Reactive distillation processes used as unique operation or finishing stage: A comparison

Abstract

The main objective of this work is modeling and designing a reactive distillation unit with rigorous hydraulic constraints, determining the configuration at the same time. The RD unit is modeled using an equilibrium model (Taylor and Krishna, 2000) based on first principles (mass and energy balances and phase equilibrium relationships). Two different process schemes are proposed: one in which the reactive distillation unit represents the only reactive operation within the process and one in which reactive distillation is used as a finishing process, after a fixed bed adiabatic reactor stage. In both schemes, the reactive stages are modeled as trays with catalyst bags (Jones Jr., 1985). Two different optimization targets are evaluated: maximizing MTBE product purity and maximizing isobutylene global conversion to MTBE. The unit is sized according to rigorous hydraulic considerations (Stichlmair and Fair, 1998), while the number of reactive and non-reactive trays is estimated within an MINLP problem. Cost estimations are provided. Results for process variables are summarized for industrial size designs. Results show that lumping reaction and separation unit operations reduces the cost for the process equipment and also the energy consumption, still obtaining a similar process performance since product purity and reactant conversion are very similar to those of conventional processes.