(430h) Multiobjective Optimization of Forced Periodic Operation of Methanol Synthesis | AIChE

(430h) Multiobjective Optimization of Forced Periodic Operation of Methanol Synthesis


Seidel-Morgenstern, A. - Presenter, Max Planck Institute for Dynamics of Complex Technical Systems
Kienle, A., Otto von Guericke University Magdeburg
Nikolic, D., University of Belgrade
Leipold, J., tto von Guericke University Magdeburg
Methanol is produced in large amounts using heterogeneous catalysts e.g., Cu/ZnO/Al2O3 and synthesis gas under steady-state conditions. In some cases, performance improvements with regard to traditional process realization can be achieved by using a forced periodic input modulations of one or more inputs . Identifying operation conditions and suitable inputs for a forced periodic operation superior to the steady-state operation is quite difficult. A suitable approach is a combined application of the nonlinear frequency response (NFR) method and rigorous numerical optimization.
In this contribution, a theoretical study of the potential of this forced periodic operation regime for improving reactor performance for methanol synthesis is conducted considering an isothermal gradientless reactor. Two objectives are investigated: (I) methanol production rate and (II) yield based on carbon. The approach is based on a detailed kinetic model. The nonlinear frequency response analysis is able to identify the most promising inputs for a forced periodic regime, showing that simultaneous modulation of two eventually phase-shifted inputs often results in superior reactor performance.
This knowledge is used as starting point for rigorous numerical multiobjective optimization with additional constraints arising from practical reactor operation . Optimal steady-state operation is compared with the optimal forced periodic operation for two benchmark problems with and without inert nitrogen using a simultaneous optimization method. It is shown that in both cases, significant improvements are possible through forced periodic operation if the volumetric flow rate and the concentration of CO in the feed are modulated simultaneously with optimal frequency, phase shift, shape of the input functions, and amplitudes, which are all determined by the theoretical approach. This method is then applied to an isothermal and non-isothermal fixed bed reactor in a second step. The results reveal a significantly increased reactor performance for a more industrially relevant case of applying optimized forced periodic operation.