To gain a basic understanding of process and temperature control, you must look at a number of control aspects as they relate to industries such as food, pharmaceutical, chemical and petrochemicals. In these applications, multiple-loop systems can be used to control a single variable to satisfy the requirements of a combination of controlled variables. In other words, there can be only one independent setpoint at any given time; however, this does not exclude the use of several other controlled variables to achieve optimum control.
Cascade ControlCascade control is the most common multiloop control scheme in use. It decreases deviations of the primary variable and minimizes out of spec product after a process disturbance. In this control scheme, the output of one primary controller is used to manipulate the setpoint of a secondary controller. The two controllers are cascaded with each other, but each has its own process variable input. The primary controller has an independent setpoint; its output goes to the secondary controller. It accepts the primary controller's output as its setpoint input and has its control output as the output to the process. This secondary controller is a closed loop within the primary loop.
Consider a simple heater temperature control without cascade control (figure 1). In this example, water flows at a constant rate through a pipe being heated by a burner. The temperature controller adjusts the fuel valve position to maintain the feed outlet temperature, which is monitored by the temperature transmitter. A mechanical linkage between the air and fuel valves maintains the fuel/air ratio. However, if any fluctuations in fuel pressure occur, the amount of energy to the heater will fluctuate and temperature at the water outlet will deviate from setpoint. A simple feedback loop cannot adequately control this system because it does not monitor fuel pressure or compensate for pressure fluctuations. Once the error is detected by the temperature transmitter, it is too late for corrective action to maintain setpoint. Even optimum PID tuning will not correct this situation.
Controlling fuel flow alone will not satisfy the application because it will not respond to water temperature load changes. This can be accomplished with two independent controllers or one controller with two internal control loops. Use cascade control when:
- The primary variable has a slow response to disturbances.
- Process changes cause serious upsets in the controlled variable.
- Another variable is affected by disturbances and is closely related to the control variable.
- The secondary variable can be controlled, and it responds quickly to the primary controller.
Feed Forward ControlFor cascade control to be an option, the variable causing the disturbance must be controllable. However, in some applications, a disturbance that cannot be controlled may occur. Feed-forward control bases its control action on a known disturbance forward of the PID algorithm. It measures the disturbance and feeds it directly to the output of the PID algorithm. Feed-forward control minimizes the disturbance long before the PID algorithm can because PID is not capable of reacting until after the error is measured. An illustration of this could be a disturbance in the product's feed rate through the water heating example. A typical feed-forward diagram is shown in figure 3.
In this example, a flow transmitter has been added to measure the feed inlet flow, and its output is multiplied by the setting on the adjustable gain relay. Combining two signals by a summing amplifier or circuit results in a feed-forward signal, which becomes the temperature controller's second input. As inlet water flow increases, the flow transmitter instantly feeds an increased signal to the summing amplifier. This signal is added to the temperature controller output and will produce an immediate change in fuel/air valve position. The increased fuel inlet can now prevent large feed temperature deviations from occurring. This control scheme can be accomplished using one controller that has a second input with math capability (gain input and summing functions).
Considering Ratio ControlRatio control is used in applications where two process inputs must be controlled to meet the process demand. The values of the two inputs must be kept in a particular ratio to achieve the end-product specifications. Examples include blending base product with thinner or water. Two ratio strategies are available: series and parallel.
In a parallel ratio control strategy, the master controller manages both loops simultaneously (figure 5). The advantage is that the systems are independent, and any electrical signal noise in one loop will not affect the other. Another advantage is that if the application requires one flow to be interrupted for maintenance purposes, this interruption will not affect the second flow.
Gaining a basic comprehension of control strategies can provide a good starting point for understanding process control.