Insulation (Megger) testing of conductors

[mez1st]

We had an incident occur at a customers location this past week. Parrllel 500 kcmil in seperate 2 1/2" EMT install 50' between panelboard and customer equipment, 480v 800A three phase. The conductors were all tested using our Ideal 61-791 insulation tester, which showed no problems. the 800A breaker was closed and the circuit energized. 2-3 hours after energizing, the wire short circuts inside the conduit and blows a dime size hole and an arc flash. Fortunately no one was in the vacinity of the explosion and the breaker operated correctly and tripped as intended.

Here is the question- has anyone had a problem like this where the conductors insulation failed even after a megger test? We are planning to investigate the test procedure that the tech used and the method of wire installation. We are also having the 61-971 inspected and recalibrated. ANy other suggetions would be appreciated.

 

[don_resqcapt19]

You can have condutors with completely bare spots that will pass an insulation resistnace test if there is no contamination...then after a lenght of time when water works it way in the condutor will fail. The time frame is much too short for that in this case. I would guess you have a faulty tester.

 

[renosteinke]

Folks can mean several different things whan they say 'megger' testing. I suppose the key question to as is: was there any documentation present of past tests, to which the new test results were compared?

Here's a little primer, for those who have not explored the full capabilities of megger testing:

First, there's the 'quick and dirty' test, where the megger applies perhaps 600 volts to the wires, looks for minute amounts of leakage, and gives you an instant 'pass/fail' answer. This is a good test for troubleshooting problems, and prooving new runs ... but it's of no use in identifying how close something is to a future failing. This is the only test the cheapest meggers can perform.

Fancier meggers, when supported by records of past tests, can predict the future failure of conductors that are, at the time of testing, still OK. The two tests used here are the 'polarization index' and the 'saturation test.' Both involve applying the charge for a significant period.

In the 'polarization index' test, successfully higher voltages are applied until some leakage is detected. A given wire might start to 'leak' at, say, 1350 volts, for several years. Then, as the insulation deteriorates, you will see this figure start to decline just prior to complete failure. To make sense of this test, you must have records of previous tests.

The 'saturation test' treats the wire as if it were a capacitor; a charge is applied, and you see how much of the charge is still there after, say, 20 minutes. Again, you will see a sudden drop in your test results as the insulation approaches failure.

I suspect that only the 'quick and dirty' test was performed, and the 'weak' insulation was not found.

 

[mez1st]

This was a new installation (so there was not a previous test to campare to) which only had the quick test perform as I believe that is the limitation of the unit we have. It was an overhead run of conduit in an electrical room, so I do not think moisture was an issue. Hopefully we will know about the tester after having it checked.

 

[brian john]

What was your test voltage?
What was the duration of the test?
What are the results of the test?

A-Gnd
B-Gnd
C-Gns
A-B
B-C
C-A