What happened here?

[rhamblin]

Attached are pictures of an overload unit that got hot. At first glance it would appear to be a loose connection on the B phase and as it loosened up it got warm. This runs a 40 HP pump. It was installed last summer and is used for the cooling of the building, so it didn't even run that much as we are in Wisconsin, and we don't cool that much in the winter. If you look closely you can still see a piece of the #6 AWG wire in the lug. The copper wire actually got so hot that it appears to have melted (separated just outside of the lug). The interesting thing is that 6 inches down the conductor from this point, the insulation was still flexible. The overload, I'm told never tripped and was never reset. So in my mind the wire got really hot, enough to melt the copper into to pieces, but not stay hot long enough to transfer any heat to further down the conductor. Any thoughts what would cause all of this? Someone mentioned lightning, which kind of makes sense as this combination starter is the shortest run from our 3000 amp service. But this was the only damage. If it were lightning, I would expect to see the lightning finding a path to ground somewhere. Any thoughts are appreciated.

 

[Fulthrotl]

Attached are pictures of an overload unit that got hot. At first glance it would appear to be a loose connection on the B phase and as it loosened up it got warm. This runs a 40 HP pump. It was installed last summer and is used for the cooling of the building, so it didn't even run that much as we are in Wisconsin, and we don't cool that much in the winter. If you look closely you can still see a piece of the #6 AWG wire in the lug. The copper wire actually got so hot that it appears to have melted (separated just outside of the lug). The interesting thing is that 6 inches down the conductor from this point, the insulation was still flexible. The overload, I'm told never tripped and was never reset. So in my mind the wire got really hot, enough to melt the copper into to pieces, but not stay hot long enough to transfer any heat to further down the conductor. Any thoughts what would cause all of this? Someone mentioned lightning, which kind of makes sense as this combination starter is the shortest run from our 3000 amp service. But this was the only damage. If it were lightning, I would expect to see the lightning finding a path to ground somewhere. Any thoughts are appreciated.

i'm guessing here, but i'd call it an internal failure of the overload, not from an external factor.

probably failed when the hole melted and let out the magic smoke.
magic smoke is a lot like SF6 gas on HV breakers.

 

[infinity]

Looks like the stab that connects to the bus was loose. I've seen them with the attachment bolt loose straight from the factory.

 

[Jraef]

That appears top be an A-B E1 solid state overload relay, so it would not directly sense the heat of a bad connection as a bi-metal one would. A loose connection then would concentrate the heat of the added resistance right at the terminal, then if mounted vertically, the heat would rise INTO the rest of the OL relay and the tang that is connecting it to the contactor. Heat doesn't tend to migrate downward very much.

What kind of wire was this by the way? Mechanical lugs like this are not rated for high strand count conductors. I have seen this issue many times as a result of people not understanding this and using extra flexible cable because it's easier to work with.

 

[infinity]

I think that the first guess would be a loose connection. Figuring out where it was loose is the hard part.

 

[Jraef]

I think that the first guess would be a loose connection. Figuring out where it was loose is the hard part.

Agreed. But I worked for a number of years with a PE that did forensic investigations of electrical equipment failures and accidents. One thing I picked up from him is that you can usually tell where something started by looking at the damage path from worst to least, and the source will almost always be the worst, but following the rule that heat rises. So in this case the wire lug is the worst and was likely at the bottom, the tang for the contactor was likely at the top and shows less damage, so it was likely collateral to the bad wire connection.

 

[rhamblin]

Thanks for all the reply's guys. It was mounted vertically, and had THHN wire under the lugs. And it was an Allen Bradley E1 Plus.


Sent from my iPhone using Tapatalk

 

[steve66]

Agreed. But I worked for a number of years with a PE that did forensic investigations of electrical equipment failures and accidents. One thing I picked up from him is that you can usually tell where something started by looking at the damage path from worst to least, and the source will almost always be the worst, but following the rule that heat rises. So in this case the wire lug is the worst and was likely at the bottom, the tang for the contactor was likely at the top and shows less damage, so it was likely collateral to the bad wire connection.

Thanks for all the reply's guys. It was mounted vertically, and had THHN wire under the lugs. And it was an Allen Bradley E1 Plus.


Sent from my iPhone using Tapatalk

So with it being regular THHN wire, I'm going to guess maybe the wire lug was never tightened properly?

 

[Jraef]

Thanks for all the reply's guys. It was mounted vertically, and had THHN wire under the lugs. And it was an Allen Bradley E1 Plus.


Sent from my iPhone using Tapatalk

Thanks for the update.
THHN should have been fine, so it was most likely just not tight enough. That size unit needs 40 lb-in of torque.

My guess:
Cell phone rang right when the installer was supposed to be paying attention...

 

[JFletcher]

Loose connection was my first guess.

 

[wwhitney]

but following the rule that heat rises

My understanding is that hot air rises (convection), but heat transfer via conduction or radiation doesn't care about up or down. So is the upshot that in these forensic analyses, the heat flow via convection is usually a sufficiently large portion of the total heat flow that the damage will follow the rule "heat rises"?

Cheers, Wayne

 

[Sahib]

it would appear to be a loose connection on the B phase and as it loosened up it got warm. This runs a 40 HP pump. . The copper wire actually got so hot that it appears to have melted (separated just outside of the lug). The overload, I'm told never tripped and was never reset. Any thoughts are appreciated.

The extreme heat at B phase terminal of CB might have changed its resistance and so current flow to the pump motor and so there might be voltage unbalance/single phasing to the pump motor. If the pump motor had any unbalance protection, it might have tripped.

 

[Jraef]

The extreme heat at B phase terminal of CB might have changed its resistance and so current flow to the pump motor and so there might be voltage unbalance/single phasing to the pump motor. If the pump motor had any unbalance protection, it might have tripped.

That solid state OL relay has phase loss protection, but not phase current imbalance for levels above 20% of the setting (meaning it detects a phase loss when one phase is less than 20% of the set amount after the first 3 seconds). A slight imbalance increases the heat in the motor, but would not damage the OL relay terminal like that.

 

[ggunn]

My understanding is that hot air rises (convection), but heat transfer via conduction or radiation doesn't care about up or down. So is the upshot that in these forensic analyses, the heat flow via convection is usually a sufficiently large portion of the total heat flow that the damage will follow the rule "heat rises"?

Cheers, Wayne

I heard an interesting talk on the radio the other day by a forensic fire investigator about forest fires. He said that the burned area in a forest fire is usually roughly triangular in shape with the point of ignition at one of the triangle's vertices.

 

[Jraef]

I heard an interesting talk on the radio the other day by a forensic fire investigator about forest fires. He said that the burned area in a forest fire is usually roughly triangular in shape with the point of ignition at one of the triangle's vertices.

I heard that too, very interesting. I immediately thought about wind, ANY wind, being the reason for the spread to be angular, but triangular? That implies that at some point, the progression of the fire proceeded in a straight line opposite the starting point.

 

[Smart $]

I heard that too, very interesting. I immediately thought about wind, ANY wind, being the reason for the spread to be angular, but triangular? That implies that at some point, the progression of the fire proceeded in a straight line opposite the starting point.

He said burn area... not burn path.

 

[GoldDigger]

And some would be perfectly happy to call a sector of a circle triangular.....

 

[Smart $]

My understanding is that hot air rises (convection), but heat transfer via conduction or radiation doesn't care about up or down. So is the upshot that in these forensic analyses, the heat flow via convection is usually a sufficiently large portion of the total heat flow that the damage will follow the rule "heat rises"?

Cheers, Wayne

Actually, conduction also transfers heat [more] upward. Just nowhere near as much as it does with gaseous elements.

 

[Smart $]

Actually, conduction also transfers heat [more] upward. Just nowhere near as much as it does with gaseous elements.

Replace it with convection.

 

[GoldDigger]

Actually, conduction also transfers heat [more] upward. Just nowhere near as much as it does with gaseous elements.

I do not see the physics or thermodynamics to support that.