Mistake proofing denies the ability of the control system to perform functions that could cause harm or damage to personnel, equipment, or the process itself. To properly institute mistake proofing, an analysis of potential failures with significant consequences needs to be undertaken. Knowing potential significant failures, a failure analysis, or failure tree, must be performed to find all of the potential circumstances and causes for the possible failures. Once the circumstances and causes for a potential failure are known, an approach to “mistake proof” the process and prevent the failure can be initiated. There are four main strategies for mistake proofing:
• Elimination of the possibility,
• Prevention of the mistake,
• Detection and alarm, and
• Loss control.
Elimination is not as important as the other mistake-proofing strategies in automatic control because it typically is not possible to eliminate the process function. Conversely, prevention is the most common method of mistake proofing used in automatic control. The control system can prevent mistakes from occurring. Detection can be used to identify potential or actual problems. Loss control is used to control a problem once it is detected and to keep it from getting worse. For example, the control system can stop chemical flows once a leak is detected.
Automatic control mistake proofing is carried out with interlocks and permissives. The system must be able to sense when a mistake could occur (or has occurred). This is done with sensors such as proximity switches that indicate valve position, level sensors on a tank, pressure sensors, and so on. If a mistake occurs, it is necessary for the system to provide actions such as stopping a pump, closing or opening a valve, and so on.
An example of mistake proofing is preventing pump damage by ensuring an adequate path for sludge before allowing a positive displacement sludge pump to start. The control system can monitor the position of all valves in the potential pump discharge path. If an adequate path does not exist to ensure safe flow to an appropriate destination, the control system will prevent the pump from starting.
4.0 CONTROL CONCEPTS
Based on sensor measurements, control loops change manipulated variables via some type of actuator. They convert a mathematical control strategy into a mechanical or electronic system. For example, a thermostat from an older home senses a room’s temperature via the bending of a bimetallic strip and responds by using a switch to open or close the heater’s circuit.
Control options may be classified in several ways. The following sections explain the various options and classifications of control.
4.1 On–Off Control Versus Modulating Control
In the thermostat example in the preceding section, the control is considered on–off control because there are only two states the heater can be in: on or off. This is often called discrete control because there are only two discrete states that can be controlled. When the temperature in a house drops below the setpoint, the thermostat responds by turning the furnace on. When the temperature has risen sufficiently, it turns the furnace off. The thermostat is designed with a temperature deadband to prevent on–off cycles from occurring too rapidly; however, its main drawback is that the temperature constantly oscillates around the setpoint.
The other control type is modulating control. An example of modulating control is a modulating air valve that varies its opening to maintain the dissolved oxygen level in the aerobic section of an activated sludge basin. The valve’s position is controlled. If dissolved oxygen is low, the valve opens a little more and, if dissolved oxygen is high, the valve closes a little more. This is also called analog control because the position continuously varies from completely closed to completely open. There are many examples of both types of controls in this section.
4.2 Open-Loop Versus Closed-Loop Control
Processes without a direct feedback of the process variable are called open-loop systems. Those with a direct feedback control loop are called closed-loop systems. An example of an open loop would be driving a car (without cruise control). The control action is depressing the gas pedal. If the pedal remains in the same position, the car will speed up and slow as the car goes over hills. Using cruise control is an example of closed-loop control. The cruise control adjusts the amount of gas feed to maintain a constant speed for the car. While going uphill, cruise control will increase the gas feed rate, while, going downhill, it will decrease the gas rate.
Closed-loop control is used when there is a relationship between the control action and the process variable and the process variable can be easily measured. Closed-loop control is common in wastewater treatment control. An example includes dissolved oxygen control in an aerobic activated sludge system, which was mentioned previously. However, there are instances when open-loop control is needed and can be applied effectively. An example is “flow pacing”, or adjusting chemical flow proportionally to plant flow. This keeps the dosage rate of the chemical constant, but does not control the chemical feed based on a measurement of the desired effect of the chemical.
4.3 Feed Forward Versus Feedback Control
Closed-loop systems are also called feedback control because they provide direct feedback of the process variable to the control system to allow manipulation of the controlled variable. Feed-forward control is an open-loop control that contains a prediction or model of how the process is expected to respond based on data other than the actual process variable. The aforementioned flow-pacing example represents feed-forward control. In this instance, the prediction is that the chemical effect will be the same if the dosage rate is constant. The chemical flowrate is then varied in accordance with facility flow to maintain a constant dosage rate.
4.4 Step Control
Step control provides control response in discrete steps or levels. An example of step control is a pump station with constant speed pumps. Possible steps in a two-pump station are listed in Table 7.1.
Step control is advantageous because it is simple and it eliminates the cost of variable control equipment; however, it does result in more variability. For example, the pump station will keep the level within a range, but it cannot control precisely enough to keep the level constant. This is likely not necessary for lift stations; however, for many control applications, more precise control is needed.
TABLE 7.1 Possible steps in a two-pump station.
4.5 Proportional–Integral–Derivative Control
Perhaps the most common control mechanism in the wastewater treatment industry is the proportional–integral–derivative (PID) controller. To understand how it works, consider its use in controlling the speed of an automobile via a cruise control system.
If a car maintaining a speed of 88 km/h (55 mph) on a flat road slowed down because of a change in wind or road slope, the proportional (P) control loop would press the accelerator in proportion to the error (E), as follows:
Where
PV = actual speed (process variable) and
SP = desired speed (setpoint).
The accelerator position is the manipulated variable (M). This type of control can be expressed as
Where