Near miss of the day .........

[VernB]

So, I'm sitting there in my office in the mill and I get the panic phone call "Get your tools and get down here!". When I get down to the hall and ask what the problem is, they just point up to the ceiling and there's an octagonal box on a conduit with the box all discolored and a lovely black scorch mark above it on the ceiling.

When I get the juice shut off and open it up, the inside of the box is burned to a crisp but the antiquated 50 year old Square D breakers never tripped!

Since this was part of an obsolete (but still energized) 240V lighting circuit, I killed the whole thing and ripped out about 20 feet of wire to insure that it would never be reactivated.

I still don't have any idea what kept it from totally going. It still gives me the willies to look at that pic.

Vern

 

[480sparky]

Sounds like you could save some company money by not having to buy that new welder they've needed.

"Yea, just hook it up to that old breaker in the panel there. I'll get a good spot-weld with that!"

 

[mdshunk]

I still don't have any idea what kept it from totally going. It still gives me the willies to look at that pic.

Hey, a 15 amp fault on a 20 amp circuit will go on like that for quite some time. I suspect very much that your conduit as EGC was only marginal, and not passing enough current to trip the breaker but enough to make a little bit of black in the box.

 

[LarryFine]

Or maybe the duration of the fault was not long enough.

 

[augie47]

burn baby burn

burn baby burn

The few times I've seen almost identical, it's been what Marc states. We used to maintain an old saddlry company that had 1950's wiring using RMC with double locknuts as the only grounding means. I could have taken a few identical pictures. (over the years we were able to pull EGCs in most of the conduits)

 

[LarryFine]

I'd bet dollars to doughnuts (I have no idea where that phrase comes from) that comparing the impedance of an EGC to a decently-installed conduit system (and a properly-bonded service) would result with the conduit being the better conductive fault-clearing pathway.

In other words, in my opinion, the same fault would have resulted in the same non-tripping of the breaker, simply because that event's fault-current level and/or duration never reached the trip point of that breaker (with the usual all-other-things-being-equal disclaimer thrown in).

 

[augie47]

Larry, by no means being argumentative with you (in a war of wits with with you, I am totally unarmed)..my experience was with a rigid conduit sytems some 50 years old in a totally wood frame muti-story building with fork truck traffic on each floor. Obvioulsy I have no idea if the original conduit system was "decently installed" (your term) but, over time at least, it had become a poor pathway for current. Oft-times you could spot arcing at the locknuts. Vern's picture so resembled that situation I could not help but comment.
I will yield to your experience and that of others, but from my perspective, you can't beat an installed equipment grounding conductor. I've seen a lot of loose locknuts and loose emt connections in my time.

 

[chris kennedy]

but from my perspective, you can't beat an installed equipment grounding conductor. I've seen a lot of loose locknuts and loose emt connections in my time.

100% behind this statement!!!

 

[480sparky]

I've seen a lot of loose locknuts and loose emt connections in my time.

As well as missing locknuts.

 

[mdshunk]

I'd bet dollars to doughnuts (I have no idea where that phrase comes from) that comparing the impedance of an EGC to a decently-installed conduit system (and a properly-bonded service) would result with the conduit being the better conductive fault-clearing pathway.

Absolutely true. (Georgia Tech Model). It's a struggle to find, however, a well installed RMC installation that has also remained that way as time passes.

"Being willing to bet dollars against doughnuts (viewed as worthless) means that you're totally confident that you're right, so confident that you'll bet money against nothing."

 

[chris kennedy]

Absolutely true. (Georgia Tech Model). It's a struggle to find, however, a well installed RMC installation that has also remained that way as time passes.

Marc, I've seen that study here a couple times. I don't disagree a properly installed conduit run has lower impedence. But I don't recall that study taking fittings into account.

 

[LarryFine]

Oft-times you could spot arcing at the locknuts. Vern's picture so resembled that situation I could not help but comment.

I've seen a lot of loose locknuts and loose emt connections in my time.

I cannot argue with these observations, and yes, they certainly do affect the ability of a conduit-based EGC system to function. I have also seen very old conduit system completely intact and fully functional that perform as intended and as when they were new. My comments are obviously based on the latter.

However, that much carbon shows a relatively high-energy arc flash. The question is whether the presence of a separate EGC would have made a difference in whether the breaker would have opened with that fault. Did any lock-nut arcing occur at that time? We may never know for sure.

 

[augie47]

great point

great point

However, that much carbon shows a relatively high-energy arc flash. The question is whether the presence of a separate EGC would have made a difference in whether the breaker would have opened with that fault. Did any lock-nut arcing occur at that time? We may never know for sure.

again, I profit from your knowledge...I had not given that thought.

 

[Pierre C Belarge]

Larry
Not arguing your point of the flash, but it was mentioned by the OP that the circuit was old. That old conductor insulation burns very black and it is possible that could have caused the black mess as well.

 

[VernB]

Sounds like you could save some company money by not having to buy that new welder they've needed.

"Yea, just hook it up to that old breaker in the panel there. I'll get a good spot-weld with that!"

Heh, don't even get me started on that.

Vern

 

[VernB]

Hey, a 15 amp fault on a 20 amp circuit will go on like that for quite some time. I suspect very much that your conduit as EGC was only marginal, and not passing enough current to trip the breaker but enough to make a little bit of black in the box.

This was a non ground referenced circuit (1 phase of an ungrounded 240V delta).

Vern

 

[VernB]

I cannot argue with these observations, and yes, they certainly do affect the ability of a conduit-based EGC system to function. I have also seen very old conduit system completely intact and fully functional that perform as intended and as when they were new. My comments are obviously based on the latter.

However, that much carbon shows a relatively high-energy arc flash. The question is whether the presence of a separate EGC would have made a difference in whether the breaker would have opened with that fault. Did any lock-nut arcing occur at that time? We may never know for sure.

Larry, this was a single phase of an ungrounded 240V delta, if it had faulted to the conduit (nothing was connected to the entire branch), it wouldn't have tripped the breaker anyways.

My opinion after examining it closely was that the failure was likely caused by overheating of lousy splices in that box.

Vern

 

[VernB]

Larry
Not arguing your point of the flash, but it was mentioned by the OP that the circuit was old. That old conductor insulation burns very black and it is possible that could have caused the black mess as well.

The wire was 14awg type R, canvas wrapped natural rubber. Burning rubber does tend to make a heckuva mess, that's for sure :grin: . What really gave me the heebie jeebies was the scorch on the ceiling. The ceiling in that area is some sort of an old fashioned fiber wallboard, rather than sheet rock. I'm glad I didn't have to find out how flammable it was the hard way :roll: .

Vern

 

[VernB]

Larry, by no means being argumentative with you (in a war of wits with with you, I am totally unarmed)..my experience was with a rigid conduit sytems some 50 years old in a totally wood frame muti-story building with fork truck traffic on each floor. Obvioulsy I have no idea if the original conduit system was "decently installed" (your term) but, over time at least, it had become a poor pathway for current. Oft-times you could spot arcing at the locknuts. Vern's picture so resembled that situation I could not help but comment.
I will yield to your experience and that of others, but from my perspective, you can't beat an installed equipment grounding conductor. I've seen a lot of loose locknuts and loose emt connections in my time.

This was certainly old rigid, building is a 154 year old mill, brick shell with wood interior. The grounding of the conduit is kind of moot in this case, since the circuit wasn't ground referenced to start with (240V delta), but you make a darned good point. I've been retiring a lot of the old rigid but I'll have to check a few box connectors for curiosities sake.

Vern

 

[LarryFine]

Larry, this was a single phase of an ungrounded 240V delta, if it had faulted to the conduit (nothing was connected to the entire branch), it wouldn't have tripped the breaker anyways.

Okay, so now that we know it was a line-to-line fault, the conduit-conductivity discussions are moot. What is definite is that the current in this event did not rise fast enough and/or high enough to trip this breaker.

The OP was centered around the breaker not tripping, and suggested that it may have failed to perform properly. It also assumed that the conductors remained in contact, but I do not agree with that assumption.

I am saying that the arc cleared itself quickly enough to not reach the trip curve of the breaker. Whether that fell within the design parameters of that breaker cannot be determined without examination.

To my field-educated eyes, the maginitude of the soot indicates that there was a small explosion, and not a (relatively) long-duration fire. I believe the arc extinguished as fast as it started, and no tripped breaker.

As for starting a fire, In my opinion, the conduit system performed its function, which is as much to protect the building and people from damage by the wiring as it is to protect the wiring itself from damage.