Burn't GFCI recpt fix proper?

Status
Not open for further replies.
ritelec

ritelec

Senior Member
Location
Jersey
Hello all,

Just want to confirm I did right with this repair (with the exception of installing an arc fault interrupt ??)

Gentleman called about exterior gfi recpt with "scorch" marks.

I went, yes there where scorch marks on building (couple inches).

Vertical GFI fs box cvr. Flush mounted. Plastic old work box.

Cover was alittle scorched, opened cover receptacle face was Burnt.

Removed cover and receptacle.

As I had the receptacle in my hand and was pulling it out of the box, there was sparking and flames running around across the bottom of the receptacle (line connection side, no load wires, hot was detached and floating, and arcing also across the receptacle plastic) and the internal area of the box.

I'm not sure if it was a loose connection that did this. I can't tell because of the melt down but the screw on the receptacle still seemed tight. The hot was off. Could it have broke off from maybe a poor strip and knick in the wire? The back of the receptacle has had a great deal of melt down at the bottom where the line connections were made, but the face of the receptacle seems that the fire burnt up and the down part was un effected.
Any possibility that the internal electronics burnt and it wasn't a loose connection or poor termination (any known issues or recalls?)? The owner said the receptacle and addition went on in the 90's.

The breaker. It did not trip as I'm speculating there was no ground fault and the sparking and flames where "just" a load on the circuit but no "overcurrent".
Is carbon that combustable that fire was just running around the box as it did (wish I had a picture of that).

I'm also speculating that an arc fault device would have picked this up and de-energized the circuit??


I stripped back the wiring to non damaged nm. Replaced the plastic box. Installed new gfci receptacle.
Should I feel good with this repair? Should I install an arc fault, or is it just another 15 amp non arc fault protected circuit like in the rest of the house and world?

Thank you
 
Last edited:
This is an incident the AFCI manufacturers want you to believe you are protected from when using their devices. Whether or not they would open the circuit is still out for debate, and maybe some do open the circuit and some may not.

In general 120 volts is not quite high enough to sustain such arcing, you must have had something present that was conductive enough to allow this to happen. The fact it was an outdoor receptacle maybe does contribute to having some moisture and conductive contaminants being present that otherwise may not have been likely if this were in a completely dry location.

Also AFCI protection though not prohibited is not NEC required for outdoor receptacles.
 
I just fixed a burnt up receptacle and melted wire ends last week for someone.
Now the circuit works where before it did not.
I feel great about it because it wasn't going to get any better. I fixed it.
 
I don't think the AFCI would have helped.

Was there a heavy load connected to it?

It could very well been a short in the GFCI but not a dead short to trip the circuit breaker. And as you were pulling the device out you saw the sparks on the back of the device.
 
Last edited:
kwired said:
In general 120 volts is not quite high enough to sustain such arcing, you must have had something present that was conductive enough to allow this to happen. The fact it was an outdoor receptacle maybe does contribute to having some moisture and conductive contaminants being present that otherwise may not have been likely if this were in a completely dry location.
Click to expand...


?? I don't know. Yes it does seem extreme for 120. But she was a flaming :happyyes:

Contaminates? Moister? It's located about 18 inches above a four foot or so high deck under a two foot or so soffit overhang. I don't know??

It did look though like any melting was from the rear and not front where something was plugged in. (and I don't think it was just melt but an actual fire. And contained as I guess the box was designed :thumbsup:

Don't know what caused it but I "think" yesterday flame ball was ''carbon" related" ??


"The low-weight end of the group is flammable in low-density forms. Carbon powder is highly flammable in fine grades. Light-weight charcoal burns but is hard to ignite. A 3-inch single chunk of dense graphite can oxidize if heated but will not sustain its own fire when the heat is removed. The next heavier element in the group, Silicon, can burn as very fine powder, but pieces any larger (like sand) will grow a self-adherent surface oxide (quartz-glass!) which usually stops combustion. Germanium has a similar adherent surface oxide and lower oxidation energy release than silicon. Not what you would call flammable. Tin and Lead are low-melting metals with no flammability to speak of.

All of these things could add some energy to a pre-existing huge fire."
 
edward said:
Was there a heavy load connected to it?
Click to expand...

Don't know, the owner mentioned he noticed the "scorch" marks this past summer. I'm thinking not. They are an elderly couple so they don't use appliances that draw lots of current :blink:
 
ritelec said:
?? I don't know. Yes it does seem extreme for 120. But she was a flaming :happyyes:

Contaminates? Moister? It's located about 18 inches above a four foot or so high deck under a two foot or so soffit overhang. I don't know??

It did look though like any melting was from the rear and not front where something was plugged in. (and I don't think it was just melt but an actual fire. And contained as I guess the box was designed :thumbsup:

Don't know what caused it but I "think" yesterday flame ball was ''carbon" related" ??


"The low-weight end of the group is flammable in low-density forms. Carbon powder is highly flammable in fine grades. Light-weight charcoal burns but is hard to ignite. A 3-inch single chunk of dense graphite can oxidize if heated but will not sustain its own fire when the heat is removed. The next heavier element in the group, Silicon, can burn as very fine powder, but pieces any larger (like sand) will grow a self-adherent surface oxide (quartz-glass!) which usually stops combustion. Germanium has a similar adherent surface oxide and lower oxidation energy release than silicon. Not what you would call flammable. Tin and Lead are low-melting metals with no flammability to speak of.

All of these things could add some energy to a pre-existing huge fire."
Click to expand...
I not a chemist, but plastics are made from petroleum products, start burning them somehow and you are likely to release some carbon from the plastics. You do however need some catalyst to get the whole thing going - probably heat from either a weak connection or a malfunctioning logic component in the GFCI. The plastic as manufactured is a good insulator, but add some heat to break it down and who knows what may happen next. Again 120 volts is usually low enough voltage it will not sustain arcing, but if conditions are right and enough continuity is present and for some reason the conductive component is not consumed/destroyed in the arc/fire it could continue to arc.
 
edward said:
It could very well been a short in the GFCI but not a dead short to trip the circuit breaker. And as you were pulling the device out you saw the sparks on the back of the device.
Click to expand...

Yes.. as I pulled it out the arc went from the hot broken off wire (make have been off may have been hanging on by a thread and broke off, not sure) across the plastic of the receptacle to the neutral side then continued into a flame clockwise into and around the box. It was building up energy and arcing then stopped as the distance between the wire and receptacle increased. I would say it was arcing at about 1/8 to 1/4 of an inch then stopped at about 1/2 inch. Was interesting.
 
Buck Parrish said:
I speculate the gfci was also protecting a bath outlet. They likely had a space heater in the bath.
Click to expand...

The OP states no LOAD side connections.

ritelec said:
Yes.. as I pulled it out the arc went from the hot broken off wire (make have been off may have been hanging on by a thread and broke off, not sure) across the plastic of the receptacle to the neutral side then continued into a flame clockwise into and around the box. It was building up energy and arcing then stopped as the distance between the wire and receptacle increased. I would say it was arcing at about 1/8 to 1/4 of an inch then stopped at about 1/2 inch. Was interesting.
Click to expand...

WOW I did not think is was still arcing while you were working on it.
I have seen this meltdown in GFCI's.
I think it is the internals melting down. The ones I have come across had no loads and were in a vacation home. We think there was a surge or the POCO lost a nuetral..
 
Sierrasparky said:
I think it is the internals melting down. The ones I have come across had no loads and were in a vacation home. We think there was a surge or the POCO lost a nuetral..
Click to expand...

Thanks for the info, the owner did mention in our phone conversation that maybe it was lightning.

I couldn't say yes or no, I could only ask if anything else in the house was affected...
 
With 120 Volts and some carbon, you can get a very nice arc. Think of carbon arc lamps, they put out a lot of light and heat and sustain a good sized arc with less than 120 V (since there is a ballast resistor).
 
141112-1326 EST

It is likely that all GFCIs have an MOV across the GFCI input terminals. If the MOV fails, lightning is a possible cause, then an arc ccould easily develop in that region, or just plain high heat. The MOV will be at the input terminals end of the unit.

Typical arc voltage drops are 10 to 50 V. An arc is a relatively constant voltage load over a considerable current range. A spark (lower current than an arc) is also a moderately constant voltage load over a fairly large current range. Because these are constant voltage loads there must be current limiting sonwhere.

.
 
gar said:
141112-1326 EST

It is likely that all GFCIs have an MOV across the GFCI input terminals. If the MOV fails, lightning is a possible cause, then an arc ccould easily develop in that region, or just plain high heat. The MOV will be at the input terminals end of the unit.

Typical arc voltage drops are 10 to 50 V. An arc is a relatively constant voltage load over a considerable current range. A spark (lower current than an arc) is also a moderately constant voltage load over a fairly large current range. Because these are constant voltage loads there must be current limiting sonwhere.

.
Click to expand...

Surge suppressors have these and I have yet to see one fail like I have seen these GFCI's fail.
 
Thanks. Sorry for not getting back sooner.

Interesting read gar. I'm all MOVed out.

Wonder if I smash the thing with a hammer and open it up if I could locate it and determine if it was the source.

IMO. mov=plausible

Have a great day all.
 
141113-1003 EST

ritelec:

Don't smash the GFCI. Remove the screws holding the unit together, and/or cut away the melted plastic to try to get to the area of the greatest heat.

You can see photos of a Leviton GFCI at my website http://beta-a2.com/GFCI.html . The third and fourth photos labeled P5 show the MOV at the right side (the input terminal end). Note that on this particular GFCI the MOV is close to the end and side of the Hot Input Terminal.

The MOV was bent to the right for the picture to make its shape more obvious.

If an MOV gets too warm, self or external heating, and the voltage is in a range to produce substantial heating, then thermal runaway occurs, and the MOV and close objects get destroyed.

I need to correct the web page to eliminate the multiple P5s.

.
 
Last edited:
Status
Not open for further replies.
Top