Plant-Wide Control Issues

Plant-wide control is concerned with designing control systems for a large number of individual process units that may be highly interacting. A typical plant-wide control system will consist of many single-loop controllers as well as multi-variable controllers such as Model Predictive Control (MPC),1 10 and may involve thousands of measurements, hundreds to thousands of manipulated variables and hundreds of disturbance variables. Fortunately, a plant with a large number of processing units can be analysed as smaller clusters of units. 

It should be noted that interactions between control loops is not just limited to interacting units but will also occur within single units. A typical example is an exothermic reactor where a change in the control loop which controls the level in the reactor will have an effect on the amount of material in the reactor. This in turn will affect the heat removal requirements and, therefore, the cooling water control loop. Many strategies for reducing loop interactions, and for selecting control loops so as to minimise interactions, can be found in most standard control textbooks.1-11 

An exothermic reaction A - B is taking place in a CSTR which has a cooling jacket with cooling water. The input stream is coming from an upstream unit. The main disturbances, controlled variables and manipulated variables are  

In order to ensure safe operation and a satisfactory quality of the reactor product, the outlet concentration CB is to be controlled. Due to possible side-reactions, the reactor temperature T must also be controlled. The temperature can be measured easily but a measurement of the concentration of the product B is only available from a laboratory every hour. 

Propose a control system based on feed-forward - feedback control, cascade control and inferential control to achieve these control objectives. 

---

While the development of flue gas clean-up processes has been progressing for many years, a satisfactory process is not yet available. Lime/limestone wet flue gas desulfurization (FGD) scrubber is the most widely used process in the utility industry at present, owing to the fact that it is the most technically developed and generally the most economically attractive. In spite of this, it is expensive and accounts for about 25-35% of the capital and operating costs of a power plant. Techniques for the post combustion control of nitrogen oxides emissions have not been developed as extensively as those for control of sulfur dioxide emissions. Several approaches have been proposed. Among these, ammonia-based selective catalytic reduction (SCR) has received the most attention. But, SCR may not be suitable for U.S. coal-fired power plants because of reliability concerns and other unresolved technical issues (1). These include uncertain catalyst life, water disposal requirements, and the effects of ammonia by-products on plant components downstream from the reactor. The sensitivity of SCR processes to the cost of NH3 is also the subject of some concern. 

Obviously, from a purely mathematical point of view, it would be optimal to use a centralized on-line optimizing controller with continuous update of its model parameters and continuous reoptimization of aU variables. However, for a number of reasons, we almost always decompose the control system into several layers, which in a chemical plant typically include scheduling (weeks), site-wide optimization (day), local optimization (hour), supervisory/predictive control (minutes) and regulatory control (seconds). Therefore, we instead consider the implementation shown in Figure 1 with separate optimization and control layers. The two layers interact through the controlled variables c, whereby the optimizer computes their optimal setpoints (typically, updating them about every hour), and the control layer attempts to implement them in practice, i.e. to get c Cj. The main issue considered in this chapter is then What variables c should we control ... 

Traditional methods of additive analysis and the required instruments are often expensive and require the efforts of a skilled technician or chemist. In some cases a single instmment can not provide analyses for the wide variety of additives a particular organisation utilises. Additionally, laboratory techniques rarely provide results in a timely fashion. Determination of physical properties is not the least important if one thinks of pigments, talc and other fillers. Application of spectroscopic techniques to polymer production processes permits real-time measurement of those qualitative variables that form the polymer manufacturing specification, i.e. both chemical properties (composition, additive concentration) and physical properties (such as melt index, density). On-line analysis may intercept plant problems such as computer error, mechanical problems and human error with respect to additive incorporation in the resin production. Characterisation and quantitative determination of additives in technical polymers is an important but difficult issue in process and quality control. 

---