ECN Forum Forums Code Related Discussion NEC & other Code issues AFCI Branch Circuit length

Per the "Newest" GFCIs:


This is discouraging news to me!
I really thought the devices had some type of "Crowbar" feature, which opened the Line (input) circuitry to the device, when it became "Non-Protecting" (lost the GFCI trip ability) - and the device would be Non-Resetable and unusable...PERIOD!!!

I guess it's just too difficult to fuse the Line side, and create a crowbar which knocks out the fuse if the device looses its trip / sensor circuitry.

Very sad news, indeed.

As far as the AFCI thing, it really seems to me that predicting and reacting to an Arc pattern is difficult at best.

Not that it can not be done, just that it is such a variable state to log. That, along with the many many factors of a "good" pattern (non problematic circuit components) applied to the nearly infinite loads and installations, makes me wonder if the approaches should be taken from different ways.

I would like to see (and if possible, even assist in the developments of) accurate fault interruption devices, which will drammatically reduce the hazards of fires in dwelling units.

The ideas I have would _Possibly Work_, but _Definitely_ increase the costs of Residential Wiring - which is not a good tradeoff.
Need to "think outside the box" on this one, and maybe think up something which would be effective!

Just wanted to toss in my 2¢

Scott35

As I said before I have had long talks with Harvey Johnson from C-H about these AFCIs. The real problem is a series arc that would burn your house down if it occured in the pile of dust bunnies under a bed from a pinched zip cord might look a lot like the normal arc you get in a hair dryer "pulse heat" switch if you are slow on the trigger.
The circuits are really trying to find a sweet spot between a "safe" arc and an "unsafe" arc without knowing where the arc is. You can see some nasty arcs in your garden variety snap switch if it is dark in the room where the switch is. Try it sometime. Use an extension cord on a switched receptacle and put the load in another room, then turn the light off and flip that switch a few times. It makes you happy that thing is in a box. That's a safe arc. Imagine the same arc in some cotton fuzz and you see the problem.
Parallel arcs are easier to detect because you are just looking for verey high current spikes but you also need the sweet spot between the normal LRC of a big motor like a vacuum cleaner (brushes etc) and that pinched zip cord under the bed.
This would be a lot easier if NFPA allowed us to put the protection closer to the load.
Then you might be able to dial in more protection where you should not expect "normal" arcing.
I am guessing this will be tossed around a lot between now and 2008 or 2009 when most places will actually adopt the code

Ryan,
I don't buy it...they went to great lengths to show that there is no such thing as a series arc a few years ago. They proved that you couldn't sustain a series arc at 120 volts. While the UL information uses the term series arc there is nothing in the testing information for any of the AFCI listing types that says anything about testing for a series arc. They imply it in this section, but in reading it closely they are still only testing for a parallel arc. More misinformation designed to scam the code writers and the public. (as you can see I have no trust for any organization involved in this issue)

Don

I don't know how they can say there is no such thing as a series arc. If that was true we could eliminate all of those explosion proof enclosures. A <hazardous> series arc happens when you have a broken wire that makes and breaks rapidly in a flamable environment. They seem to be implying that a bedroom is an "HO guage" class III div 2.

The information does not pertain to the short arc that occurs when a circuit is broken. An series arc cannot be sustained at normal dwelling unit voltages. The poor connection of high resistance fault or glowing connection is not really a series arc as it is not an arcing fault. It is just a high resistance point where high heat is produced by the flow of current. The resistance limits the current to a point far below the trip point of the OCPD. I think that the AFCI people went to great lengths to prove that there are no sustainable series arcs at dwelling unit voltages because most were calling the high resistance connection a series arc and they know that their device cannot directly detect such a fault.
Don

For the purposes of the "bedroom fire" discussion a "The poor connection of high resistance fault or glowing connection" is what we are trying to prevent. When that is a zip cord buried in carpet fibre or lint under the bed we have a fire. Isn't what an AFCI is for?
I didn't really think these fires were caused by a crowbar short somewhere.

The currently available AFCIs do not respond these types of faults and I can find no information in the UL test standards that indicates that the new combination type will directly respond to these types of faults.
Don

[This message has been edited by resqcapt19 (edited 05-21-2006).]

Yes Ryan, it is Warren Schill. Sorry, I did not attend your class, but twice through code changes, no matter how good the instructor (and you are a good one), is more than I can handle. Thanks again for being there and your input.