Modelling Reaction Kinetics — an Interdisciplinary Challenge for Mathematics

Abstract

Chemical process modeling is of increasing importance in the chemical industry. Most chemical processes can be split into a reaction part, where the educt species react to the product, species, and a separation part, where the product is purified and the reaction byproducts are eliminated. The separation part, for example a series of distillation columns, is often much better understood than the reaction part, such that process modeling is already widely accepted and successfully applied in optimizing separation processes. There are, however, hardly any quanitative models available for the reaction part, and therefore this part is often the bottleneck for a process model integrating all the relevant process parts into a global model. This is caused by two reasons. First, most, chemical reactions are nonlinear, such that the well understood linear modeling techniques often fail. Second, data acquisition is very expensive, if concentrations of chemical species have to be determined, such that there is always a lack of data even for a purely black-box modeling or for a parameter fitting in models containing a lot of parameters.