## Process Control

Process Control is the study of automatic control principles applied to chemical processes. It applies principles of mathematics and engineering science to the regulation of the dynamic operation of process systems. To be successful, you need strong applied mathematics skills and process understanding (most of which is just common sense).

The skills and tasks you've been exploring in the first three years of ChE classes are predominantly analytical. They are used for diagnosis and understanding of processes and problems. This year, your design classes will work on synthesis skills for devising new processes. In Process Control, we will use some analytical tools (old and new) and synthetic skills to understand the dynamic (time dependent) behavior of processes and ways to regulate plant operation.

Since the primary function of control systems is to compensate for dynamic changes in process systems, we need to understand the dynamics of processes -- how their behavior changes with time -- if we are to develop workable solutions. We address this need through dynamic modeling of the chemical processes. Mathematically, this means we will be dealing with differential equations.

### Why is Control Necessary?

Process plants do not operate at steady state, no matter what you have been assuming in other classes.

Consider what might happen to a distillation column operating in a plant:

• market price of feedstocks and products vary, so sources and suppliers change; the 50% feed you are using today may be 45% tomorrow.
• ambient temperature changes continuously; this changes cooling water temperatures, which change condenser duties, which changes in column operating pressure and overhead flows.
• steam system supply pressure varies as sources and users are switched on and off; this causes the reboiler temperature and load to vary, changing the flows at the base of the column.
Meanwhile, the people who buy your products want to get exactly the same material all of the time, and often require statistical proof of minimal variance. Since plant behavior is variable, but your customers won't accept variation, control systems are needed. A good control system keeps a plant running at predictable, regulated conditions.

Putting all this together, the two main functions of control systems are:

1. setpoint tracking -- the ability to shift from one desired operating point to another (like you driving your car)
2. disturbance rejection -- the ability to maintain an operating point despite fluctuating conditions and external forces (like your thermostat)
More specifically, control systems
• allow basic operation by managing throughput and inventory
• ensure safety by preventing dangerous conditions
• guide transitions between operating states (startup, shutdown, setpoint tracking)
• maintain product quality by rejecting outside forces which act on the process
• implement optimizing strategies devised by management or supervisory computers

Disturbances can never be completely eliminated; however, a good control system can greatly attenuate their consequences and reduce the variability of process parameters. If we can reduce variability, we need smaller margins of error and contingency allowances, and so can operate much closer to optimum conditions, reducing waste and saving money.

### Course Organization

Control courses can be difficult for an instructor to organize. There are often multiple ways of approaching concepts, each with its own "dialect" of terminology and equation. Topics often wrap back around, so books and instructors sometimes have a tendency to use terms and ideas before they are fully defined. I'm guilty of this myself -- so PLEASE do not hesitate to ask questions when you spot things that feel out of place.

### Chemical Process Control

The same basic control methods, principles, and tools apply whether the "process" is chemical, electrical, or mechanical. Control theory has been developed by ChEs, EEs, and MEs, so the terminology reflects concepts from all three disciplines (as well as mathematical systems and optimization theory).

Differences in the application are what separate ChE control from other practitioners. Chemical process systems are distinguished by:

• longer time constants (minutes for a HX, hours for many columns)
• long transportation lags or "dead time" (minutes)
• nonlinearities (reaction kinetics)
• distributed parameters (coupled material and energy balances)

### Safety Systems

No feedback control loop, no matter how well-designed and tuned, can guarantee safe operation. Consequently, a regulatory process control system cannot be trusted as the primary safety system. Almost all chemical plants have a second, parallel control system to handle safety alarms and shutdown. While we will always consider the safety aspects of control systems, we will not study the design of these alarm/shutdown systems.

### Control Objectives

The objectives of a control system fit into a hierarchy -- that is, some objectives are given priority over others. One way of ordering the hierarchy is by the purpose of the control system components:

1. Safety
2. Environmental Protection
3. Equipment Protection
4. Smooth Plant Operation
5. Product Quality
6. Profit Optimization
7. Monitoring and Diagnosis
According to this structure, control loops responsible for safety-related tasks will always have priority over all other tasks; loops for product quality will have priority over loops whose primary task is optimization; and so forth. Most of the techniques we study in this course will apply directly to the operating and quality objectives.

Be aware that a loop can serve more than one purpose and that its place in the hierarchy is not always cut-and-dried.

One of the themes of our study this semester will be the tradeoffs between plant design and plant operation. Control systems are part of the day to day operation of a plant. This suggests another way of ordering the hierarchy of control objectives: "achievability". After all, until your plant is operating, controls aren't needed at all. This sort of hierarchy tends to group loops by function as much as it does by objective:

1. Production Rate & Inventory Controls
2. Safety/Environmental Controls
3. Equipment and Operating Constraint Controls
4. Product Quality Controls
5. Optimization

References:

1. Coughanowr and Koppel, Process Systems Analysis and Control, McGraw-Hill, 1965, pp. 1-10.
2. Luyben, Process Modeling, Simulation and Control for Chemical Engineers (2nd Edition), McGraw-Hill, 1990, pp. 1-12.
3. Marlin, T.E., Process Control: Designing Processes and Control Systems for Dynamic Performance, McGraw-Hill, 1995, pp. 1-10, 21-28.
4. Riggs, J.B., Chemical Process Control (2nd Edition), Ferret Publishing, 2001, pp. 3-4.

R.M. Price
Original: 9/29/93; 12/7/96
Modified: 1/10/94, 1/6/95, 12/15/95, 12/7/96, 1/6/98, 4/24/2003, 8/10/2004