Closed-loop dynamic real-time optimization (CL-DRTO) of a bioethanol distillation process using an advanced multilayer control architecture. (5th December 2020)

• Record Type: Journal Article
• Authors: Pataro, Igor M.L.; da Costa, Marcus V. Americano; Joseph, Babu
• Is Part Of: Computers & chemical engineering. Volume 143(2020)
• Publication Date: 2020-12-05
• Subjects: Ethanol -- Distillation -- Dynamic optimization -- Software in the loop simulation -- Predictive control -- Robustness; Chemical engineering -- Data processing -- Periodicals; 660.0285
• DOI: 10.1016/j.compchemeng.2020.107075
• Languages: English
• ISSNs: 0098-1354

Abstract

Abstract: The ethanol production has become an important part of the worldwide economy driven by its use as renewable energy and environmentally clean fuel. To obtain a high-level quality product, a large quantity of energy is used in the distillation stage. This work proposes to optimize the bioethanol production process applying a closed-loop dynamic real-time optimization (CL-DRTO) framework associated with advanced control strategies in the ethanol distillation facilities to improve production and minimize energy losses. A high-fidelity computational platform is developed using a software-in-the-loop (SIL) technique to demonstrate the feasible application in real cases. OLE (Object Linking and Embedding) Automation integrates Matlab and Aspen Hysys software to simulate practical scenarios currently applied in the industry. The optimization and control algorithms are developed in Matlab and the ethanol distillation process is modeled in the Aspen Hysys. Different advanced control strategies, such as IHMPC (Infinite Horizon Model Predictive Control), MIMO FSP (Filtered Smith-Predictor) and DTCGPC (Dead-Time Compensator Generalized Predictive Controller), are used to overcome the process dynamic issues, for instance, strong nonlinearity, dead-times, long time constants and coupled loops, on ethanol distillation process. A CL-DRTO layer is developed in the proposed control structure to consider economic and production aspects, guiding the process to optimal operating points. Performance indices and computational effort are used to evaluate the control behavior for disturbance rejection and reference tracking scenarios. Furthermore, economic performance is analyzed to ensure the advantages of the proposal. Results show that the proposed computational platform is able to reproduce industrial scenarios with high-fidelity. In addition, preliminary results demonstrate that advanced control structure can improve the production and profitability of the bioethanol distillery and present a powerful alternative to replace classical controllers currently used in this type of industry.