Insulation Testing II

[George Stolz]

I have a similar question to this, but I didn't want to threadjack him.

I am replacing old services and panels, formerly FPE. Since the FPE has such a great reputation for not tripping under faults and overloads, I decided it would be a good idea to megger all the aluminum feeders, between the services and the interior panels to see if the feeders needed replacement.

In my searching, I am familiar with the 1 Megaohm rule of thumb, but these feeders are all over the boards, and nowhere near 1 megaohm, they're well over that. The lowest I've seen so far is over 100 megaohms, IIRC. Am I just lucky, or am I not being picky enough? The copper feeders I installed yesterday instantly maxed out beyond 11 gigaohms, which got me thinking maybe I'm not being picky enough.

I am just testing each ungrounded conductor to the EGCs, with the breakers in the panels off, to test the feeder.

1. What would your minimum resistance be to recommend replacing the feeder?

2. Should I go ahead and test between ungrounded conductors too? Ungrounded to grounded? I figured the most glaring result would be from ungrounded to EGC.

All comments welcome, thanks.

 

[brian john]

I have a similar question to

I am replacing old services and panels, formerly FPE. Since the FPE has such a great reputation for not tripping under faults and overloads,

That was not the issue with the recall I was involved with the AIC rating was the issues.

In my searching, I am familiar with the 1 Megaohm rule of thumb, but these feeders are all over the boards, and nowhere near 1 megaohm, they're well over that.

Forget you ever heard there was/is a one meg rule.

The lowest I've seen so far is over 100 megaohms, IIRC. Am I just lucky, or am I not being picky enough? The copper feeders I installed yesterday instantly maxed out beyond 11 gigaohms, which got me thinking maybe I'm not being picky enough.

VERY GOOD document and move on assuming you allowed the reading to stabilize as you meggered, not just a spot test go, no go.

I am just testing each ungrounded conductor to the EGCs, with the breakers in the panels off, to test the feeder.

Test the neutrals. Typically phase to phase, phase to ground, neutral to ground, neutral to phase

1. What would your minimum resistance be to recommend replacing the feeder?

NEW conductors infinity of the meter you are utilizing, but I would like to see with new at least 100 meg also if there is a large difference between the conductors I would suggest investigating.

2. Should I go ahead and test between ungrounded conductors too? Ungrounded to grounded? I figured the most glaring result would be from ungrounded to EGC.

YES, grounded neutrals can result in power quality issues, GFP operational issues and possible life safety issues.

All comments welcome, thanks.[/QUOTE]

IM me if you would like or need any forms

 

[George Stolz]

Thanks, Brian.

The lowest I've seen so far is over 100 megaohms, IIRC. Am I just lucky, or am I not being picky enough? The copper feeders I installed yesterday instantly maxed out beyond 11 gigaohms, which got me thinking maybe I'm not being picky enough.

VERY GOOD document and move on assuming you allowed the reading to stabilize as you meggered, not just a spot test go, no go.

I tested for exactly thirty seconds for a benchmark. From the Stitch in Time book, I assumed that the reading would never truly stabilize.

I am just testing each ungrounded conductor to the EGCs, with the breakers in the panels off, to test the feeder.

Test the neutrals. Typically phase to phase, phase to ground, neutral to ground, neutral to phase

Currently I am testing just before re-energizing the feeders, so everything is already landed. So, my neutrals are all interconnected, my EGCs are all interconnected, and the main bonding jumper is in place in the service.

It also makes it easier as far as accessing the units; I walk through the units with the inspector after everything is landed, and as we inspect the new panels, I shut all the branch breakers off. Afterwards I walk back outside, megger the units, energize the feeders, and make one last trip in to turn the branch breakers back on.

Is it worth meggering the units earlier in the upgrade, before I've landed anything?

1. What would your minimum resistance be to recommend replacing the feeder?

NEW conductors infinity of the meter you are utilizing, but I would like to see with new at least 100 meg also if there is a large difference between the conductors I would suggest investigating.

Okay - I was testing both phases, and recording the worst reading of the two. If I have one at 800 MΩ and it's companion is 2.1GΩ, then that is indication of a problem?

And by investigating, this is all the investigating I can do without digging up the yard. The feeders generally run underground in PVC, make probably five nineties and reemerge inside the unit in EMT. Since I have replaced the panel, the end has been cut off and reterminated.

YES, grounded neutrals can result in power quality issues, GFP operational issues and possible life safety issues.

You've got a point, I should be more thorough to verify that my main bonding jumper is the only one present.

Edit:

IM me if you would like or need any forms

Love to see one, thanks!

 

[K8MHZ]

Okay - I was testing both phases, and recording the worst reading of the two. If I have one at 800 MΩ and it's companion is 2.1GΩ, then that is indication of a problem?

Are you testing from phase to ground?

If the lowest is really 800megs, I see no problem. Remember you are testing the insulation and each conductor will have different amounts of contact to it's surroundings and thus we expect that the readings won't be the same. It's the low impedance readings that get's the red flag.

Also, moisture is not your friend when it comes to insulation testing. More than once I had to use heaters to dry out the ends of cables that got wet and were screwing with my readings.

I am not sure what you are saying about your readings. Are you concerned with fluctuations during the test? Many times erratic readings indicate a moisture or gunk problem. The sticky stuff from tape will conduct at high voltages. Sometimes ends of conductors are taped and not all the tape gets removed.

Also, what type of megohmeter are you using and at what voltage? Do you trust that your meter is giving you an accurate reading?

 

[George Stolz]

Correction:

Resistance in GΩ ... Apartment Number

0.641...1
0.211...2
0.710...3
4.100...4
0.450...5
3.100...6
1.119...7
0.928...8
1.428...9
3.600...10
4.200...11
0.204...12
0.218...13
2.145...14
0.108...15
2.100...16
1.541...17
18
11.000...19
11.000...20

Unit 15 is my lowest reading, at 108 MΩ.

Units 19 and 20 are brand new copper feeders, since they were above ground.

I am testing each unit phase to equipment grounding conductor.

I am using a brand new Fluke 1507 at 1000VDC. I trust it, since I can't tell the difference from the taste.

 

[K8MHZ]

Correction:

Resistance in GΩ ... Apartment Number



Unit 15 is my lowest reading, at 108 MΩ.

Units 19 and 20 are brand new copper feeders, since they were above ground.

I am testing each unit phase to equipment grounding conductor.

I am using a brand new Fluke 1507 at 1000VDC. I trust it, since I can't tell the difference from the taste.

If there are no specs from the manufacturer the rule of thumb is NETA Table 10-1. 600 volt insulation tested at 1000 volts should be 100 megs or better to pass. Using that table even the 108 meg reading would be a pass.

Remember that temperature and humidity also affect the tests. I would hazard to guess that doing the exact same test again will yield different results. When I was testing at a new peaker power plant we had to record the temperature and humidity on the forms we used.

I would re-test the conductors with the lowest readings at a later time just out of curiosity.

Methinks you are taking the new fangled digital read outs too seriously. My Biddle has an analog meter and anything over 200 megs is tough to determine as the difference on the meter between 200 megs and infinity is less than a quarter of an inch. Insulation testing is a pass or fail test. I don't see the need for readings over 200 megs on a 1000 volt tester and the meter makers didn't either. It's just that the resolution of digital meters is higher and can display readings in the gig range that would normally be interpreted as infinity.

 

[George Stolz]

Found my first .5 MΩ feeder today...

 

[ptonsparky]

Aren't you a lucky fellow. Now the sales pitch begins as you try to explain why things still work but it really should be fixed.

Aluminum feeder in underground PVC? You may be able to find that with a fault locator if the Ford extraction tool does not work.

 

[Cow]

Found my first .5 MΩ feeder today...

Is that with the wires still landed on the breaker at one end or do you have the individual wires isolated and capped off? No moisture on the ends or in the jacket?

 

[K8MHZ]

Is that with the wires still landed on the breaker at one end or do you have the individual wires isolated and capped off? No moisture on the ends or in the jacket?

Ya beat me to it.

If you are testing conductors then they should not be landed during testing. If it were me, the next step would be to determine the integrity of the ends of the conductor in question. Check to see if they are clean and dry. Do a re-check.

 

[ptonsparky]

Wires that are tested while connected to a starter, that has been used, may indicate a lowerer resistance than with one that is new. A lot of times you can see the carbon that has built up from normal use in a starter. I would assume a circ breaker would be the same.

 

[brian john]

Wires that are tested while connected to a starter, that has been used, may indicate a lowerer resistance than with one that is new. A lot of times you can see the carbon that has built up from normal use in a starter. I would assume a circ breaker would be the same.

CBs are
Typically we megger Line to load with the switch, CB, contactor open - A to A, B to B, C to C, then Line and load A to B A to C, C to B, then A, B, C line load to ground.