PRACTICE OF ADVANCED PROCESS CONTROL

Abstract

In recent years, Advanced Process Control (APC) is gaining popularity in Asia. Reliable Distributed Control Systems (DCS) have been widely established to provide safe, stable and profitable operations for process industries. With increasing computing and communication capabilities of DCS, many advanced control techniques can now be implemented in DCS directly. The main motivations of implementing APC in DCS are two folds. Firstly, it allows APC to be resident in a highly reliable DCS platform. Secondly, applying APC in DCS greatly reduces software maintenance time. Industrial APC system consists of three main layers, Enhanced Regulatory Control (ERC), Multi-variable Control (MVC) and Real Time Optimization (RTO). In this thesis, a general APC application methodology is proposed. This provides a simple model for control engineers in the industry to develop an industrial APC system. The main steps of this methodology are as follows: 1) Process Analysis and Process Knowledge Acquirement 2) APC Benefit Estimation 3) Process Modeling and Controller Design 4) Installation and Commissioning of Controller 5) Training 6) APC Performance Audit The APC methodology model is illustrated in the author's representative APC practices on several industrial processes. These practices include enhanced regulatory control strategies for heating furnace pass balancing control, surge tank level control, and design and development of a complete APC system for air separation plant. The results of these applications showed that great benefits could be achieved by implementing advanced control logic in the standard DCS hardware. Most of the existing multivariable controllers are very expensive. This thesis has proposed a low cost multivariable de-coupling control strategy that is suitable for simple distillation units.