A dynamic approach to the dimension reduction of chemical kinetic schemes

Abstract The use of both experimental and theoretical techniques has improved our understanding of chemical reaction mechanisms considerably in recent years. As a result, highly detailed chemical reaction mechanisms are being developed that incorporate large numbers of species and elementary reaction steps. However, even with modern high performance computers, it is still difficult to incorporate such large comprehensive schemes within commercial software, such as computational fluid dynamic codes used to simulate industrial processes, due to the computing burden involved. Much research is therefore currently focussed on the development of methods for reducing the dimension of comprehensive schemes whilst retaining the important behavioral features of the original scheme. In this paper, a dynamic approach based on a net rate of production analysis for scheme reduction is proposed. The success of the new approach is illustrated for a CO-H 2 oxidation reaction system in a continuously stirred tank reactor that exhibits complex oscillatory behaviour. The main advantage of the approach is that it does not depend on any training data set, and therefore does not suffer from domain restrictions that limit the applicability of many established reduction methods. Consequently, it can be implemented in any numerical simulation package.