Info

Introduction

Reactive separations are widespread operations. In typical reactive separation processes such as reactive absorption or distillation the superposition of reaction and separation is deliberately used. In other cases, simultaneous reaction and separation simply cannot be avoided. This is, for instance, the case when side reactions occur in separation equipment or when intrinsically chemically reactive mixtures, such as solutions of weak electrolytes or formaldehyde solutions, have to be separated. Furthermore, in many reactors products are directly removed, which is basically a reactive separation.

Thermodynamics plays a key role in understanding and designing all these processes. The fact that reaction and separation occur simultaneously gives rise to special challenges both in experimental investigations and modeling the processes. It is these challenges that we will focus on in the present chapter. This will be done using case studies, which illustrate general facts, rather than by trying to cover the subject comprehensively.

There are several contributions of thermodynamics to the field of reactive separations. Thermodynamics provides the basic relations, such as energy balances of equilibrium conditions, used in the process models, which again are the key to reactive separation design. Furthermore, thermodynamics provides models and experimental methods for the investigations of the properties of the reacting fluid that have to be known for any successful process design. We will focus on equilibrium thermodynamics here but discuss relations to kinetic models.

The present chapter is aimed at readers interested in reactive distillation (RD). The basic ideas can, however, easily be applied to other reactive separation processes.

Process Models for Reactive Distillation

4.2.1 Outline

Process models for RD have to take into account both the chemical and the physical side of the process. Two basic types of model are used: stage models, which are based on the idea of the equilibrium stage with phase equilibrium between the outlet streams, and rate-based models, which explicitly take into account heat and mass transfer. Similarly to the physical side of RD, the chemical reaction is either modeled using the assumption of chemical equilibrium or reaction kinetics are taken into account. Note that a kinetic model, either for physical transport processes or for chemical reactions, always includes an equilibrium model. The equilibrium model is the stationary solution of the kinetic model, for which all derivatives with respect to time become zero. Hence, whatever model type is used, it has to be based on a sound knowledge of the chemical and phase equilibrium, which is supplied by thermodynamic methods. Starting from there, kinetic effects can be included.

Only a limited amount of data is needed to develop models for RD processes if the assumption of both phase and chemical equilibrium is used. Nevertheless, even the development of a reliable equilibrium model alone is a challenging task, which is often underestimated. The amount of information needed to develop reliable kinetic models greatly exceeds that for the equilibrium models. Finding reliable model parameters is often the bottleneck in model development.

An excellent comprehensive survey on the fundamentals of different types of model for RD processes has recently been given by Taylor and Krishna [1]. In that paper the focus is on modeling concepts, we focus here on the application of such models, especially the comparison of model predictions with experimental data and the background of the complexity of the model and the effort needed for its parameterization. We do account for the fact that the results of any such comparison will depend on the chosen example, but emphasize that comparison with experimental data, especially in predictions, is the final test for any modeling strategy. Unfortunately, there is only a very limited amount of data on RD experiments available in the open literature for such comparisons.

Was this article helpful?

0 0

Post a comment