Process control, automatic controller specification

The calender was developed over a century ago to produce natural rubber products. With the developments of TPs, these multimillion dollar extremely heavy calender lines started using TPs and more recently process principally much more TP materials. The calender consists essentially of a system of large diameter heated precision rolls whose function is to convert high viscosity plastic melt into film, sheet, or coating substrates. The equipment can be arranged in a number of ways with different combinations available to provide different specific advantages to meet different product requirements. Automatic web-thickness profile process control is used via computer, microprocessor control. 

The detailed design and specification of the automatic control schemes for a large project is usually done by specialists. The basic theory underlying the design and specification of automatic control systems is covered in several texts Coughanowr (1991), Shinskey (1984) (1996) and Perry et al. (1997). The books by Murrill (1988) and Shinskey (1996) cover many of the more practical aspects of process control system design, and are recommended. 

Automatization of all stages of the analytical process is a trend that can be discerned in the development of modern analytical methods for chemical manufacture, to various extents depending on reliability and cost-benefit considerations. Among the elements of reliability one counts conformity of the accuracy and precision of the method to the specifications of the manufacturing process, stability of the analytical system and closeness to real-time analysis. The latter is a requirement for feedback into automatic process-control systems. Since the investment in equipment for automatic online analysis may be high, this is frequently replaced by monitoring a property that is easy and inexpensive to measure and correlating that property with the analyte of interest. Such compromise is usually accompanied by a collection of samples that are sent to the analytical laboratory for determination, possibly at a lower cost. 

Experienced process control engineers are usually responsible for the design and specification of automatic control schemes on large chemical plants. The book by Shinskey (1979) provides details of the practical... 

Cell number and viability determination is not only an excellent direct process parameter, but also the basis for all specific and calculated parameters (growth rate, specific consumption and production rates). Therefore, the apphcation of a sophisticated process control system is largely dependent on a reliable cell number measurement. Most laboratories determine the cell number within the reactor by sampling and off-line analysis. This is not satisfactory because of the necessity of frequent handling, overnight attendance and problems with reproducibility. On the other hand, automatic devices are either very complicated (hke sampling photometers or sampling cell counters) or are relatively new and not validated. 

In this chapter we have advocated the use of online model-based optimization for the automatic control of SMB plants. This approach has the advantage that the process is automatically operated at its economic optimum while meeting all relevant constraints on purities and flow rates. Application to a pilot-plant-scale reactive SMB process for glucose isomerization showed that implementation at a real plant is feasible - the requirements for additional hardware are moderate (a high-level PC and online concentration measurements in the recycle line). The experiments confirmed the excellent properties of the proposed control scheme. The scheme is extremely versatile, and the cost function and constraints can easily be adapted to any specific separation task. 

A control plan is the sum of the procedures and equipment used to ensure that the internal requirements are met. It includes control charts, sampling plans, 100% inspection, feed-forward/feedback mechanisms, and mistake-proofing tech-niques/devices. Some items may be performed by equipment like automatic controllers, and some are performed by operators and checkers. There may not be a single document called a control plan. Instead, these controls might be spread across a number of documents including a statistical process control plan, an inspection plan, an operator manual, and various other standard operating procedure and specifications. 

The preceding sections have summarized briefly what is involved in automatic process control and how control problems may be attacked quantitatively. In the course of the presentation some soft spots have been manifest. For one thing it is clear that we do not know how to specify the kind of controlled behavior that is optimum for a given plant or process. General or approximate specifications can be given but they are not optimal. Actually, control quality specification can be made only on the basis of the disturbances affecting the process, and at the present time there is virtually no reliable knowledge about process disturbances and where they apply. 

The on-line instrumentation development needs can be divided along the traditional lines of quality control and process control instrumentation The interrelationship between the two and the proposed mode of interaction are shown in Figure 9 The quality control instruments, in an operational and a practical sense, are located outside the preparation plant As a result of data obtained on product quality, the quality control instrument interacts with the plant feed and the process control instrumentation to optimize the use of raw material consistent with the desired product specifications This feedback loop can be automatic however, current practice places the preparation plant superintendent or engineer in the loop and the quality control measurement in a remote laboratory, resulting in manual control scheme with a long time delay The process control instrumentation controls a single unit operation and the instrumentation is unique to each unit operation Their set points are established by the output from the quality control Instruments ... 

Process capability is shop specific. Each shop needs its own process capability database of its own before process planning can be done automatically. Collecting and analyzing capability data can be tedious. Some of these data can be collected through inspection, such as from the control charts. Others require experiments on the machines. Most of the processes that remove material quickly, such as milling, drilling, and turning, create poorer surface finish and accuracy. 

Many books have been written concerning the methods of process control for specific unit operations. Basically, automatic control is relatively easy to comprehend. In many ways it is like manual control. However, the automatic controller does not necessarily duplicate what the human operator does by hand. Automatic equipment gives continuous, minute attention to the control application. Automatic controllers can compute and remember, but they cannot reason from new conditions, nor can they forecast beyond the data which are built into them. 

Servomechanisms. Automated processes controlled by digital logic require other devices by which the function of the machine can be measured. Typically, a measurement of some output property is automatically fed back into the controlling system and used to adjust functions so that desired output parameters are maintained. The adjustment is carried out through the action of a servomechanism, a mechanical device that performs a specific action in the operation of the machine. Positions and rotational speeds are the principal properties used to gauge machine function. 

---