Overpotential test voltage for low-voltage switchgear

[AlexT90]

Should a overpotential test on a low-voltage switchgear be done with ac or dc? I've seen techs bring in a 5000v megger and basically do insulation-resistance at 1000v and then repeat the same tests at 2300v. This seems kind of dumb to me, wouldn't an ac hipot be better?

[randyw8483]

Should a overpotential test on a low-voltage switchgear be done with ac or dc? I've seen techs bring in a 5000v megger and basically do insulation-resistance at 1000v and then repeat the same tests at 2300v. This seems kind of dumb to me, wouldn't an ac hipot be better?

It is common to perform insulation resistance tests with a DC Megger instead of an AC Hipot. Low voltage gear is not tested with true overpotential methods as Medium Voltage gear per the NETA ATS/MTS. Once you get to medium voltage levels 4160/2400V and higher does the hipot testing come in handy as the max ouput on the Megger is generally 5kVDC. Testing with DC has its own advantages since the required output for cable/bus testing is a lot lower due to testing at unity power factor.

[BrickSalad]

I mean, generally you aren't going to overpotential test LV gear anyways, but if you're asked to then doing it with the same 5kv megger is more convenient than bringing in an AC hipot.

It may seem like you're repeating the same test at a higher voltage, but that's not really the case. The lower voltage test is to find any problems without damaging the gear. The higher voltage test is to fail any equipment that's bad. The different voltages make them quite different tests.

Where you want to use AC (especially VLF) instead of DC is for field-aged cables. The AC test isn't more informative, it's just less damaging to cable that passes. The reason why is a bit complicated, but the gist of it is that DC tests can leave a charge in older cable that causes significant damage when you re-energize.

[SecondGen]

I mean, generally you aren't going to overpotential test LV gear anyways

Dielectric withstand is required by NETA ATS for low-voltage switchgear (7.1.2.3) and this test should be done with AC unless any cables connected to the switchgear can't be removed (see explination below). Because this test is usually done during acceptance there are a few number of reasons, if any, not to remove the cables.


NETA ATS 2013 Table 100.2 The column headed “DC” is given as a reference only for those using dc tests to verify the integrity of connected cable installations without disconnecting the cables from the switchgear. It represents values believed to be appropriate and approximately equivalent to the corresponding power frequency withstand test values specified for voltage rating of switchgear. The presence of this column in no way implies any requirement for a dc withstand test on ac equipment or that a dc withstand test represents an acceptable alternative to the low-frequency withstand tests specified in these specifications, either for design tests, production tests, conformance tests, or field tests. When making dc tests, the voltage should be raised to the test value in discrete steps and held for a period of one minute.

[BrickSalad]

I stand corrected on the LV switchgear. I must have gotten it mixed up with LV switches (which you don't overpotential test). Thanks for that clarification :)

As far as DC vs AC goes, however, the wording beneath that table is a bit ambiguous. They don't actually say that you have to disconnect the cable if possible, nor do they say that a DC test is unacceptable. It's pretty easy to get that impression though, because they word it in a way to strongly suggest you shouldn't use the DC testing ever. NETA clearly doesn't recommend it, but they haven't yet explicitly forbade it either.

What's going on here is that the industry has been shifting away from DC testing ever since a couple of studies showed that it does more harm than good on aged cable. This doesn't apply to acceptance testing, but there is still evidence that VLF is slightly more effective at finding faults. DC testing works just fine, and there's still a lot of DC test equipment out there, so they haven't yet taken the final step to ditch the tables.

Now, do I recommend DC testing? No, not really. But for acceptance testing on low voltage switchgear, especially for short 1-minute tests, the increased practicality might outweigh the minor benefits of using a VLF. That's something you should probably discuss with the customer though.

[AlexT90]

Thanks for the response guys, you have provided valuable insight. Is there another standard other than NETA that talks about this in greater detail? You all make good points, but I'm leaning more to the AC side, mainly for acceptance testing. What do manufacturer's typically do/recommend?

[Kalbi_Rob]

Thanks for the response guys, you have provided valuable insight. Is there another standard other than NETA that talks about this in greater detail? You all make good points, but I'm leaning more to the AC side, mainly for acceptance testing. What do manufacturer's typically do/recommend?

This might be a tad late to reply, but this standard is where NETA derives its values from (C37.20.2-2015 for Medium-Voltage):

IEEE Std C37.20.1-2015 Standard for Metal-Enclosed Low-Voltage Switchgear

Table 1 Note b:
The column headed, "dc withstand" is given as a reference only for those using dc tests to verify the integrity of connected cable installations without disconnecting the cables from the switchgear. It represents values believed to be appropriate and approximately equivalent to the corresponding power frequency withstand test values specified for each voltage class of switchgear. The presence of this column in no way implies any requirement for a dc withstand test on ac equipment; or that a dc withstand test represents an acceptable alternative to the power frequency withstand tests specified in this standard, either for design tests, production tests, conformance tests, or field tests. While power frequency withstand tests are preferred, it is recognized that users may choose to use dc withstand tests for field tests of assembled switchgear for reasons of convenience and availability of test apparatus. When making dc tests, the voltage should be raised to the test value in discrete steps and held for a period of 1 min.