Chemical Engineering Science, Vol.202, 347-356, 2019
Modeling of molybdenum transport and pressure drop increase in fixed bed reactors used for selective oxidation of methanol to formaldehyde using iron molybdate catalysts
A dynamic model was developed for a single reactor tube, in which methanol oxidation to formaldehyde over an iron molybdate/molybdenum oxide catalyst takes place simultaneously with transport of MoO3 from the catalyst through the reactor. A previously developed dynamic 1D mathematical model for a single ring-shaped cylindrical catalyst pellet, in which volatilization of MoO3 takes place, was implemented in the reactor model. Known axial profiles in a pilot scale reactor with respect to MeOH and H2O concentration and temperature were used as input to the model. MeOH forms volatile Mo-species with solid MoO3 in the catalyst pellets, which diffuses to the bulk gas phase and is transported through the reactor, leading to MoO3 depleted pellets. Volatilization of MoO3 from the pellets occur at the inlet of the reactor. As MeOH is oxidized down the reactor, the volatile Mo-species decomposes via the reverse reaction that formed them. Deposition of MoO3 downstream in the reactor decreases the void space between the catalyst pellets leading to increased pressure drop. The hydraulic diameter of the catalyst pellets and the porosity of the deposited MoO3 were fitted to experimental data obtained in a pilot plant unit containing a single reactor tube. Furthermore, the model was used to simulate a tube under industrial conditions for up to two years (feed composition 8.4% MeOH, 4% H2O, 10% O-2 in N-2, bed length = 100 cm and a temperature of 190-346 degrees C). Finally, two cases where catalyst pellets with no excess MoO3 or shaped as filled cylinders are used in the initial 21 cm of the catalyst bed were simulated. The simulations show that this significantly decreases the rate at which the pressure drop increases. This model is a first step towards a useful tool to predict MoO3 transport, pressure drop increase and estimation of process life time at varying reaction conditions. (C) 2019 Elsevier Ltd. All rights reserved.