Up on the Roof

Shutterstock/ Tfoxfoto
Shutterstock/ Tfoxfoto
Published On
Mar 15, 2020

When I was doing fire and electrocution investigations, I received a phone call from an insurance company about a fire in a home the company insured.

Upon arrival at the home, I discovered that the fire department had responded to heavy smoke in the attic and the smell of burning wood. After turning the main circuit breaker off on the 200-ampere panel supplying power to the home, the firefighters proceeded to determine the source of the smoke in the attic area.

The fire investigator and I went into the attic and discovered a wooden truss that was part of the roof support structure had been heavily damaged with wood charring. At this point, the fire department used a portable fire extinguisher to douse the area surrounding the charred wood, determined that the fire was extinguished and left the scene. It was up to me to investigate the reason for the fire and whether it was electrical in nature.

I started by studying the wooden truss and the point of charring since that was the origin of the smoke and smell of burned wood. Closer examination of the truss determined there were two metal plates, one on either side of the splice point on the truss. These metal plates were installed at the time the truss was manufactured and provided connection between the 2-in.-by-6-in. pieces of wood.

I also noticed that there was a metal vent pipe that was connected to the range hood on one side of the truss and there was a metal duct from the air conditioner on the other side. Both the range hood vent pipe and the air conditioner duct were touching the metal plates, one on either side of the truss. The wood was charred between the two metal plates, which spurred me to further examine why the wood would have been charred without any clear indication of an ignition source.

I exited the attic and went into the kitchen to look in the range hood to see if anything looked wrong there. The branch circuit supplying the range hood appeared to be properly installed and the range hood was connected to an equipment grounding conductor. Everything was normal there.

I went up to the roof to look at the air conditioner. At the bottom of the air conditioner disconnecting means, I found a piece of cardboard with a ground screw mounted on the unit. I looked at the connection of the EGC from the air conditioner branch circuit coming into the bottom of the disconnecting means. The EGC was connected to the grounding point in the disconnect, but the equipment ground screw was not installed from the grounding point in the disconnect to the disconnect enclosure. It was still sitting in the bottom of the disconnect, which indicated to me that the original electrician never installed it.

Using a multimeter, I determined that one of the air conditioning motors had shorted to ground. Without the equipment grounding conductor connection to the disconnect enclosure frame, the fault current could not flow back on the EGC, which was isolated within the enclosure. The fault current eventually went from the air conditioner frame onto the metal duct, down to the metal plate on the truss, through the wood, to the other plate on the other side of the truss, and to the metal duct connected to the range hood. Since the range hood was properly connected to an EGC, the fault current attempted to go back by that equipment grounding connection. The metal plates on the side of the truss created a high-impedance current path that had enough impedance that the fault in the air conditioner motor did not trip the circuit breaker for the air-conditioner, but it did provide enough current to char the wooden truss. Had the fault current lasted long enough, I am sure that ignition would have occurred, and a full-blown attic fire would have resulted.

Over the years, I have determined that a fire report is only as good as the investigator and that, at times, luck can play a part in determining a potential fire or the reason a fire occurs. At times, electricity gets blamed for fires when there clearly is no reason even to assume that it was caused by the electrical system. There are times when the electrical system contribution is secondary to the origin of the fire, and, other times, it is the true source of the fire. Care must be taken to determine the true origin and not jump to conclusions.

About the Author

Mark C. Ode

Fire/Life Safety, Residential and Code Contributor

Mark C. Ode is a lead engineering associate for Energy & Power Technologies at Underwriters Laboratories Inc. and can be reached at 919.949.2576 and Mark.C.Ode@ul.com.

Stay Informed Join our Newsletter

Having trouble finding time to sit down with the latest issue of
ELECTRICAL CONTRACTOR? Don't worry, we'll come to you.