Cause of electrical fire

[journeygal]

Hello! A couple of months ago the rental house, built in 1999, we were living in caught fire. The fire marshall determined it started at a receptacle. There was nothing plugged into the receptacle at the time. However, the receptacle was the first receptacle on the circuit, and it was not pigtailed, so all the current for the circuit passed through this particular receptacle.

When we first rented this house, this circuit kept tripping the 20 amp breaker (12 gauge wire). Upon investigating, we realized that the garage was also fed from this circuit. We isolated the garage and the breaker stopped tripping. (This house was located in the country, was owner built, and probably never inspected).

At the time of the fire, we had a portable air conditioner, a small ups protecting a computer, large flat screen monitor, and two printers plugged in to the circuit. Again, nothing was plugged into the receptacle which caught fire.

The fire marshall and the insurance investigator both said that the cause of the fire was indeterminable, although both agreed it started at the receptacle. Does anyone have any insight into why this receptacle would have started on fire? I have some ideas but would like to hear others' thoughts.

Thanks!

Anne Thornton

 

[ptonsparky]

Most likely poor connection(s) of the wires to the device. You may have had other factors at play as well.

 

[kwired]

Current still has to pass through the wires and terminals on this receptacle, as you mentioned early in your post. An poor connection will be a place that can develop heat, which makes the connection get even poorer, eventually it gets to a point where you have a "glowing connection".

Often these do burn themselves out and leave you with an open circuit condition, and damaged conductor insulation for a couple inches and a damaged device it was connected to. But should there be combustible material in the vicinity you have something that can continue burning and eventually start the structure on fire. This is one reason we need to use boxes at junctions and other devices, to help contain such incidents should they happen.

 

[GoldDigger]

The poor connections may also have been made worse by whatever overloads were repeatedly tripping the breaker.
Thermal cycling can lead to bad connections.
Was the receptacle in question wired with screws on the side or push in connections at the back?

 

[journeygal]

Re: Cause of electrical fire

The receptacle had the wires wrapped around the screws on the side. Also, the receptacle was a 15 amp receptacle, on the 20 amp breaker. I'm guessing that contributed as well...

Anne

The poor connections may also have been made worse by whatever overloads were repeatedly tripping the breaker.
Thermal cycling can lead to bad connections.
Was the receptacle in question wired with screws on the side or push in connections at the back?

 

[GoldDigger]

The receptacle had the wires wrapped around the screws on the side. Also, the receptacle was a 15 amp receptacle, on the 20 amp breaker. I'm guessing that contributed as well...

Anne

Not necessarily. Code allows 15A receptacles on a 20A circuit. And AFAIK 15 receptacles are listed for pass through of 20A.

 

[Bladnaster]

The receptacle had the wires wrapped around the screws on the side. Also, the receptacle was a 15 amp receptacle, on the 20 amp breaker. I'm guessing that contributed as well...

Anne

What Gold said. Only time that wouldn't hold true is if you had only one receptacle.

 

[journeygal]

Re: Cause of electrical fire

So would something like this happen quickly, or would it be something that was brewing for a while? Basically I'm trying to figure out how to avoid (or minimize) the chance of this happening in the future. Would an arc fault breaker work in a case like this?

Anne

 

[Bladnaster]

So would something like this happen quickly, or would it be something that was brewing for a while? Basically I'm trying to figure out how to avoid (or minimize) the chance of this happening in the future. Would an arc fault breaker work in a case like this?

Anne

First thing I would probably do would be to go through all the outlets in the place to see what I find and if there is a common theme. When you look at the wiring, does it seem like they knew what they were doing or does it look like a handyman hack job? Check the wire's themselves for nicks and all connections to the devices, and if any splices, check those also.

Havent really worked with afci's. Cant say they would hurt, but if you do have issues with your wiring you may have to spend some time hunting down the causes of nuisance trips.

At least that would be my take on it.

 

[normbac]

I went on a service call years ago and the owner complained of sparks at an outdoor duplex receptacle no cover. I walked up to it, gave it a nudge and without exaggeration a large flame shot out at least four to five inches. Can still picture it to this day. it was broken and not really sure how it caused the large flame maybe metal was arching slowly and overtime it was thin enough to arch.

 

[Fulthrotl]

So would something like this happen quickly, or would it be something that was brewing for a while? Basically I'm trying to figure out how to avoid (or minimize) the chance of this happening in the future. Would an arc fault breaker work in a case like this?

Anne

this was probably a chronic problem that has been cooking for a while.

pigtail the devices, so you have a wirenut carrying the circuit load, and not
a side wired device. i'm partial to wagos, but some people hate them.

i have strong opinions on arc fault. it wouldn't be on my choice of protective
measures.

 

[ptonsparky]

The glowing connections Kwire talked about can happen at less then 1 amp of current. From what I understand, someone correct me please, some of the AFCI require 6 amp or above to work on the arc detection.

 

[growler]

Hello! A couple of months ago the rental house, built in 1999, we were living in caught fire. Does anyone have any insight into why this receptacle would have started on fire? I have some ideas but would like to hear others' thoughts.

Most likely poor connection(s) of the wires to the device. You may have had other factors at play as well.

I agree with the loose connection idea.

The thing about these loose connections is they happen all the time but most are caught before they cause any serious damage.

Even if every other connection in the house was good this one bad connection would be a problem.

To be on the safe side I would inspect other receptacles in the house to make sure they are terminated correctly. Many times there may only be this one problem and other homes may have wiring that looks like a fire waiting to happen. Have a "qualified" electrician inspect the wiring.

 

[kwired]

So would something like this happen quickly, or would it be something that was brewing for a while? Basically I'm trying to figure out how to avoid (or minimize) the chance of this happening in the future. Would an arc fault breaker work in a case like this?

Anne

AFCI?

This is a hot topic for some, the AFCI manufacturers would like you to think they would have detected this and responded by tripping.

There have been many that have simulated this situation with an AFCI in the circuit and it never tripped.

Do a search for "glowing connection" and maybe plus "AFCI" and you can come up with information and comments on these tests - just searching this site alone should give you some good reading on this.

 

[mgookin]

And it's also possible it was just a bad receptacle.

 

[al hildenbrand]

So would something like this happen quickly, or would it be something that was brewing for a while? Basically I'm trying to figure out how to avoid (or minimize) the chance of this happening in the future.

Among the list of stressors of electrical wiring is simple Newtonian physics, the kind of physics we were taught in high school. Common real world conductive materials all have resistance. When current passes through conductive materials, heat is released equal to the square of the current times the resistance. The heat causes the temperature of the conductive material to rise. Conductive materials physically expand in dimension as their temperatures rise, thus changing shape.

The dimension changing, shape changing, of terminal screws, wires, backstab springs, wirenut metal spirals, etc., is the thing to hold in mind, in my opinion, while making each and every connection or splice that one makes on the job.

Over-tightening a connection can be as bad as making a loose connection. Following the manufacturer's torque specifications for making a splice places "active" compression on the conductors in the splice in an active range that the splicing device can tolerate, allowing a certain amount of movement (conductor thermal expansion and contraction). If the splice is too tight, there is more of a possibility that the conductors will deform, when they are heated and expand, and when the conductors cool off they will be a tiny bit looser. This additional looseness adds to the "contact resistance" of the connection of the conductors inside the splice, and, the next time the current flows, the resultant heat will be a bit greater, resulting in a slightly higher temperature and an additional bit of dimensional expansion, and the cycle goes along, repeating, until the first tiny plasma arc appears in a gap.

Plasma arcs are so hot that they liquefy and boil the metal around them, and the process of the destruction of the splice accelerates, adding more resistance (and thus more heat under current) every cycle of the load current passing through the splice.

The lesson: pay attention to each and every splice, following the specific material instructions.

 

[edlee]

So would something like this happen quickly, or would it be something that was brewing for a while? Basically I'm trying to figure out how to avoid (or minimize) the chance of this happening in the future. Would an arc fault breaker work in a case like this?

Anne

It seems likely that the problem developed over time unless it was caused by something like rodent damage or an electrical storm. It would also likely have been noticeable if for some reason you had the receptacle outlet removed from the box for examination. The best thing to do would be to take apart every switch and outlet and check the connections. Time consuming! Pigtailing is certainly the preferred method for feeding through a box.

Don't know about arc-fault protection in this case. Given how easily they trip it probably would have shut the the thing off.

 

[kwired]

It seems likely that the problem developed over time unless it was caused by something like rodent damage or an electrical storm. It would also likely have been noticeable if for some reason you had the receptacle outlet removed from the box for examination. The best thing to do would be to take apart every switch and outlet and check the connections. Time consuming! Pigtailing is certainly the preferred method for feeding through a box.

Don't know about arc-fault protection in this case. Given how easily they trip it probably would have shut the the thing off.

AFCI can not tell the difference between a glowing connection and a resistance heater drawing same amount of current. Should a ground fault develop during the process it may trip because of ground fault feature within the AFCI device, but some don't have GF component anymore.

 

[K8MHZ]

I have in my collection somewhere, an old Slater receptacle that started on fire with nothing plugged into it. It was wired as a pass through and hidden from view behind a large entertainment center. Once we found it, we noticed black scorch marks above the receptacle.

Luckily, the downstream receptacles failed and the burnt receptacle was located before the fire got outside of the box.

 

[kwired]

I have in my collection somewhere, an old Slater receptacle that started on fire with nothing plugged into it. It was wired as a pass through and hidden from view behind a large entertainment center. Once we found it, we noticed black scorch marks above the receptacle.

Luckily, the downstream receptacles failed and the burnt receptacle was located before the fire got outside of the box.

I had a customer that had a house fire last winter. Fire investigators said it started in a specific receptacle, and based on how much burn damage was near and above that location I would agree. What they did not determine was what failure occurred at the receptacle to cause the fire. My best guess was glowing connection. There was a television plugged into it but was not on at the time of the fire. Feed through did go to a blower on the fire place - which was running. If there was an outlet box it was plastic and completely gone from the fire, same for the receptacle - only thing left was metallic parts. Was kind of hard to determine specifically what happened, but glowing connection seemed most logical.