Your plant uses an industry association certified electrical testing firm to perform electrical infrastructure testing at the service and feeder level during your annual November holiday shutdown. Data from the testing is entered into the CMMS, and a report is generated to identify what needs repair or replacement. Repairs and replacements are then performed by a local electrical services firm during the December holiday shutdown.
The plant hired a new controller in June, and she had questions about this process. She had worked at several other plants, about half of which had electrical infrastructure testing programs. She said that in her experience, this level of expense for repairs and replacements seems outsized. She asked if you had analyzed the data over multiple years to identify any “repeat offenders.” Nobody had brought up this idea before, so the answer was no. So the analysis was done, and it turned out that most cables replaced one year were also be replaced the next year.
What might be going on, and how can you get to the bottom of this?
Answer to Quiz. First, determine which test and testing data are used to identify whether a cable is bad. Typically, an insulation resistance test is used, and the data are plotted over time. The numbers should get slightly worse each year and look like a very gently sloping (almost horizontal) line when plotted. Continuing on the plot, when the slope of that line makes a much sharper slant or “break,” that means the insulation is deteriorating at an accelerated pace, and the cable should be scheduled for replacement.
If it is this test, then when a new cable is installed the insulation resistance data history must be reset. Otherwise, you will see that increased slope in the plot. So you replace the cable, which is perfectly good, and due to the short notice and short timeframe for the work, you probably don’t take baseline testing data. And that’s a common policy to “save money” even though it’s costly in multiple ways. With bad data and late testing, your new cable can be untested but flagged as defective based on the data from the originally defective cable.
Any time new equipment or cabling is installed, always take baseline testing data at the time of installation. This will give you some history by the time the next testing rolls around, and it would also provide assurance that what you installed isn’t defective.
Cables can be damaged during shipping and/or installation. It isn’t necessarily the electricians causing the damage either. At a plant in northern Illinois, baseline testing data kept finding defective cabling. It turned out that some operators would take a shortcut through the work area during lunch break and walk on the cables. Once this problem was identified, there were no more defective cables found. Had the testing not been done, the plant would have started up with several cables that had marginal, at best, insulation integrity.
One particularly lame “reason” given for not conducting baseline testing is the cables are new. As we’ve just seen with the Illinois plant example, that’s not relevant. With equipment, a “reason” is the equipment is under warranty. But no warranty will cover the loss of revenue or the endangerment of people and property resulting from equipment failure.
So set up clear rules on how testing data will be stored, archived, and used — and never allow an installation to go untested before putting it into service.
But what if the electrical services firm is doing shoddy work? Installation testing will show this (on the off-chance a licensed contractor does shoddy work at your facility). It will also exonerate them if the issue is something else. And it could be unrelated to bad data management or installation, but you can’t know that if you have bad data management.
If the plant has a spring or summer shutdown, try to get repair/replacement work scheduled for then instead of only a month later. This gives much more time to gear up for what must be done.
But what about those other issues not related to bad data management or bad installation? Here are some to check for:
- Transient spikes. The best way to catch these is with a power monitor. If you don’t have one installed, then use a portable power analyzer that can be connected for at least 24 hours to collect event data. A large motor starting across the line can produce cable-damaging transients; this problem can usually be solved with a soft-starter or electronic drive.
- Deteriorating agents. If anything oil-based comes into contact with the cable insulation, you can expect damage. Commonly used lubricants such as grease or spray oils can cause this. Commonly used solvents may also cause damage, always check for compatibility if using around electrical cables.
- Insufficient ampacity. This will cause the cables to run hot, which will cause the insulation to degrade. Don’t just do calculations from the NEC ampacity tables to figure this out. Also account for voltage drop unless the cable is fairly short. Remember that power monitor (or analyzer)? Check the power factor at the load end of each suspect cable. Often, power factor is corrected at the service to avoid a utility penalty but it’s still too low at the load and that means a higher current draw.
- Excess harmonics. A harmonic analyzer can show you if a given cable is also carrying useless current that heats up the cable.
