GFI Caused a Fire!

A handle tie might be acceptable if AFCI breakers only detected arc faults. But they also include ground fault protection, which requires the neutral to go through the breaker, which means you need a 2-pole breaker for an Edison circuit.

The alumiconns are the ones that require a trained, licences, and certified installer and cost about $10 each installed???

So what is currently used in England for wire splices in house wiring? the push in "chocolate block" things? I've seen them used in LV applications but they seem too close to a backstab receptacle for me to feel really comfortable with them.

Not really in any way pertinent to anything I'll be doing any time soon, but I just find this stuff interesting.

nate

That is COPALUM, which is an AMP product, and I suspect there are not many installers left.

With AlumiConn the connection is made with a set screw.

The current recommendations for aluminum connections by the engineer in my previous post is at

In merry old England wire nuts have not been used in a long time, and I think AlumiConns are similar to what is used for all small branch circuit splices.

Cheaper than a COPALUM at about $3.50 each locally, and more secure than a wire nut and not as much of an issue for box fill issues

I have a similar issue: A GFCI outlet in the apartment we rent was pouring smoke, white smoke that smelled like ozone. The wall behind the outlet wa s hot to the touch. My GF only found out about it when stopping by the apartment during a break from work. Luckily she works in a building just next door to our apartment building. She smelled a strange smell, the lights were off, and there was white smoke pouring out of the GFCI outlet in the kitchen. She called the fire department, who checked the wall, and also felt the wal l panel was still warm to the touch. The lights were off in the apartment at the time. (we usually turn them off ) and there has been NOTHING plugged into the GFCI plug for months. So ,yeah, we are not staying there any more.

Any ideas on how something like this starts?

Some background: It's a 139 year old historic building that last had a int erior structural rehab in the 1980s. Mixed retail and living units across three floors and a basement.

Ideas?

What else was on that circuit? In an old building like that the GFI was usually going to be feeding all of the down stream outlets.

ng smoke, white smoke that smelled like ozone. The wall behind the outlet was hot to the touch.

ak from work. Luckily she works in a building just next door to our apartme nt building. She smelled a strange smell, the lights were off, and there w as white smoke pouring out of the GFCI outlet in the kitchen.

all panel was still warm to the touch.

ff) and there has been NOTHING plugged into the GFCI plug for months.

nterior structural rehab in the 1980s. Mixed retail and living units acros s three floors and a basement.

I have seen loose connections cause a GFCI outlet to burn, but that was wit h a load on it.

Was there any indication of water getting onto the GFCI outlet somehow? Th at would cause it to burn without a load.

A faulty plug would cause heat to develop, but that would also have to have a load on it.

John Grabowski

A feed through installation would allow a downcircuit load to pull through the GFCI and an internal failure in the GFCI could cause this.

That's true too and one more possibility. Maybe it was not the GFI at all, but a mouse chewed tthe wires real close to that box.

You are all saying that the GFCI had a loose connection or had a heavy load on it. I had a Legrand Brand (Menards) GFCI installed in an outside rated box with proper cover and all gaskets in place. It was about 5 years old. It's on the end of the line of a 20amp breaker and was rated at 20 amps, so no continuing connections to pull load. It also had nothing plugged into it. At time of removal, the wire lugs were very tight and showed no signs of heat damage. However, while away for a week of work on 6/1/20, the GFCI device burnt up. I noticed it upon return when the breaker in that outbuild was tripped, Upon resetting it, it immediately tripped again. After inspection, I found the outside water tight box (PVC?) holding this GFCI device was distorted. The outlet also had some soot under the cover and and around the door gasket. Disassembly proved that this was truly a failure of the device internally. It was not caused by a "Loose" connection to or out of the device. The repair required replacement of the entire box and outlet. None of the wiring was replaced as it was not damage and looked it's age (5yrs old). I'm just throwing this out there to say that they can fail internally. It does not have to be a loose connection or overload/high current situation for them to fail. I replaced with a standard outlet this time and replaced the 20amp breaker with a GFCI breaker as this could have been disastrous, especially with me out of town.

Sounds like why NFPA started requiring AFCI protected circuits ... almost 20 years ago.

On Tuesday, June 9, 2020 at 10:38:04 PM UTC-4, snipped-for-privacy@gmail.com wrote:

I wonder what kind of failure analysis has been done with GFCI? Might be a good idea if they had a thermal fuse like heating appliances.