Implementation of advanced operational and control techniques for a pilot distillation column within a dcs environment

A well-instrumented pilot-scale continuous distillation column has been interfaced with an industrial distributed control system (DCS-ABB MOD 300) and a VAX cluster through the network. The column separates ethanol and water and the set-up is used to implement different advanced control and operational techniques. An ideal supervisory scheme was proposed and accordingly gross error detection, identification and estimation, data reconciliation, simulation, efficiency modification and optimisation modules were implemented to provide the grounds for advanced operation of chemical processes. it is shown that these techniques play a vital role in control and optimisation of a given process. Furthermore, a novel robust multivariable, low order, high performance, model-based controller was designed and implemented as a standard PID block within the DCS. Finally, the answer to an important question of how close to the optimum can an operator operate without fear of violating the constraints in the face of disturbances is discussed; this led to a semi-online optimising control program acting as another informative module that completes a unified framework. The paper is accompanied by some experimental evidence which demonstrate the importance of each implemented module individually and as a unified set-up. (C) 1998 Elsevier Science Ltd. All rights reserved.