(203b) Semicontinuous Reactive Distillation for Specialty Chemical Production: Economic Comparison with Batch and Continuous Processing

Semicontinuous reactive distillation (SRD) is introduced as a novel design technique for performing separations and reactions together. This process builds upon the forced-cyclic principles of semicontinuous distillation (SD), and tightly integrates the reaction step with the distillation process as in reactive distillation (RD). The controlled, periodic interaction of a reaction in an auxiliary vessel with distillation allows multi-purpose usage of the column during alternating phases of the cyclic campaign. This is accomplished without the startup, shutdown, or cleaning stages in batch processes, saving both time and energy. Moreover, the SRD design may be optimal for larger production rates that are too inefficient to achieve in batch processes. Additionally, the semicontinuous method requires fewer columns when compared to continuous systems, and operates at lower production rates that may be less profitable in continuous processes. Thus, SRD performs reaction-separation operations at intermediate production rates, having its greatest potential in the fine and specialty chemical industries.

An SRD case study is presented involving the reversible reaction of acetaldehyde and propylene glycol to form 2,4-dimethyl-1,3-dioxolane (DMD) and water. The two products form a low-boiling azeotrope, with the products and azeotrope having intermediate boiling points. Moreover, DMD and water are close-boilers (with an azeotrope), reaction temperatures cannot be controlled, and the reaction is reversible, so ordinary reactive distillation is difficult to operate efficiently. Thus, the SRD system is shown to be a viable candidate.

Simulation results show that SRD creates nearly-pure reaction products for this difficult reaction system, with minimal process equipment and without excess reagents to drive the reaction. Economic optimization studies over a range of operating parameters indicate that the ideal SRD production rate is close to the heuristic production-rate boundary that separates continuous and batch processes. Further economic comparisons with continuous and batch designs at the same production rates indicate when the designer should choose one method over the other two. A generalized algorithm for the design of a SRD process for other reaction systems is also presented.