Model predictive control of azeotropic dividing wall distillation column for separating furfural-water mixture

Xing Qian; Shengkun Jia; Kejin Huang; Haisheng Chen; Yang Yuan; Liang Zhang

doi:10.1016/j.isatra.2020.10.068

Model predictive control of azeotropic dividing wall distillation column for separating furfural-water mixture

ISA Trans. 2021 May:111:302-308. doi: 10.1016/j.isatra.2020.10.068. Epub 2020 Oct 28.

Authors

Xing Qian¹, Shengkun Jia², Kejin Huang³, Haisheng Chen¹, Yang Yuan¹, Liang Zhang¹

Affiliations

¹ College of Information Science and Technology, Beijing University of Chemical Technology, Beijing 100029, China.
² School of Chemical Engineering and Technology, Chemical Engineering Research Center, Collaborative Innovation Center of Chemical Science and Engineering (Tianjin), State Key Laboratory of Chemical Engineering, Tianjin University, Tianjin 300350, China. Electronic address: jiask@tju.edu.cn.
³ College of Information Science and Technology, Beijing University of Chemical Technology, Beijing 100029, China. Electronic address: huangkj@mail.buct.edu.cn.

PMID: 33158551
DOI: 10.1016/j.isatra.2020.10.068

Abstract

Dividing wall distillation columns (DWDCs) are effective technologies for distillation process intensifications. Azeotropic DWDCs (ADWDCs) are effective process intensification technologies to further intensify distillation process through combination of azeotropic distillation (AD) with DWDC. Investigations on the controllability and operation ability of ADWDC are mainly employing the traditional proportional-integral (PI) control schemes. Because of the inherent complicated configuration of ADWDC, the PI control leads to relatively poor dynamic performances. So as to enhance the dynamic performances, composition control scheme employing the advanced model predictive control (MPC) for an ADWDC separating furfural and water is studied in this paper. Although there are several studies employing MPC for DWDC, few investigation has been done on using MPC for ADWDC which is more interactive than DWDC. The dynamic results obtained in this paper using MPC are compared with the corresponding PI control scheme. Although the steady-state deviations and numbers of oscillations are similar employing PI control and MPC, the settling time and maximum deviations are smaller using MPC. MPC can considerably suppress the maximum deviation in face of +20% feed flow rate disturbance from -1.449% to -0.013%, and the corresponding settling time is reduced from 4.28 h to 2.20 h. These prove that MPC is more suitable than PI control for ADWDC. This paper provides new effective control scheme for the complex and highly interactive ADWDC, and verifies that MPC is a promising method for ADWDC to provide better dynamic performances with reduced maximum deviations and shortened settling time.

Keywords: Dividing wall distillation column; Model predictive control; Process control.