ECN Forum Forums Code Related Discussion NEC & other Code issues Arc Fault Breakers

Don,

"More people die in fires in the bedroom so that's where we'll start with mandatory AFCI" seems a very simplistic and flawed approach.

I would've thought that more fatalities occur in the bedroom simply because someone has to wake up before he realizes there is a fire. By that time, the blaze has well and truly taken hold and maybe even rendered its victim unconscious before he has a chance to waken.

As someone involved with the aftermath of fires, would you agree with my logic?

Paul you are correct. That is the very reason that smoke detectors are required in most areas for dwelling units. Unfortunately many times after a fire, the detector is found without batteries. This is a reason why some areas now require dual (utility and battery) powered detectors. I fully support these requirements and would also support the AFCI rule if I am convinced that they will prevent a significant number of fires. It does not appear to me that the AFCI can do that at this time.
Don(resqcapt19)

I am having engineering review my technical reply, hopefully sometime tomorrow. In the interim, our AFCI devices are both UL1699 (AFCI) and UL1043 (equipment ground fault).

Obviously, arcs can occur anywhere in a home wiring system. Anything I say about why the code making panel chose bedrooms woult be speculation since I wasn't there. The fact that most people die in bedrooms had something to do with it, but I'm sure that consideration was given to the nature of wiring in a bedroom, and the types of devices that are used. By that I mean that many bedrooms have large furniture against the wall such as headboard, dresser, chest, nightstands, etc. Therefore, it is more likely that wires would be abused by having the furniture pressed up against and bending the wire. Also, unlike standard lighing loads, the loads may tend to be more continuous such as alarm clocks, telephones, air cleaners, electric blankets, etc. Therefore, the loads would be running while people are sleeping. Just a thought.

Sorry...UL1053...finger slip

For NM-B, with ground, the device will trip on ground fault.

Brendan,

I'm beginning to realize that the AFCI (at least from what I gather from the above tech) is intended to operate on a narrower set of fault conditions than I had assumed earlier.

Don(resqcapt19) describes the "glowing connection". I've seen far too many heat destroyed connections. Connections whose conductive metal to conductive metal resistance is not as low as it should be. When current flows through the connection, current supplying a legitimate down-stream load, the watts of heat released in the connection will be the current squared times the resistance. This heat moves out into the surrounding metal and insulations at a rate determined by their physical properties. The parts "fill up" with heat with gradual temperature rise because much of the heat is able to move on into air and other surrounding materials. But, at a point, the parts are "full", and if more heat is being created in the resistance of the connection than can be sloughed off into the surround, the temperature of the connection starts shooting up extremely fast. Physics calls this point the Black Body Radiation Limit. Our real world connection gets to this limit quicker than the theoretical. The connection "glows" at circuit normal load current levels. High temps are attained that result in destruction of the connection that creates the heat that generates the high temps. Circular, huh? At some point the connection may so degrade as to result in contact with either ground or the other side of the circuit, and there is a parallel fault or a ground fault that the AFCI can "see". Up to this point, the AFCI is blind to what's happening. Am I right? This is in a dwelling bedroom with 15 or 20 amp overcurrent branch circuit protection, a load that might be a space heater, and the AFCI starts "looking" above 50 amps, so the connection that is glowing, where ever it is, cord and plug contacts, wire nuts, terminal screws, hot or neutral, remains invisible while heat is released.

Do actuarials exist for this scenario?

The statistic about %50 of the GFIs failing, I believe that was in the IAEI magazine a couple of years ago. The next month they came out and pretty much disowned the writter and the article itself. Of course, there could have been another source with the %50 number, I don't know.
Don

A senior research engineer at UL has made the following statement.
"...For example, electrical ignition most frequently occurs as the result of Joule heating or electrical arcing. It should not be expected that those ignition scenarios representing Joule (I^2R) heating would necessary be prevented by an AFCI..."
The complete statement was published in the IAEI News. It is available at: http://www.ul.com/regulators/afci/Dini.pdf

This seams to support my point that many electrical fires are caused by poor connections and AFCIs do not detect these types of faults.

Don(resqcapt19)

indeed....
Mr. Dini performs an amazing run of interference for UL. The sanctioned widget of the century may well turn out it's edsle.

After reading the article when it came out i contacted UL , asking if AFCI's were listed for ungrounded circuits. This was not a 'if it'll work' question,or 'do you think it's a good idea?' question.
It was a listing question, as it IS or it IS NOT listed for such use.
UL's reply did not resolve this.

First,let me identify myself as a lobbyist for the Michigan Association of Home Builders. I deal with code and land use issues.

From this perspective, one of the real problems with the code development process is the ability of manufacturers to use codes to create a market for their products.

That is precisely what happened with arc-fault circuit interrupters.

Those of you who are properly skeptical of this claim can go to http://www.holmesreport.com/holmestemp/story.cfm?edit_id=1176&typeid=4

which outlines the specific public relations and lobbying efforts used by Cuter-Hammer create a market for their device by creating a code requirement in the NEC and a corrsponding "consumer pull."

I wonder what happened to Brenden from Cuttler-Hammer. We need a manufacture's rep to help answer the questions.
Don(resqcapt19)