reactions but to identify the population and evolution the whole of the possible reactions at the present moment. The whole of the possible reactions is gigantic, but the whole of the elementary types of reactions, i.e., isomerization, cracking and oligomerization, are limited and it can be seen that in each step generation of all reactions of oligomerization is possible. In this point of view, we have calculated the population and conversion of species up to C16 which can be generalized for higher carbon numbers. It is shown that the stochastic simulation could be an advisable approach for simulation of the oligomerization network. On the other hand, a certain number of parameters of simulation are difficult to estimate, duration of simulations is relatively high and adequate simulation runs are required to obtain a stable average.
References
[1] Gilbert F.Froment, "Kinetic modeling of acid-catalyzed oil refining processes," Catalysis Today, vol. 52, no. 2-3, pp. 153-163, Sept.1999.
[2] Denis Guillaume, "Network Generation of Oligomerization Reactions: Principles," Ind. Eng. Chem. Res., vol. 45, pp. 4554-4557, 2006.
[3] Daniel T.Gillespie, "A general method for numerically simulating the stochastic time evolution of coupled chemical reactions," J. Comp. Phys., vol. 22, pp. 403-434, 1976.
Presented Monday 17, 11:15 to 11:33, in session Chemical Reaction Engineering: Kinetics & Modelling (T2-2a).
