Abstract

An industrial steam reformer of a methanol plant was modeled at a dynamic condition in which a one dimensional homogeneous model was coupled with a verified kinetics from the literature. A close agreement was observed between the results of the model and industrial data from a real plant at steady state conditions. The open loop response of the system to switching between two operating conditions was investigated and shown that the produced synthesis gas during the transition period would be unsuitable for the downstream methanol converter. The genetic algorithm was then employed to perform a multi-objective dynamic optimization on the reactor performance in case of switching the feed and operating conditions. Maximization of methane conversion and minimization of a stoichiometric parameter, were considered as the two objectives' functions that were optimized for a fixed feed rate of methane to the existing unit. The results of the dynamic optimization for the specified reformer configuration were achieved after switching the operating condition. Results of the optimization showed that the produced synthesis gas would stay in its acceptable limits in terms of quality of the feed of the methanol converter and also, the final conversion of the reformer would be improved compared to the steady state condition. This procedure could be applied to the advanced process control of the methanol plant.

Recommended Citation

Alizadeh, Ali; Mostoufi, Navid; and Jalali-Farahani, Farhang (2007) "Multiobjective Dynamic Optimization of an Industrial Steam Reformer with Genetic Algorithms," International Journal of Chemical Reactor Engineering: Vol. 5: A19.
Available at: http://www.bepress.com/ijcre/vol5/A19