Implementing predictive maintenance (PdM) on equipment that operates intermittently, seasonally or not at all can be quite challenging. Most of the available predictive technology is designed to monitor equipment at a steady state with repeatable conditions so that users can compare current data with previous readings. Additionally, there is the problem that failure modes change depending on the operating profile of the equipment. The same type of equipment will have different failure mechanisms depending on whether it is constantly running, operating intermittently, in standby condition with only periodic testing or on long-term lay-up.
The good news, though.
Anywhere there exists a problem, there is also an opportunity. The trick is to move away from trying to adapt the mainstream technologies to an irregular situation and start with a custom-made PdM plan that is tailored to your individual situation. Here is how to do it:
Take advantage of the downtime
A great place to start with PdM on standby equipment is to understand that some technologies are meant to be applied offline. If your equipment has a lot of standby time, there is no excuse not to have these programs implemented.
- Precision alignment: I have seen figures from laser- and precision-alignment equipment vendors claiming that up to 50 percent of rotating machinery failures can be directly attributed to misalignment. I take that with a grain of salt. Even if the real figure is a tenth of that, precision alignment is worth implementing in most industries. Misalignment is the root cause of several machine failures, which can lead to higher energy consumption, higher vibrations, premature wear and overheating, as well as other failure mechanisms. If you want your equipment to run smoothly when you need it, precision alignment is a must.
- MCA offline: Motor circuit analysis (MCA) is often divided into two categories: online and offline. While some analysis is capable of doing both, it comes with a hefty price tag. For standby motors, offline MCA goes beyond insulation resistance and allows for fault finding and condition monitoring of both rotor and stator. It´s a growing PdM tool that allows for trending and monitoring of the overall health of electric motors.
Remember that some things still apply
Even if the plant is not currently producing, some predictive testing technologies still apply. As long as electrical panels are energized, it’s still valid and recommended to perform periodic ultrasonic electrical inspections to detect signs of arcing, tracing and corona. The same applies to dissolved gas analysis (DGA) of dielectric oil in power transformers.
Lubricating and hydraulic oil analysis of standby equipment should also be a part of your monitoring program for standby equipment, but because your failure mode changes, so does what you monitor. Instead of focusing on wear particles, the oil analysis program will tend to shift its emphasis to oil contamination, corrosion, additive depletion, etc. Special attention should be paid to the sampling procedure to make sure representative samples are being obtained.
Seize the opportunity
Standby equipment will, by its nature, have higher start/stop ratios per running hour than its counterparts. This is something you can and should take advantage of to customize your condition monitoring program.
- Coast-down time: Chances are you perform periodic starts on standby equipment but are you trending coast-down times? Coast-down time is a basic condition-monitoring technique that is often underutilized. It refers to simply measuring the time it takes an equipment to come to a complete stop once the power or fuel is cut off.
My experience trending coast-down times started on large frame gas turbines. Under normal conditions, it took around 20 minutes for a turbine to come to a complete stop from 3600 rpm. I have seen that time decrease to eight minutes under a misalignment condition. That equates to a 60% decrease and it´s hard to miss. A myriad of variables can affect a machine´s coast-down time, but the two most common are unbalances and misalignment.
- Rate of rise in bearing temperature: I once read the following phrase on an electric motor operation and maintenance manual: “Following the initial start-up, the bearing temperatures should be closely monitored. The rate of rise in bearing temperature is more indicative of impending trouble than is the actual temperature.” Therefore, we started monitoring and trending bearing rate of temperature rise during start-up to validate its usefulness as a condition monitoring technique.
Your standby equipment doesn’t need to suffer from lack of periodic check-ups when it comes to predictive technology. A full asset health report based on multiple condition-monitoring techniques is possible, it will just look different.
REFERENCES
Gupta, Piyush, and O. P. Ghandi. "Coast-Down Time Monitoring for Defect Detection in Rotating Equipment ." International Journal of Performability Engineering 10, no. 2 (March 2014): 197-210
TECO-Westinghouse. OPERATION & MAINTENANCE MANUAL FOR THREE PHASE INDUCTION MOTORS. Round Rock, TX: TECO-Westinghouse Motor Company