Insulation resistance - domestic and industrial

[Skogsgurra]

"Always check insulation before connecting to mains!"

This is in the bone marrow of most electricians. Some use a Megger, some use an ohmmeter. The latter method is not quite what I would call checking the insulation - but it has one advantage: it doesn't kill anything.

"A 500 V or 1000 V Megger doesn't either", I can hear someone say. "But you should connect all phases, and neutral, too. Then you are safe..."

Am I? I remember a paper machine drive back in the seventies. A distributed installation with miles of cables. The capacitance of all these cables combined was in the close to one microfarad range. So when the guy that cranked the Megger saw a quite low resistance, but increasing, he thought that something was wrong with the Megger.

I told him to crank on and voltage rised slowly. And then resistance fell to zero. There had been a flash-over somewhere in the plant. God knew where. New cranking and new flash-over. I stopped the excersise and we did a separate check on the high amp circuits to make sure that the flash-over didn't happen there.

The machine was started and I went to the hotel for a much-coveted night's sleep.

The phone rang just as I had fallen asleep. There was something wrong in the machine. I went there again to have a look. There was a little free-wheeling diode across a relay coil that was dead short. And there were traces of a flash-over between coil pin and chassis (yes, relays were monted in sockets that were monted in metal chassis at that time). So, know we were two to know.

We had been very careful. We had tied everything together. But that doesn't help when the whole system capacitance decides to discharge through a tiny diode. A few volts would have been OK, but we probably were close to 1000 V this time.

So. Whatever the "specialists" say about meggering and what little harm it does "if you do it right": DO NOT LET THEM DO IT! The distance between live parts and grounded parts is not any greater today - I would say probably smaller.

Next thing to consider: What shall the insulation resistance be? What is an acceptable level? What do the different Codes say?

I have had this discussion so many times. So I leave it open for you to join in. The European codes usually say that the resistance shall be Ux1000 ohms, where U is system voltage. An industrial installation with lots of inverters and other electronic loads connected to a 690 V grid should thus have a 690 kohm insulation resistance. Which is absurd.

How do you handle these things? Do you measure one wire at a time or do you measure all wires connected? Do you measure with load connected or do you measure "between fuses" i.e. one shrt piece of wire at a time? And what limits do you use?

It is about time that we bring some order to this swamp. Please help!




Gunnar Englund

http://www.gke.org

 

[rbulsara]

As far as I know meggering should only be done on disconnected cables or windings...and I woul

If a switchboard..all control fuses should be taken out..

 

[Skogsgurra]

A couple of corrections:

"So, know we were two to know" Shall read "So, now we were two to know".

shrt = short



Gunnar Englund

http://www.gke.org

 

[RalphChristie]

Skogs

I megger a lot of cables, motors, transformers etc. up to 66kV. (At least once a week, mostly trailing cables, and most of the times due to mining activities) One thing I noted is that it depends a lot on experience to decide how, when and where to megger and how to decide if the readings are correct. A lot depends also what type of equipment you are meggering. I have also to add, most of the times when I am meggering equipment, it is not on new installations, but for faultfinding purposes. And the best way to learn is to do it practically.

Personally, I do not like these electronic, digital testers of today. My first choice would be a "crank" megger, and if it is an electronic one, it must have an analog display. (but this is my opinion)
If there is any sensitive equipment connected, disconnect it from what you want to megger.
Most failures are between a phase and earth. Remove any possible grounding points (link between star-point of trsf and ground, resistor between pilot-wire and earth, etc.) and do not forget to remove your own safety earths.
Remember if you megger cables, especially long cables with high voltage ratings, to discharge the phases after you have meggered it. You will never forget it when it discharge through you. (believe me, you will not let it happen twice if you feel how a 11kV cable "feels" if you are the discharger!)
Megger long cables for a time, because it has to charge up. Many times the resistance first falls to zero after several minute's meggering if there is a fault.

Now for values. A rule of thumb is not less than 1 M-ohm, but this depends on equipment. I have seen a lot of submersible pump-motors with very low values to ground, but what work for many years after I have meggered it. You have to have a feeling on what is wrong and what is right - and this comes just with experience. Maybe not the answer you are looking for, but this is how I am experiencing it.

Regards
Ralph

I do not know if megger, meggered, meggering, etc. are valid words, but we use it if we are doing insulation resistance tests, even if the tester (equipment) is not a Megger.

 

[Skogsgurra]

Thanks Ralph,

Yes. I know the feeling. I was the "discharger of a 70 kV cable once.

But I am talking about low voltage installations here. Such as the ones you find in a typical industrial installation (in our part of the world anything up to 1000 V AC is considered low voltage).

The reason that I bring this up is that the IEC 60364 part 6-6 (initial verification) says that the lowest acceptable resistance is 500 kohm for an installation with less than 500 V and 1 Megohm above 500 V.

I try to use my judgement when evaluating the status of an installation. But judgement is one thing and what the "law" says is another thing.

There are so many more or less "private" interpretations of the 60364 that a meaningful discussion is hard to carry through. The "connect all phases and neutral together" is one such issue. Another one is if the measurement shall include all segments of the installation (from busbar down to the smallest lamp fixture) or only one segment at a time (i.e. busbar, then cables, then wires for the different devices connected to the system). There doesn't seem to be much consensus there.

Lots of inputs needed here. Come on! Tell us how you do it - and why!



Gunnar Englund

http://www.gke.org

 

[ScottyUK]

Skogs,

Our 'official' standard is as you describe - 500k[Ω] below 500V and 1M[Ω] below 1000V. If I saw a new cable indicate remotely near those values I would be asking 'why?'. When a cable is connected to equipment, those values might be more acceptable. For instance, if a motor tripped on earth fault, as a quick check I'd megger the cores and the attached stator to ground to determine whether a more detailed investigation was needed.

If I was concerned about a busbar or switchgear installation I'd disconnect incoming and outgoing circuits or open the associated breakers so the measurement was exclusively on the switchgear. I would be concerned to see readings that were not off-scale for modern switchgear. As you and Ralph have said, much of it is gut feeling about what is 'good' and 'bad'. Readings which wander about make me suspicious, as they often indicate dirt or water ingress, even though they may be in the 'good' M[Ω] range.



----------------------------------

If we learn from our mistakes,
I'm getting a great education!

 

[Andy32821]

Hi Guys,

I'm going over to Meg some 480-volt feeder cables for a small commercial facility. I thought I would I would beef up the report by quoting 2005 NETA standard. I was amazed to find the lower limit is 100 megohm at 1000 volt.

I come from an industrial paper mill where any 480-volt cable was “good to go” above 2 meg.

I realize that the rest of the world does not operate like a paper mill, but do you guys recommend a customer pull new cable in a commercial facility or even a public assembly building if a 480-volt feeder cable megs 30 megohm?

Seems like overkill but I wish to conform to the current “standard of care”.

Plus more work for the electricians.

Does the insurance industry have a post water intrusion lower resistance limit that they are willing to pay for cable replacement?

Thanks,

Andy

 

[wfowfo]

Does anybody still utilize a P.I. (polarization index)?
Forgive my memory here because it has been a long time since I've done any meggaring.
An ideal insulation is capacitive in nature, not resistive. Polarization idexes (if I remember right) are the 10 minute megohm reading divided by the 1 minute megohm reading. This ratio is an indication of the rate of charging of the capacitive characteristics of the insulation. The closer this ratio is to one, the worse your insulation is. The pass/fail ratio eludes me at the moment (perhaps from being the discharge point for too many cables?) so someone feel free to jump in here with the details.

 

[eemotor]

Polarization index is still used to check condition of motor winding insulation. The capacitive charging current usually decays to a negligible value before the first measurement at 1 minute is taken. There is also absorption current that's present. However, this also decreases to nearly zero. What's left at the 10 minute point of the test is the conduction as well as surface leakage current which are fairly constant. A polarization index of 1.5 or higher is usually satisfactory as far as motors go, but may vary a bit when it comes to cables.