Should this be worrisome?

[CanuckEngineer]

Hello All,

I have recently been asked to review the Independent Testing Organization (ITO) testing documents pertaining to one of the new projects I am working on.

What I have attached is a test sheet scan for one of the 27.6kV loadbreak switches (which comprises as a whole the 5 bay 27.6kV switchgear).

I have circled the large difference between A-A' versus B-B' and C-C' as well as large difference between A-G versus B-G and C-G.

The order of difference measured greater than 10x. Is this a concern? Or as a whole acceptable.

Thanks,

RK.

 

[marks1080]

Looks acceptable to me. The difference between phases is large, but the lowest reading is 44000 Mohms (44,000,000,000) ohms. I would accept this for a 27.7kV system.

Sometimes when you see drastic differences in resistances in tests like this it might be an indicator that something is failing... If this is a new installation I wouldn't think that. If you are still concerned try having all the equipment cleaned and dried very well then run the test. Amazing the difference a little bit of moisture and dirt can make on a hit pot test.

If you are still worried, schedule a maintenance test again maybe 1 or 2 years out and reference those data against the original.

Hope this helps,
Mark

 

[CanuckEngineer]

Thanks Mark.

Great points. The ITO did not flag it as a concern as well and it is good to know some tips on what may effect the readings.

Btw, the gear has been energized for some time.

Regards,

RK.

 

[oldfieldguy]

When you perform insulation testing with common test equipment and start getting readings in the multiple gigohm range, keep in mind that the tiniest amount of moisture, dust, oil from fingerprints, etc. can cause a big swing in the readings.

This makes it difficult to apply 'difference between like specimens' criteria. In my understanding, a lot of that sort of evaluation criteria is a holdover from times when test equipment simply didn't measure accurately above a gigohm or two. We used to write "insulation resistance exceeds the scale of the test instrument" and people were happy. Now we get to write 44 gigohms, 61 gigohms, 39 gigohms, and somebody gets to get excited over the difference.


old field guy

 

[ScottyUK]

Check the test leads themselves. Dirt - especially oily dirt - on the insulation can give really confusing results. I wipe mine with alcohol wipes if I'm making measurements in the high G[Ω] range.

 

[marks1080]

I don't know if there is any kind of standard to back this up, but I have often used the rule of thumb that you should get 1Mohm per kV, plus another Mohm margin at minimum. That being said, I might be conserned if I was close to this level depending on the scenario. For healthy equipment readings are generally much higher.

One very important thing is to keep all of this kind of test data so you can reference back in the future. You can catch failing equipment before it actually fails by observing trends.

 

[unclebob]

Most of the time, the conductors aren't disconnected from the switch when testing it.

 

[BigJohn1]

I agree with Oldfieldguy who said that above a certain resistance value even very minor leakage can cause a drastic swing in readings. This is the reason IEEE standards suggest exercising caution if a PI reading is above 5kMΩ, because minor fluctuations in test current can drastically affect your ratio.

To put OPs measurements in perspective: Between the 44kMΩ value and the 65kMΩ value you're only looking at a different of 30μA of current flow, which by itself wouldn't concern me.

Regarding the rule-of-thumb for 1kV per MΩ plus 1: It may be in other standards as well, but I know that comes directly from IEEE 43 for testing rotating equipment. Be careful with that, though, because on newer equipment they set a baseline of 100MΩ minimum, so if you find yourself significantly below that value, you should do some investigation.