Design and Optimization of Pressure Swing Distillation using a Stochastic Algorithm based in the Boltzmann Distribution

Abstract

The separation of azeotropic mixtures with help of distillation is an important unit operation in chemical and pharmaceutical industry. The literature focuses on different alternatives for the separation of azeotropic mixtures, such as extractive distillation, azeotropic distillation, heterogeneous azeotropic distillation, vacuum distillation and the Pressure Swing Distillation (PSD). This work describes an approach for PSD in continuous flow for binary azeotropic mixture separation sensible to the pressure changes. The main advantages of this process compared to the others are: i) no additional substances (entrainer) have to be used, and ii) for the continuous flow operation heat integration is possible and it can save energy. The PSD has been optimized with a novel stochastic algorithm called Boltzmann Univariate Marginal Distribution Algorithm (BUMDA). The performance of BUMDA is robust and highly efficient as shown by the experiments conducted (although it does not guarantee optimality). This work makes two specific contributions: 1) the optimization of the PSD in continuous for a binary azeotropic mixture; 2) the application of the BUMDA stochastic algorithm, and the constraint handling technique.