Towards improved automation for desalination processes, Part I: Advanced control

Abstract

The desalination process industries have largely tried to solve control problems by conventional rather than modern control techniques that are attracting widespread industrial applications and revolutionizing process control as a result of advances in control theory. Control systems are designed from mathematical models that are generally imperfect descriptions of real processes. It is widely accepted that conventional PID or PI controllers may not provide satisfactory performance in view of non-linearity and uncertainty in the real world of desalination processes. It is essential that control systems operate satisfactorily over a wide range of process conditions. The practice of multi stage flash (MSF) desalination process control, process design, optimization, conspicuously lags behind contemporary theoretical developments. Conventional process control techniques do not ensure the safest and the most profitable operation of processes when compared with advanced process control which uses the knowledge of process economics and provides insight into the process to maintain essential variables at optimum setpoints. This paper is an attempt to promote excellence and innovation in desalination engineering and to remove barriers to advanced process control. Attempted here is brief clarification of some aspects of advanced control strategies in terms of answers to the questions: Advance control strategies: What are they? Where and when are they applicable? and What can be expected of them? Although the reference here is to MSF plants, some of the observations and remarks are also relevant to process control practice in general. In the end, the conclusions point out the complexity of the real world control problem which behoves us to advocate the application of techniques such as self-optimizing control strategies and artificial intelligence (AI) leading to an integrated system of intelligent control.