Roofing screws through conduits, how to test live circuit insulation?

[CurSol1!]

Here's a good head scratcher for you guys and gals.....

I've been in the trade for over 40-yrs now and yet I cannot figure out how you can test to see if a live conductor that is nicked by a screw in a conduit and could be leaking to ground but yet not bad enough to trip a breaker.

One of our customers had a very large rubber roof (50,00SF) where the roofer has penetrated many conduits with their 6" rubber roofing screws.
My guys found a few in just the first look. We usually provide insulation resistance test reports using our meggars but always on de-energized circuits.
When this building was originally built, our code did not address conduits above bar joists and in the corrugated metal pan roofing and of course in this case, there could be upwards of at least thousands of screws, 50 or more conduit runs with hundreds of live circuits.
To make matters worse, we do not have as-builts to help us determine where all the branch circuits go and what are connected to them, so now that just by pulling the conductor off the breaker and inducing upwards of 1000V into the conductor with our meggar is going to result in countless printers, computer, tv's damage and who knows what..... so with that said:

Is there any way to test a live conductor's insulation resistance value (leakage to conduit/ground) with unknown loads on that conductor's circuit?

Playing the devil's advocate and hoping for some valued input:

If we turn off the breaker, could we read the resistance value from conductor to conduit? Whatever is plugged into that circuit is probably going to provide bad readings though, correct?

Do you think that if we data log amperage we may find some anomalies? Motor start or any load turned on is going to blow that theory.

Talk about a needle in a haystack, I told our customer to initiate an insurance claim against roofing company because a dead short could happen now or it can happen in a year if you have a nicked wire and the roof heats up or the insulation breaks down over time, you'll then get that fault. I feel bad for our customer, but we are at a stand-still as to how we can attack and dig into this one...

Norman

 

[hillbilly1]

Short of megging all of the circuits (which would be difficult without removing all loads) one possibility, though not fool proof, would be measuring current on the SBJ. Turning off feeds watching for current drop. Then divide and conquer until the offending circuits are found. A caveat though, it could be a leakage from a faulty piece of equipment too.

 

[Deltaforce]

Here's a good head scratcher for you guys and gals.....

I've been in the trade for over 40-yrs now and yet I cannot figure out how you can test to see if a live conductor that is nicked by a screw in a conduit and could be leaking to ground but yet not bad enough to trip a breaker.

One of our customers had a very large rubber roof (50,00SF) where the roofer has penetrated many conduits with their 6" rubber roofing screws.
My guys found a few in just the first look. We usually provide insulation resistance test reports using our meggars but always on de-energized circuits.
When this building was originally built, our code did not address conduits above bar joists and in the corrugated metal pan roofing and of course in this case, there could be upwards of at least thousands of screws, 50 or more conduit runs with hundreds of live circuits.
To make matters worse, we do not have as-builts to help us determine where all the branch circuits go and what are connected to them, so now that just by pulling the conductor off the breaker and inducing upwards of 1000V into the conductor with our meggar is going to result in countless printers, computer, tv's damage and who knows what..... so with that said:

Is there any way to test a live conductor's insulation resistance value (leakage to conduit/ground) with unknown loads on that conductor's circuit?

Playing the devil's advocate and hoping for some valued input:

If we turn off the breaker, could we read the resistance value from conductor to conduit? Whatever is plugged into that circuit is probably going to provide bad readings though, correct?

Do you think that if we data log amperage we may find some anomalies? Motor start or any load turned on is going to blow that theory.

Talk about a needle in a haystack, I told our customer to initiate an insurance claim against roofing company because a dead short could happen now or it can happen in a year if you have a nicked wire and the roof heats up or the insulation breaks down over time, you'll then get that fault. I feel bad for our customer, but we are at a stand-still as to how we can attack and dig into this one...

Norman

Use leakage current clamp meter putting both conductors through the meter detect ground fault current on damaged insulation. Individual circuits below roof may also check , locate damaged portion

 

[synchro]

Use leakage current clamp meter putting both conductors through the meter detect ground fault current on damaged insulation. Individual circuits below roof may also check , locate damaged portion

Yes, it would be worth doing this with a leakage current clamp meter capable of reading ~1mA and preferably less.
You could first put the clamp around all of the wires in a conduit. Then see what kind of variation in the leakage currents between conduits you are seeing. Starting with the conduits having the highest leakage, identify individual 2-wire and MWBCs and measure each of their their leakage cuurents. As @hillbilly1 mentioned, some equipment could be contributing leakage current and may need to be identified before blaming the leakage on the wiring.

 

[Deltaforce]

You could first put the clamp around all of the wires in a conduit. Then see what kind of variation in the leakage currents between conduits you are seeing.

EGC to spare, it reduce actual leakage current

As @hillbilly1 mentioned, some equipment could be contributing leakage current and may need to be identified before blaming the leakage on the wiring.

Better, off such equipment or all equipment with power in conduits

 

[synchro]

As @hillbilly1 mentioned, some equipment could be contributing leakage current and may need to be identified before blaming the leakage on the wiring.

Better, off such equipment or all equipment with power in conduits

Having the load drawing current will help identify any screw penetration to the neutral, due to its causing the diversion of some of the load current.
And so having the load turned ON, OFF, and even totally disconnected could provide additional information as the testing proceeds.