Reaction/distillation matrix algorithm development to cover sequences containing reactive dividing wall column: Validation in process of cumene production

Abstract

The Reactive Dividing Wall Column (RDWC) offers a powerful approach to process intensification by combining chemical reactions and multi-component separations within a single unit, delivering significant economic and operational benefits. This study investigates various RDWC configurations for Cumene production generated through a modified Reaction/Distillation matrix-based algorithm. Rigorous simulations of both the conventional Reactive Distillation Column (RDC) and RDWC setups were conducted in Aspen Plus, while the Teaching-Learning-Based Optimization (TLBO) algorithm in MATLAB minimized total annual costs (TAC) and met product specifications. TLBO was chosen for its parameter-free nature, computational efficiency, and strong optimization performance, making it a robust tool for this case study. Comparative analysis revealed that the optimized RDWC configuration substantially outperformed the conventional RDC, reducing TAC by 27.62 % and total energy consumption by 27.98 %. Furthermore, the optimized RDWC configuration led to a 57.66 % decrease in capital costs and a 29.13 % reduction in operating costs by minimizing the required number of columns, condensers, and reboilers. These results demonstrate that the RDWC is a highly efficient and economically viable option for energy-intensive processes such as Cumene production, achieving substantial cost savings and enhanced operational efficiency.