Feasibility and Process Alternatives for Reactive Distillation
S. B. Gadewar, N. Chadda, M. F. Malone, and M. F. Doherty 6.1
Reactive distillation (RD) has the demonstrated potential for capital productivity improvements (from enhanced overall rates, by overcoming very low reaction equilibrium constants, and by avoiding or eliminating difficult separations), selectivity improvements (which reduce excess raw material use and by-product formation), reduced energy use, and the reduction or elimination of solvents. Some of these advantages are realized by using reaction to improve separation, such as overcoming azeotropes or reacting away contaminants; others are realized by using separation to improve reactions, for example, overcoming reaction equilibrium limitations, improving selectivity, or removing catalyst poisons. The potential is greatest when both aspects are important. This technology is potentially attractive whenever there is a liquid phase reaction that involves an excess of reactant. Many existing, and potential, applications are discussed by Sharma and Mahajani .
Some of the successes of RD are so dramatic that we might ask if all liquid-phase chemical processes should be based on simultaneous reaction and separation instead of more traditional separate steps. The answer is no, because combining reaction and distillation is not always advantageous; in some cases it may not even be feasible! A key question is 'How can we decide quickly whether RD is a good process concept?' This question is addressed mainly by studies in conceptual design, which is the major focus of this chapter.
For conceptual design of RD, systematic methods are needed for deciding its feasibility. We use geometric methods which have been used extensively to assess the feasibility for non-RD. For non-RD, a number of algorithms can be used to find the feasible splits. Stichlmair and Herguijuela  calculate the products at total reflux while Wahnschafft et al.  and Fidkowski et al.  use pinch tracking techniques to estimate the feasible products for finite as well as total reflux. These methods rely on geometric visualization, so have been applied to ternary mixtures. Algorithms building on these ideas, but that do not rely on visualization are described by Safrit and Westerberg  and Rooks et al. . These methods can also be applied in the limit of reaction equilibrium using transformations of the compositions that have properties similar to those of mole fractions in non-reactive mixtures [1, 9, 46]. There is a one-to-one correspondence between the real mole fractions and the transformed variables at chemical equilibrium.
Such one-to-one transformations are not known for kinetically controlled RD although some aspects of these transforms remain useful in the kinetic regime. Studies at finite rates of reaction are important because most RD devices operate in this regime. There are several approaches available that capture some of the effects of finite reaction rates, such as, geometric design methods [2, 3, 34]; mixed integer nonlinear programming (MINLP) methods [6, 28, 36]; attainable region methods [32, 33]; residue curve/bifurcation methods [37, 45, 47, 48]; and difference point methods . However, there are few tools that assess the feasibility of reactive mixtures in the kinetic regime. Giessler et al. use static analysis to determine feasibility of reactive columns operated with large internal flows [17, 18]. Chadda et al. generate feasible product regions at finite rates of reaction for ternary systems . However, that method cannot be extended to treat a larger number of components or multiple reactions.
In this chapter, we describe an algorithm for predicting feasible splits for continuous single-feed RD that is not limited by the number of reactions or components. The method described here uses minimal information to determine the feasibility of reactive columns: phase equilibrium between the components in the mixture, a reaction rate model, and feed state specification. This is based on a bifurcation analysis of the fixed points for a co-current flash cascade model. Unstable nodes ('light species') and stable nodes ('heavy species') in the flash cascade model are candidate distillate and bottom products, respectively, from a RD column. Therefore, we focus our attention on those splits that are equivalent to the 'direct' and 'indirect' sharp splits in non-RD. One of the products in these sharp splits will be a pure component, an azeotrope, or a kinetic pinch point; the other product will be in material balance with the first. The proposed algorithm is based on the following.
• A bifurcation study to predict the distillate and bottoms products for the entire range of reaction rates from the limit of no reaction to the limit of chemical equilibrium. This provides a global view of the direct and indirect split products from a continuous RD at all rates of reaction.
• Flash calculations and the application of the lever rule (overall mass balance relating the feed, distillate and bottoms product streams) to predict feasible sharp splits for a given feed condition.
We begin by describing a simple example that demonstrates the advantages gained by combining reaction and separation. These advantages are represented in terms of the attainable region.
Was this article helpful?