On-line optimization system of pilot scale multi-effect batch distillation system

Abstract

The operation system, which generates a dynamic optimal operation procedure, is developed for a multi-effect batch distillation system (MEBAD) and is installed in a pilot scale MEBAD. It is said that in the MEBAD, the separation performance can significantly be increased by optimally operating the reflux flow rate as a function of time. However, because of the disturbances and modeling errors, the optimal operation profile calculated off-line is not always optimal when it is applied to the real plant. In the proposed system, by executing the optimization repeatedly within a batch of operation, the deviation from the optimal trajectory is minimized. The proposed system consists of four subsystems: In the composition measuring subsystem, the compositions of the products in all vessels are measured on-line by the near-infrared analyzer (NIR). In the composition estimation and model update subsystem, the unmeasurable compositions on all trays and tray efficiency are updated periodically so as to minimize the difference between the measured and estimated composition profiles of the products. In the optimization subsystem, the optimal composition profiles in the reboiler, the middle vessel, and the reflux drum are calculated so as to maximize the separation performance. The optimal profiles are sent to the control subsystem, and the process is controlled so that the product compositions follow the given trajectories. The proposed system is applied to the problem of separating ternary mixture. Using the pilot plant, it is experimentally verified that the energy consumption of the MEBAD can considerably be reduced when the optimal operation scheme is adopted.