Plantwide control
Plantwide control refers to control structure design for complete chemical plants. This has been the main topic of my specialization course in advanced process control. This topic deals with the structural decisions of the control system, including what to control (controlled variables selection) and how to control (pairing these variables with operating valves and form control loops).
Although these are very important issues, these decisions are in most industrial cases made in an ad-hoc fashion, based on experience and engineering insight, and often without considering the details of each problem.
Actually, a chemical plant may have thousands of measurements and control loops. However, in practice, the control system is usually divided into several layers, separated by time scale.
These layers are linked by the controlled variables , whereby the setpoints are computed by the upper layer and implemented by the lower layer. As I mentioned in the beginning, an important issue is the selection of these variables.
My professor Sigurd Skogestad has developed a systematic procedure for plantwide control. The procedure is divided in two main parts:
(I) Top-down analysis to identify degrees of freedom and "economical" primary controlled variables. In this analysis, there is one step which suggests to look for self-optimizing variables. My research project uses a method based on this step called self-optimizing control.
(II) Bottom-up analysis to determine "stabilizing" secondary controlled variables and structure of control system (pairing with the valves or any kind of manipulated variables in the plant).
As our main example, distillation control has been largely covered during the lectures. However, my research project is about a heat exchanger network.
Reference : S.Skogestad, Control structure design for complete chemical plants, Department of Chemical Engineering, Norwegian University of Science and Technology (NTNU), Trondheim, Norway, published in Elsevier Ltd, 2003.