CIESC Journal ›› 2022, Vol. 73 ›› Issue (3): 1246-1255.DOI: 10.11949/0438-1157.20211308

• Process system engineering • Previous Articles     Next Articles

Pretreatment process simulation and multi-objective optimization of C5 by reactive dividing wall column

Xiaoqing SHI(),Weixuan ZHU,Haotian YE,Zhizhong HAN,Hongguang DONG()   

  1. School of Chemical Engineering, Dalian University of Technology, Dalian 116000, Liaoning, China
  • Received:2021-09-07 Revised:2021-11-26 Online:2022-03-14 Published:2022-03-15
  • Contact: Hongguang DONG

碳五隔壁反应精馏预处理工艺模拟及多目标优化

石晓青(),朱炜玄,叶昊天,韩志忠,董宏光()   

  1. 大连理工大学化工学院,辽宁 大连 116000
  • 通讯作者: 董宏光
  • 作者简介:石晓青(1996—),男,硕士研究生,1933911099@mail.dlut.edu.cn
  • 基金资助:
    国家自然科学基金项目(21276039)

Abstract:

In view of the high energy consumption in the pretreatment section of C5 comprehensive utilization, this paper proposes the reactive dividing wall column (RDWC) pretreatment process based on the reactive distillation (RD) pretreatment process. First of all, the chemical simulation software Aspen Plus was used to build the RDWC four tower equivalent rigorous model, and the degree of freedom and univariate analysis were carried out. On this basis, the response surface Box-Behnken Design (BBD) method is used as the model fitting tool to fit the functional relationship between target variables and decision variables, and the fitting results are analyzed by ANOVA. Finally, multi-objective evolutionary algorithm based on decomposition (MOEA/D) is used to optimize the RDWC pretreatment process, and a series of Pareto optimal solutions are obtained. The solution with the smallest TAC is selected and compared with the RD pretreatment process. The results show that compared with RD pretreatment process, RDWC pretreatment process can save TAC 12.8%, save reboiler load 27.8%, and improve selectivity.

Key words: reactive distillation, reactive dividing wall column, simulation, response surface analysis, multi-objective algorithm, optimal design

摘要:

针对碳五综合利用预处理工段能耗较高的现状,在反应精馏(RD)预处理流程基础上提出了隔壁反应精馏(RDWC)预处理流程。首先,利用化工模拟软件Aspen Plus搭建RDWC四塔等效严格模型,并对其进行自由度和单变量分析。在此基础上,以响应面Box-Behnken Design(BBD)方法作为模型拟合工具,拟合出目标变量与决策变量之间的函数关系,并对拟合结果进行方差分析(ANOVA)。最后采用基于分解的多目标进化算法(MOEA/D)对RDWC预处理流程进行多目标优化,得到一系列Pareto最优解,选出其中年总成本(TAC)最小的一组解与RD预处理流程进行对比。结果显示:与RD预处理流程相比,RDWC预处理流程可以节约TAC 12.8%,节省再沸器负荷27.8%,选择性也有所提高。

关键词: 反应精馏, 隔壁反应精馏塔, 模拟, 响应面分析, 多目标算法, 优化设计

CLC Number: