The Fundamentals of Process Control
Industries such as food, pharmaceuticals, chemicals and petrochemicals, and the applications within these industries -- refining crude into gasoline or converting wood to paper, to name a couple -- involve the handling of raw materials in a continuous or batch process. Via mechanical or chemical changes, a product is manufactured. How well or poorly this product meets the manufacturer's desired characteristics depends in large part on how well the manufacturing process is controlled. By understanding control fundamentals, process measurements, controller characteristics, tuning, control strategies and final control elements, a processor can ensure that his process will produce acceptable product every time.
Process Control DefinedFirst, a definition of what is meant by process control is needed. Simply put, process control involves the measurement of a process variable (temperature, flow or pressure, for example), the comparison of that variable against a desired value (called a setpoint), and the generation of a change in the process to adjust the variable to the desired value. Any change in the process in response to commands from the process control usually involves the transfer of energy from a source to the process you are trying to control.
Temperature control is a subset of process control, and one with which most people are familiar. One example of temperature control is the thermostat used in your house. In this instance, the process variable to be measured is the temperature in your home. A bimetal sensor in the thermostat measures the home's temperature and compares it to the desired value, which is the temperature selected on the thermostat. If the two values are equal, then no action, or transfer of energy, is requested by the thermostat. If, however, the measured temperature is lower than the desired temperature (referred to as the error), the thermostat sends a signal to your furnace telling it to turn on and stay on until the room is warmed to the desired temperature. The sensor monitors the home temperature, and once the two values match, the thermostat turns off the furnace.
In the previous example, a transfer of energy occurs from the furnace, where natural gas is ignited, burned and used to heat the surrounding air. This heated air then is pushed into the room where the thermostat is located via a fan. Even in this simple example, the basic pieces of a process control system are present: controller (thermostat); sensor measuring temperature (bimetal sensor in the thermostat); final control element (furnace); and the process itself (the room to be heated). This is an example of a closed control loop or feedback control loop control scheme, which means that the entire self-contained process control system works on information directly obtained from the process variable.
How Sensors Fit InSensors are simply devices that measure a condition -- temperature, pressure or flow, for example -- and provide a signal to the controller or transmitter. Modern controllers can accept many different inputs directly from sensors without requiring a transmitter. Why choose one sensor over another? Your decision depends on what you are trying to measure. I'll limit this discussion to temperature control.
Temperature sensor choices include thermocouples, resistance temperature devices (RTDs) and noncontact sensors. Sensor selection should be based on the temperature range desired, the sensor available for that range and the accuracy required. Some thermocouples have a range of 0 to 3,300oF (-17.8 to 1815.5oC), with their most accurate sensing capabilities at the higher temperatures. Con-sequently, these thermocouple types typically are used in high temperature furnace applications. For lower temperature applications in the -300 to 1,200oF (-184 to 649oC) range, RTDs typically are used. RTDs often are employed in oven applications for food baking or environmental testing, to name a few. Finally, noncontact (typically radiant-type) sensors are used where the temperature is so high that standard thermocouples cannot be used, or where physical contact with the product cannot be tolerated.
Sometimes the distance between the sensor and the controller is so large that the sensor's output signal degrades before it reaches the controller. To eliminate this problem, a temperature transmitter is used. A temperature transmitter is a device that retransmits a sensor's output to the controller when the sensor signal alone is not strong enough to reach the controller without some degradation. Some controllers can accept sensor signals such as thermocouple and RTD inputs directly. The transmitter converts the low level thermocouple millivolt signal to a higher level signal (such as 4 to 20 mA) and sends that signal through wires for long distances.
Controller ChoicesControllers come in many shapes, forms and sizes and in different design types such as pneumatic, analog, digital mainframe computers and digital panel. Such a range of choices can be daunting; fortunately, controller selection is eased because and each controller type has advantages and disadvantages.
Pneumatic controllers were the first sophisticated type of industrial process controller developed. Operated on air pressure, today these controllers usually are used only in limited applications where a high degree of explosive atmosphere exists in the immediate area. Pneumatic controllers provide a high degree of safety for these applications, but they are expensive, utilize an old technology, have limited features, and require much maintenance.
Electronic analog controllers were developed next and today are used extensively in fundamental applications such as simple food processing and plastics manufacturing. Usually inexpensive, analog controllers have a limited functionality set, so complex control strategies are difficult to implement, and multiple units are required to achieve adequate control. Also, analog controllers are not easily modified in the field. Instead, they must be ordered with the particular input and output types needed. If you make a mistake, you must send it back to the factory or order new parts to upgrade the controller.
Digital mainframe control systems came about with the development of the mainframe computers in the mid-1970s and today include loop control used in programmable logic controllers (PLCs). These systems allow the control strategies to be manipulated using powerful software from a keyboard and monitor; however, if the mainframe fails, you lose your control. For critical loop backup, using PLCs with separate stand-alone controllers is a more acceptable strategy.
In the early 1980s, distributed control systems were developed that had all the advantages, power and flexibility of a mainframe, but this approach distributed the functionality over various components that communicated with each other. With this setup, if one piece of hardware failed, the rest of the system continued to function.
At about the same time, the digital panel-mount controller was developed. It offered single-loop integrity while allowing great flexibility of control strategies using the front keypad or a separate programming device. Today, digital panel-mount controllers are used in the majority of processes requiring less than 20 loops of control. For more complex, multiple loop applications, the PLC, distributed control and new generation multiloop process controllers are used.
Within the last decade, manufacturers of panel-mount controllers have standardized on the European format for front face sizes: 1/32, 1/16, 1/8 and 1/4 DIN. In the past, the controller's size was indicative of its functionality -- the larger the size, the more functionality it offered. However, in recent years, smaller-size controllers have become able to provide the functionality of their larger-sized counterparts.
Control AlgorithmsOn/Off Control Algorithm. A simple control algorithm, with on/off control, the output is either on (100% power) or off (0% power). The control algorithm "decides" whether to be on or off by comparing the process variable value to the setpoint value. This comparison determines the sign (positive or negative) of the error, and the on/off algorithm operates on the sign of this error. In direct-acting control, if the error signal is negative, the output is 0%; likewise, if the error signal is positive, the output is 100%. With reverse-acting control, the opposite is true. Commonly, an adjustable overlap value, called a hysteresis band, is provided between the off and on states to prevent the output from changing states rapidly when the error is very small.
PID Control Algorithm. A PID controller includes three common control modes -- proportional, integral and derivative -- within a controller. As compared to on/off control, a PID controller provides a modulating type of control output signal, capable of positioning the opening of a control valve, for example, to any degree from fully open to fully closed. A basic explanation of each control mode (without getting into a detailed mathematical explanation) follows.
Proportional control is a corrective action that is proportional to the error. Remember that error is defined as the difference between the desired value (setpoint) and the actual measured value (process variable). Proportional control is referred to as either gain or proportional band. Proportional band is the percent of change in the controlled variable that causes the controller output to move over 100% of its range. In other words, a proportional band of 100% means that if the measured variable moves over 100% of its range, the controller output signal (usually 4 to 20 mA) will move over 100% of its range; or, if the measured variable moves 1% of its full range, then the control output moves 1%. In general, as the proportional band percentage gets smaller, the controller becomes more sensitive. It is possible to make it so sensitive that a very small change in the process variable measurement will result in very large controller output. For those who like to work in gain instead of proportional band, the relationship between these terms is simple: gain is the reciprocal of proportional band. If the proportional band is 100%, then gain is 1/100%, or 1. If proportional band is 50%, then gain is 1/50%, or 2.
A proportional-only controller cannot control most processes well because it never reaches the desired setpoint. This offset, defined as the difference between the desired setpoint and the actual process value that the controller can achieve for the process, is due to the mathematical nature of the proportional control mode. Because there is only one controller signal output value associated with the particular process variable value the controller is reading, the controller sticks to its rigid mathematical association between input and output values. It cannot recognize that there is a problem of not achieving the desired setpoint value with proportional-only control. Increasing the controller gain will reduce, but not eliminate, the offset, but too much gain will result in unstable control that will oscillate endlessly above and below the desired setpoint.
To eliminate this persistent error, integral control is used. Also re-ferred to as reset action, integral control senses that an offset exists and resets, or repeats, the proportional action until the offset is eliminated.
Integral settings are defined in repeats per minute or minutes per repeat. A setting of 10 repeats per minute will instruct the controller to repeat the action 10 times each minute. By contrast, a setting of 10 minutes per repeat instructs the controller to repeat the action once every 10 minutes. In general, the more often the repeats, the faster the offset will be eliminated. Integral control is a control mode that acts only as long as an offset exists.
The third control mode, derivative control, is a little harder to explain. Basically, it is an anticipatory control mode that senses how fast the error is changing and increases the controller output signal. Therefore, it allows the process to reach the desired setpoint faster than with proportional-only control. There is no derivative action if the error is constant; therefore, derivative action does not correct offset. But, if the error is rapidly changing, derivative control will increase the gain temporarily in proportion to the change. Although derivative action can be used in many different processes, it primarily is used in slow processes or where large dead times exist. Dead time is the time difference between when a setpoint change is made to the controller and when the process first starts to react to that change.
The term "controller modes" relates to the PID functions. A single-mode controller typically is a proportional-only controller; a two-mode controller normally is a proportional plus integral (PI) controller; and a three-mode controller is a proportional plus integral plus derivative (PID) controller. Most modern digital controllers offer all three modes and allow users to select which mode combination to use for a given process.
Controller Tuning. The term "tuning a controller" refers to setting the PID parameters within the controller. Some users prefer to manually tune a controller by selecting values for each PID parameter, seeing how the process reacts, and repeating the procedure until they get the process control reaction desired. A controller can be manually tuned using basic mathematical principles developed in the 1940s. With the development of self-tuning software in most digital controllers, however, this time-consuming and often frustrating manual procedure is going the way of buggy whips. Few people take the time to tune a controller manually when dependable software can do it for them.
According to the results of a survey conducted by a controls company, more than 85% of users utilize self-tune software in controllers. Fundamentally, there are two different types of self-tuning software available. The first, on-demand self-tune, calculates the PID settings when initiated. These PID settings do not change unless the software is re-initiated by the user. The second, adaptive tune, continually monitors the process and periodically changes the PID values to optimize control for changing process conditions.
The term "fuzzy logic" often is used to describe a controller's tuning capabilities. Basically, fuzzy logic is software that works with self-tune software to minimize the time required to get to setpoint while at the same time minimizing any overshoot. (Overshoot is any controller action that allows the process variable to go above or below the setpoint value before achieving steady-state setpoint.)
This brief overview provides a good starting point for understanding process control.