(409d) An Automated Feasibility Evaluation Algorithm for Reactive Distillation

Reactive distillation offers potential reductions in the capital costs, operating costs, process complexity, and environmental impact of chemical processes because it allows reaction and distillation to overcome each other's limitations [1]. This paper discusses the feasibility criteria of batch and continuous reactive distillation and presents an automated algorithm to evaluate these criteria for any given single-reaction system. The feasibility criteria[2-4] are for columns with a reaction of the form aK+bI cL+dH or bI cL +dH: 1) A batch reactive rectifier can produce an unstable node (UN) product if it's reachable from reaction equilibrium. A batch reactive stripper can produce a reachable stable node (SN) product. 2) A batch middle-vessel, or continuous single-feed, reactive distillation column can produce UN and SN products if they are reachable from reaction equilibrium. 3) A batch reactive rectifier with a decanter can produce an almost-pure product if a reachable UN azeotrope splits to a product. 4) Continuous or batch reactive extractive distillation can produce pure product if a homogenous entrainer allows extractive section profiles to connect the reaction equilibrium manifold to an entrainer-product binary edge, 5) If the above columns don't work, then a heterogenous entrainer in continuous or batch reactive extractive distillation with decantation is needed.

The automated algorithm for evaluating these criteria requires only reaction and phase equilibrium information. The algorithm finds all of the azeotropes in the system by an isovolatility curve search [5] and determines their dynamic properties [6]. It then applies the algorithm of Rook et al.[7] to find the distillation boundaries and distillation regions. Then, the algorithm calculates which distillation regions the reaction equilibrium manifold lies in and determines if any of the simple column configurations (rectifier, stripper, MVC, continuous single-feed) can produce pure products. If the simple columns are not feasible, then the algorithm will calculate and plot the critical composition region and the regions where upper bounds are greater than lower bounds for various entrainer flow rates. Based on these plots, the user can decide if the system is feasible and what entrainer flow rate is required. Less computation-intensive tests for batch columns can be performed by predicting the still pot trajectory from an arbitrary feed composition under an assumption of constant product composition. This assumption gives the still pot trajectory as the intersection between a linear variation and reaction equilibrium. If residue curves connect every point of the still pot trajectory to pure products (or a heterogenous azeotrope that splits to a product), then the assumption is correct and pure products can be feasibly produced from the given feed composition.

References

1. Malone MF, Doherty MF. Reactive distillation. Ind. Eng. Chem. Res. 2000; 39: 3953-3957.

2. Guo Z, Ghufran M, Lee JW. Feasible products in batch reactive distillation. AIChE J. 2003; 49: 3161-3172.

3. Guo Z, Chin J, Lee JW. Feasibility of continuous reactive distillation with azeotropic mixtures. Ind. Eng. Chem. Res. 2004; 43: 3758-3769.

4. Chin J, Choe J, Lee JW. Feasible products in complex batch reactive distillation. AIChE Journal, in press (2006).

5. Westerberg AW, Wahnschafft O. The Synthesis of Distillation-Based Separation Systems. Advances in Chemical Engineering, Volume 23 p64-171 Academic Press Inc, San Diego 1996.

6. Fidkowski ZT, Malone MF, Doherty MF. Computing Azeotropes in Multicomponent Mixtures. Cumput. Chem Eng. 1993; 17: 1141

7. Rooks RE, Julka V, Doherty MF, Malone MF. Structure of Distillation Regions for Multicomponent Azeotropic Mixtures. AIChE Journal 1998; 44: 1382-1391