Optimal control of catalytic methanol conversion to formaldehyde

An optimal control methodology is applied to find the heat and oxygen flux profiles, distributed along the length of a plug flow reactor, for the conv ersion of methanol to formaldehyde. The calculations use models for the gas -phase and catalytic [MoO3-Fe-2(MoO4)(3)] reactions. The reactor designs sh ow that a distributed heat flux improves formaldehyde yields, but an oxygen flux does not affect the results. Formaldehyde mass fractions of over 90% have been achieved in the simulations. The solutions obtained, although not proven to be globally optimal, are of very high quality. A fully nonlinear robustness analysis of the formaldehyde production with respect to the cat alyst model variables is performed by the use of a high dimensional model r epresentation. This representation is similar to the ANOVA decomposition us ed in statistics but does not require an increase in the number of data poi nts as the dimensionality of the variable space increases. The most importa nt variables are the catalyst surface area and the rate of formaldehyde des orption. The yield improvement from employing optimized fluxes is found to be robust to the catalytic model parameter values.