Motor Maintenance: High Potential Testing

[wiigelec]

The debate over whether or not to high potential test motors as a "predictive" maintenance task has proponents on both sides, to test or not to test.

I am looking for personal experiences, do you use high voltage testing (DC/AC hipot, surge testing) as part of your motor maintenance program? Why or why not?

Thanks.

 

[zog]

What type and size motor are you refering to? Have you looked at the IEEE and ANSI test standards that apply?

 

[wiigelec]

What type and size motor are you refering to?

480V, 4160V, 13.8kV, 300hp-6000hp induction. Also 4160V and 13.8kV synchronous generators 4MW - 15MW...

Have you looked at the IEEE and ANSI test standards that apply?

Basic familiarity, attempting to gain access to the docs via corporate subscription...

 

[zog]

480V, 4160V, 13.8kV, 300hp-6000hp induction. Also 4160V and 13.8kV synchronous generators 4MW - 15MW...

Basic familiarity, attempting to gain access to the docs via corporate subscription...

Not sure about the "predictive" nature of a hipot test, pretty much a pass/fail test, and a required test if I recall for machines >2300V per IEEE 95

A power factor test would be more of a "predictive" test IMO.

 

[wiigelec]

Not sure about the "predictive" nature of a hipot test

Predictive in the sense of finding a weakness in the groundwall insulation of the stator, or in the case of the surge test the turn-turn insulation, in the hopes of diagnosing failure during an outage and being able to replace the motor before online unscheduled failure.

These tests are "potentially destructive" in that marginal insulation may be totally compromised during the testing. I am aware of the risk/reward nature of the high potential testing, which is why I am looking for individual experience with these tests.

If you use high potential testing for predictive maintenance on motors, what criteria did you use to justify the tests?

Thanks.

 

[zog]

Predictive in the sense of finding a weakness in the groundwall insulation of the stator, or in the case of the surge test the turn-turn insulation, in the hopes of diagnosing failure during an outage and being able to replace the motor before online unscheduled failure.

"Predictive" is usually a term used for a condition assesment test, where you can predict the time to end of life and grade equipment for priority of replacement and budgets. I would not call a hipot a "predictive" test. But all symantics, I get what you are asking, there are better ways to predict life expectancy.

These tests are "potentially destructive" in that marginal insulation may be totally compromised during the testing. I am aware of the risk/reward nature of the high potential testing, which is why I am looking for individual experience with these tests.

I agree, there are better methods and equipment to use depending on the specifics of the motors, all covered in the ANSI/IEEE specs.

If you use high potential testing for predictive maintenance on motors, what criteria did you use to justify the tests?

Thanks.

My criteria is that it is on the purchase order I am a testing guy, not an owner of equipment. Have you looked into PdMA?

 

[wiigelec]

Have you looked into PdMA?

I am trying to decide between the PdMA and Baker test instruments, hence the question

 

[RESI]

The debate over whether or not to high potential test motors as a "predictive" maintenance task has proponents on both sides, to test or not to test.

I am looking for personal experiences, do you use high voltage testing (DC/AC hipot, surge testing) as part of your motor maintenance program? Why or why not?

Thanks.

wiigelec- I work for a large utility and am in the process of trying to establish a predictive motor testing program using Baker's AWA12 surge testing machine. We also purchased Baker"s PP130 power pack that would allow us to test motors that operate at 7200 volts. Test voltage for a 7200 volt motor is 2 times operating voltage plus 1,000 volts. This particular machine can perform 5 separate tests, either manually or automatically. The hi-pot test is AC and therefore not destructive. The whole idea behind this program is that over time, we can reasonably predict when a motor will fail, thus saving money because we could then budget for a rewind rather than it bite us when on line or when demand is high. Most all motor failures start out with failure of the insulation between individual coil wires or turn to turn.

 

[zog]

Go check them both out at the Powertest conference in Feb. It will be in Dallas and we rented out Cowboys stadium for the welcoming reception.

 

[wiigelec]

I've seen the PdMA in action and the Baker folks will be onsite next week.

What I am asking for is personal experience with hipot testing of motors in a maintenance setting (when and why), as RESI has given. Thank you.

Anyone ever blown up a stator hipot testing?

 

[stew]

I in another lifetime was an owner operator of an electric motor rewind shop in Seattle. We wound motors up to 500 hp. In my opinion the only time a hi-pot tester should be used is on new motors and rewound stators if desired or required by the customer such as Navy spec. The use of hi-pot tests on older windings is very much a potential for setting up insulation failures especially in older windings that are already partly compromised by possible wet and heat affected briteleness. Although these windings may be still usable, when a hi pot is used it can cause flashovers and carbon tracking that will lead to failure. In the motor trade I dont think you will find any motorman who would use a hi pot on anything other than new or fairly new windings. We mostly shudder at the thought. Even the Navy doesnt require hi pot on anything but new windings. A surge tester is not normally an instrument that will cause any probelem and is in fact quite a good tool. Use of a good megger is actually in my opinion a better tool for the type of testing you are considering. A log of winding resistance to ground and phase to phase kept over a long period is in my opinion a much better predictor of potetial trouble than setting yourself up for failure using hi voltage testing.

 

[charlie b]

The "AC High Potential Test" will not give you any useful information in the context of predictive maintenance. It is a "go, no-go test," and does not provide trendable results. If the motor fails this test, you send it to the rewind shop, because the test will have destroyed the motor. If the motor passes this test, all you know is that the motor has not been destroyed yet.

A better test, if you are looking to trend results and predict failures, is the "DC High Potential Step Test." You apply DC voltage in increasing steps, until the onset of leakage current is observed. You stop the test at that point, so as to prevent actually damaging the motor. The trendable result is the voltage level at which the leakage current is observed. This value will drop as the motor ages, and a rapid drop (as compared to the last test result) is predictive of possible failure.

 

[stew]

absolutley with Charly on this . Hi pot is or can be destructive. Thats why his form of test and /or a megger log is much more usefull as a preventative tool.

 

[stew]

About a year ago I reinstalled a piece of equipment that was a package unit with several motors on it the largest of which was 150 hp. It had been manufactured without a ul listing and was nearly an exact duplicate of machines before and after that did have a UL label Noone at the factory knew why. The Inspector wouldnt pass it however so we we forced to have a UL field inspection and correct deficiencies.All were minor and were corrected however his last one wasnt and that was that the motors had to pas a hi pot test. I refused and told hime that I wouldnt be able to find any motor shop that would do this test on a motor that was 16 years old. Megger tests at 1000 vdc showed the indings to be perfectly acceptable to anyones specs. I had to request the test from 3 reputable shops in our area explaining to each of them the reason for the test and the age of the motors. 2 of the shops refused to test and 1 had a caveat in thier return quote that they would not be responsible for a failure. I was finaly able to convince the UL guy to pass this without a hi pot but only after he took it to his upper management peers who pretty much agreed that an old motor wasnt a good candidate for this test.

 

[wiigelec]

Thanks for the good information all...

I think we can be in agreement that high potential testing has its place in a motor maintenance program. Hipot testing at the 2V+1000 level for proving new motors, and DC step testing and surge testing as predictive maintenance tests...

 

[zog]

If you use meggeer readings to look at trending over the years it is very important that all of the readings are tempature corrected to 40 degrees C. Otherwise any trend datda is worthless.

P.S. The 2V+1000 rule was removed from all standards many years ago