Energy-saving design and multi-objective optimisation of triple-column pressure-swing distillation configuration for separating n-propanol/benzene/water mixture with multiple azeotropes

Abstract

The separation of ternary azeotropic mixtures with multiple azeotropes is often encountered in industrial wastewater treatment. This study explores separating the n-propanol/benzene/water mixture with three binary azeotropes and one ternary azeotrope via pressure-swing distillation (PSD). Two conventional PSD schemes are designed based on the ternary phase diagram. The multi-objective genetic algorithm facilitates the optimisation task for each sequence, taking the total annual cost, CO2 emissions, and thermodynamic efficiency as the objective functions. Then, the selection of the optimal solution from the Pareto front is performed using the technique for order preference by similarity to the ideal solution. Subsequently, energy-saving methods, including direct heat integration, heat pump, and heat exchanger network, are introduced into each optimal non-heat-integrated PSD process to address the drawback of tremendous total energy consumption. Finally, the proposed 12 energy-efficient PSD designs are compared with the conventional PSD cases in multiple aspects. The optimisation results highlight the importance of operating pressures as variables in the multi-objective optimisation problem. In addition, the W-B-N sequence shows superiority over the W-N-B sequence in decreasing cost and gas emissions and enhancing energy efficiency. The proposed energy-saving PSD design with direct heat integration, heat pump, and heat exchanger network gives the lowest cost amongst the 14 processes, being up to a reduction of 39% compared to the corresponding conventional PSD design.