What do Pump Pros Know and why do they know it? This series of articles highlight the brilliance of those who work with pumps.
Monitoring Energy Consumption and System Variables
Measuring energy consumption of a pumping system can be very straightforward. Simply installing an electric meter ahead of the main line that supplies electricity to an entire pumping system can indicate the power used by all electric components in the system, such as electric motors, controls, and valves. The meter can reveal how much power the system uses during a process cycle, a work shift, and throughout the year; and comparing the values from electric meters on similar machines lets us determine which machines use more, or less, energy than others.
Another important function of systemwide energy monitoring is that it can show how energy usage changes over time. Systems that adhere to production cycles may have regular periods when they consume maximum energy, and idle periods when they consume minimal energy. The best an electric meter can do to reduce energy costs is to have us stagger the production cycle of machines so that they all consume maximum energy at different times. This doesn’t actually reduce energy consumption, but it can reduce the cost of energy by reducing peak usage.
Recording electrical power consumption when a machine is first commissioned can serve as a benchmark for future energy audits. If the entire machine is well maintained, energy use can remain close to the benchmark from when it was commissioned. After years of operation, though, the electric meter will likely indicate that the system is using more energy to perform the same amount of work.
This systemwide energy monitoring is great for establishing a baseline for a system’s most efficient operation, but it cannot reveal what areas within the system have become less efficient or how much. Experienced technicians may be able to identify some areas that need improvement using diagnostic instruments, but they could overlook problem areas that had not posed problems in the past.
Planning a Strategy
A better approach is to install sensors, test points, and instruments at strategic areas that monitor conditions throughout the system. These sensors provide critical data that can be used several ways. First, the sensors provide an immediate indication of flow, pressure, temperature, and other parameters in real time. Second, the data can be used for automated machine control, which removes much of the potential human errors of manual control. Third, the data can be accumulated over time to indicate trends in operation.
Real-time monitoring — Establishing set points for sensors enables them to trigger an alarm if a threshold has been exceeded. For example, a low-pressure indication in a pump’s suction line could sound an alarm to prevent the fluid from vaporizing in the pump. If no response occurs within a prescribed time, the control could shut the pump off to prevent damage. Similar control scenarios could be implemented for sensors to signal an alarm if a high temperature or high vibration occurs.
Automated machine control — A natural progression from using sensors to monitor set points is to use them for direct machine control. For example, if a pump is used for circulating cooling water through a machine, a temperature sensor can transmit a signal to a controller that regulates flow. This controller might vary the speed of the motor driving the pump or shift a valve to match flow of the pump to the cooling demand. Lower energy usage is the ultimate result.
The sensors also enable predictive maintenance. If a machine malfunctions because of a clogged filter, a technician or mechanic first has to make sure the machine is shut down, then lockout/tagout the machine so the filter element can be cleaned or replaced safely. This is an example of reactive maintenance—action is taken to correct a fault after it occurs and without prior warning. Filters need to be changed periodically but relying on standard time periods may not be effective.
In this case, water going through the filter may have been more highly contaminated and for longer period than anticipated. As a result, the filter element should’ve been changed before its scheduled time. On the other hand, replacing a filter element on a set schedule based on time can be wasteful. If the water going through the filter has been unusually clean for a long period, the element may not need to be changed until weeks later than with a time-based schedule.
The point is that using sensors to monitor differential pressure across the filter gives an accurate indication of when the filter needs to be changed. In fact, the differential pressure reading leads to the next level, predictive maintenance.
Data Collection Over Time — Returning to our recently commissioned system, once everything has been started up, adjusted, and fine tuned, sensors can provide baseline readings of all pressures, flows, temperatures, vibration, and other operating parameters. Later, we can compare current readings to the best-case values to determine how much components have worn, or how much the system has changed (e.g. the clogged filter).
Future readings will eventually drift from the baseline values established at startup. When readings drift beyond a predetermined limit, they may indicate impending failure, or at least needed intervention. This is predictive maintenance—alerting the operator that an impending failure may occur well before it does.
A common example is when we install vibration sensors (accelerometers) at the bearing locations (or bearing housings) for both the pump and motor. Normal wear and tear of rotating machinery or operating the pump outside of the manufactures set parameters can cause the frequency or amplitude of the rotational vibration to change—usually indicated by increased amplitude of the vibration. Specialist can examine vibration signatures at startup to determine if they are acceptable, and specify critical values that signal that attention is needed. These values can be programmed into control software to signal an alert when output from the sensor reaches its critical limit.
At startup, the accelerometer provides a baseline value of vibration that can be saved in a control memory. When real-time values eventually hit the predetermined limit, the machine control alerts the operator of a situation that needs to be evaluated. Of course, a sudden extreme change in vibration would also alert the operator of a potential failure.
A technician reacting to either alert may find a simple fault, such as a mounting bolt that has become loose—which could cause the pump or motor to shift out of alignment. Realigning the equipment, and tightening all mounting bolts, may be the only action required. After system restart, the real-time vibration readings would indicate if the problem was corrected. However, if bearings in the pump or motor have been damaged, further corrective actions may still be required. But again, because the sensor provides early warning of a potential problem, it can be evaluated, and downtime may be able to be postponed until the end of the shift, a scheduled shutdown, or when production can be diverted to another pump or system.
More Than Automation and Reliability
Sensors strategically placed throughout a system generally to provide automated control, support operations, and predictive maintenance. But they also allow for taking a closer look at a system’s operation so they can be optimized to improve energy efficiency of the entire system.
In fact, applying this strategy to an existing system can reduce energy consumption by exposing pumps or components that have a lot of room for improvement.