Comparison between differential evolution algorithms and response surface methodology in ethylene plant optimization based on an extended combined energy - exergy analysis

作     者：Jahromi, Farid Sadeghian Beheshti, Masoud Rajabi, Razieh Fereydon 

作者机构：Univ Isfahan Dept Chem Engn Esfahan Iran Univ Isfahan Proc Engn Inst Esfahan Iran Univ Isfahan Chem Engn Dept Esfahan Iran 

出 版 物：《ENERGY》 (能)

年 卷 期：2018年第164卷

页      面：1114-1134页

核心收录：

学科分类：0820[工学-石油与天然气工程] 08[工学] 0807[工学-动力工程及工程热物理] 

主　　题：Distillation column Exergy destruction level Target value Response surface methodology Differential evolution algorithm 

摘      要：This paper applies an extend energy - exergy analysis as a strategy for evaluating the performance of different types of distillation columns in ethylene production process by using industrial data. Regarding to the limitation and deficiency of energy exergy combination, a new method named exergy destruction level (EDL) and conceptual diagram based on equipment target value is proposed for process equipment with pressure and chemical composition changes. The effects of different operational parameters on the component separation are evaluated by sensitivity analysis. Eventually response surface methodology (RSM) and artificial intelligence (DE) method are developed for optimization of the chemical plant. Comparing obtained results from ethylene plant optimization by using RSM and DE, it was found that annual profit percentage with DE method is 61.6% more than RSM method. Also, results of optimization showed that the most effective operating parameters consist of feed stream temperature, boil-up ratio, reflux ratio, column pressure and feed stage. It was observed that utilities and refrigeration cycle consumption work have been declined significantly by using DE optimizer and EDL analysis (12.6% and 11.6%) compared to RSM optimizer and exergy analysis (11.9% and 4.8%). It is obvious that, EDL analysis and DE method expedite process optimization and provide more precise analysis than conventional energy-exergy-analysis. (C) 2018 Elsevier Ltd. All rights reserved.