Heterogeneous and pseudohomogeneous two-dimensional models are compared to steady state and dynamic experimental data from a packed bed reactor for the partial oxidation of methanol to formaldehyde over an iron oxide-molybdenum oxide catalyst. Highly effective parameter estimation software was used to fit selected model parameters to large sets of experimental data so as to obtain small residuals. Heat transfer parameters which were successful in matching data from experiments without reaction were not capable of fitting data from experiments with reaction, and it was necessary to increase the radial heat transfer for higher temperatures or reaction rates. Axial composition profile data was represented by estimating the preexponential factors and activation energy in a half-order redox rate expression for methanol oxidation. After some decline in catalyst activity, a time-varying axial catalyst activity profile was determined from the data. A redox-type rate expression for the oxidation of formaldehyde to carbon monoxide was proposed to fit the data. The dynamics of the reactor temperature profile were accurately represented by the model. The heterogeneous and pseudohomogeneous models gave similar results in fitting experimental data, although the parameters determined for the two models were somewhat different.

No abstract

No abstract

No abstract

No abstract