Oxidation of Methanol to Formaldehyde over Ag - Kinetic Modelling of Gas Phase Reactions using COMSOL Multiphysics

Abstract

In this master thesis the goal was to create two COMSOL Multiphysics simulation models, one representing the laboratory scaled reactor and the other representing the industrial scale reactor, to study gas phase reactions in the methanol to formaldehyde (MTF) over silver process. Gas phase reactions have been suspected to occur up front of the catalyst bed in laboratory experiments as conversion of methanol to formaldehyde abated as temperatures increased. Investigations into this started by Dynea showed that they may play a bigger role than first believed as gas phase reactions will compete for the feed oxygen with the selective reactions on the catalyst and lead to lower product yield.

The simulation models were designed as a homogeneous chemical reaction systems without space dimensions and catalyst. Mass and energy balances were given by a batch reactor study adapted to operate under isothermal conditions and with constant pressure. A reaction data set, GriMech, containing 325 reactions and 53 chemical species was loaded into the model and the development of the reactions were studied over time. The simulation models were used to study the effect of varying temperature and feed composition on the response time of gas phase reactions and resulting composition. Studies were also conducted into the individual components at 650°degree C to try to pin-point the type of gas phase reaction occurring. The GriMech reaction used for studies was originally designed for the combustion of natural gas and was quite extensive so in addition to the studies of gas phase reactions it was desirable to both evaluate the GriMech performance and adapt the mechanism to fit the MTF process.

Simulation studies showed that gas phase reactions occurred in both models and that the MTF species accounted for >99% of the species at steady state. Results of response times studies showed that temperature was the biggest factor to consider since increasing the temperature from 500°C to 700°C decreased the response time from 30.2 seconds to 0.11 seconds and no feed variations could compare with this effect. However, results showed that the presence of oxygen or formaldehyde, even at low concentrations, in the feed also catalyzed the gas phase reactions and decreased the response time greatly. Increasing the concentration of these components in the feed did not further decrease the response time. None of the other feed components studied showed the same effect.

From the studies of effect of feed variations and the individual species results indicated that at low feed oxygen concentrations both methanol and formaldehyde undergo decomposition reactions forming carbon monoxide and hydrogen. The presence of oxygen in the feed catalyzed the decomposition but results suggested that the feed oxygen was spent in a secondary combustion reaction of hydrogen to form water. As the feed oxygen concentration increased results were not conclusive as to weather decomposition or combustion occurred, but indicated that decomposition and secondary combustion reactions may be the dominant reactions.

Response times of the individual species studies indicated that oxygen would be spent in formaldehyde and then hydrogen reactions before reacting with methanol and that carbon monoxide would not undergo combustion to carbon dioxide at 650°degree. For the carbon monoxide studies however the results are not conclusive as both the feed variation studies and the individual study of formaldehyde indicated that carbon monoxide did occur to form carbon dioxide.

The GriMech reaction mechanism was adapted to the MTF process by removing nitrogen species and hydrocarbon species with more than one carbon, effectively reducing it from 325 reactions and 53 species to 125 reactions and 25 species. Simulation studies using the new mechanisms gave identical results indicating that the adaption was successful. However, it was discovered that the creators of GriMech strongly discourage removal of species from the mechanism as it can significantly effect results. Further use of the mechanism must be carefully considered.