smoking GFCI

[junkhound]

WR GFCI installed on puget sound shoreling, appox 40 feet above high tide, Maury Island. Fully covered outlet boxes, the big ugly clear plastic type.

HO reported numerous trips, so turned off CB to area of outdoor convienience outlets.

Had HO turn on breaker with meter on outlet, dead, cannot reset, but GFCI makes buzzing sound for a few seconds then slight smoke, no 'snap' or other arcing sounds, just the faint buzz and slight smoke.

Replaced the outlet, all is well

Out of curiosity, disassembled the GFCI. HO had reported a storm a few seeks earlier had thrown sea splash 50 feet high.
Salt water got past the terminals, enough conductivity on the plastic to enable low current arcing from GFCI solenoid to plastic.
High impedance of coil limits any high current arc, no breaker trip, just a hiss and smoke.

 

[junkhound]

> 150 views, no comments.
Has anyone else experienced a smoking GFCI without a CB trip?

 

[electrofelon]

Nice forensic analysis, thanks. I have not had any such experiences.

 

[Buck Parrish]

You're going to need to use a SWR - Salt Water Resistance - Lol

 

[junkhound]

You're going to need to use a SWR - Salt Water Resistance - Lol

HO says first time they have seen salt spray that high, big windstorm at full moon high time

I did liberally smear the outside of the new outlet (used a NON gfci) hubble spec grade. I

It was interesting that the salt water ingress appeared to be thru the side terminals vs. the open outlet slots.

Perhaps TR/WR combo would work better in such a location.

 

[TheCats]

Did you replace the breaker with a CAFCI or regular GFCI? Or was it already on a GFCI breaker that didn't trip?

This should be a target case for AFCI protection. The deposits that you point out do appear to be from an arc, rather than just smoke from an overheating conductive path.

I'm curious if the salt spray intrusion is more likely to trigger an arc fault or a ground fault. That might a purely academic question if a standard receptacle has a big enough isolation gap, better drainage and fewer spots to trap salt water.

 

[garbo]

If it was my house would install a quality WR TR GFCI receptacle in a bottom feed FS box inside of a quality fiberglass box that is NEMA 4 rated and keep the door closed when not in use.

 

[LarryFine]

If it was my house would install a quality WR TR GFCI receptacle in a bottom feed FS box inside of a quality fiberglass box that is NEMA 4 rated and keep the door closed when not in use.

I would install a GFCI receptacle indoors to feed the outside receptacles, or use a GFCI breaker.

 

[Fred B]

> 150 views, no comments.
Has anyone else experienced a smoking GFCI without a CB trip?

Only with old federal, Curtis and Old cutler breakers

 

[Seven-Delta-FortyOne]

Had it happen with a standard receptacle on a GFCI protected circuit.

Took the receptacle apart and it crumbled to pieces on the vanity. Customer called me because it didn’t work, along with some lights.

Circuit was still live

 

[Little Bill]

Only with old federal, Curtis and Old cutler breakers

OP's GFCI was a receptacle, not a breaker.

 

[ELA]

> 150 views, no comments.
Has anyone else experienced a smoking GFCI without a CB trip?

I have had two such experiences. The coil damage you see can be due to a failure of a auxiliary mechanical contact to open which removes power from the trip coil. The trip coil has an aux contact that normally opens when the GFCI trips and its contacts open. The trip coil is only supposed to be powered just enough time for the contacts to open.

If that aux contact is corroded /damaged or fails for any reason the trip coil receives a continuous 120V and melts down and emits the smoke you see. The current drawn is in the range of 120v/50ohms =~ 2A which is not enough to trip the circuit breaker.
I have attached the circuit board from one such failure. The coil damage that can be seen is from it flaming for a moment, emitting smoke and melting open.

 

[Fred B]

OP's GFCI was a receptacle, not a breaker.

he ask if a smoked gfci could occur without tripping a breaker, thus my responce.

 

[ptonsparky]

Strange that I've never done an autopsy on a gfci device.

I have done breakers and agree with ELAs description of the coil failure.

 

[92VR4]

> 150 views, no comments.
Has anyone else experienced a smoking GFCI without a CB trip?

Not actively smoking, but yes, I've run across some catastrophic failures on multiple devices without a breaker trip. I do a lot of beachfront on the other side of the country and you just have to accept that anything outside is a "wear-item", especially on the ocean side of the home, especially with electronics. The salt air is relentless, even if salt water never touches it.

The best solution I've found is to use the Arlington InBox enclosures with regular WR receptacles on a GFCI or DF breaker. Eliminates some potential failure points both electrically and for water intrusion (when properly installed).