Catalyst Design and Analysis by CFD - 1

he steam reforming of methane to produce synthesis gas is a particularly important reaction, due to the increased availability of natural gas as a feedstock. The use of reformed gas has historically been dominated by hydrogen manufacture for commercial use, or for ammonia and methanol synthesis. Of increasing importance, however, is the manufacture of liquid fuels from remote or stranded natural gas using Fischer Tropsch chemistry, and the generation of hydrogen from natural gas liquid fuels to power mobile and stationary fuel cells. The syngas generation section of such plants comprises over 50% of the capital cost. There is thus a strong economic incentive to develop more efficient steam reforming technology, and a major step in this effort is the optimal design of catalyst particles to achieve a balance between the demands of high catalyst activity, low pressure drop in the tube, and high heat transfer rates.

This part of our research program aims to use the ability of CFD to handle complex geometries and to simulate realistic flow fields to obtain details of the gas temperature and partial pressures surrounding the catalyst pellets. This will let us examine the effects of design changes in the catalyst pellets. To date, however, commercial CFD codes have not been developed to include reaction inside solid catalysts. To get an idea of how catalyst pellets behave in low-N fixed beds when reaction is present, we decided to extend our previous simulations in wall segments to include heat sinks/sources inside the catalyst pellets. Kinetics and heats of reaction were included in the model by user-defined code.

In our first studies simulating the energetic effects of reaction, CFD simulations with temperature-dependent heat sinks inside spherical particles were performed [Dixon, A.G., Nijemeisland, M. and Stitt, E.H., 2003]. The heat sinks were selected to mimic the endothermic methane steam reforming process. Species partial pressures were fixed in the short wall segment so that reaction rates became functions of temperature only. For spherical catalyst particles, it was demonstrated that for this reaction under particular tube conditions, the particles adjacent to the reactor tube wall did not show symmetric temperature fields, contrary to the usual assumptions of reaction engineering models. Examples of these simulations are presented in the two figures, click on them for more details if you are interested.