Well, after sitting on the sidelines with this topic, I now have some thoughts
After reading the thread in the other section of this forum, and learnig about the 75 amp trip requirement, Id like to ask what happens if an arc fault (parallel) occurs that has a high enough impedance to limit the current flow to 10 or so amps. Most of us know that arcs are so hot that a welding machine will melt steel at much less than 75 amps. I was under the impression that these devices "recognized" the asymmetrical wave form associated with arcing, regardless of the magnitude. So what happens to a high impedance arc that limits the current to about 10 amps? Do we have to wait until a low enough impedance path results, or will they trip at the 10 amp level due to the "noisy" waveform?
It has been stated that AFCIs do not protect against any arcing faults of less than 75 amps. This is not 100% true. Some of the "carbonized path" arcing situations were tested with simulated load currents.(5 amps, 10 amps, and 150% of load) The two Tables below are from a UL document titled "Arc Fault Testing and Arc Fault Scenarios - January 28, 2002" http://www.ul.com/regulators/afci/AFCI_scenarios020502.pdf
Table 1 shows some details of the tests conducted. (X indicates which test was conducted) (Combination AFCI) This is the (common breaker type) AFCI which complies with the requirements for both branch/feeder and outlet circuit AFCIs. It is intended to protect downstream branch circuit wiring, cord sets and power-supply cords.)
Table 2 shows which type of device provides protection under various conditions. Here again, "Series Arcing Protection" is assumed to mean - protection against arcing across a break in a single conductor, which progresses to a ground or parallel arcing scenario.
Re: AFCIs...my turn#14235 09/17/0207:35 AM09/17/0207:35 AM
Maybe I'm stupid, but I can't see where table 2 addresses the level of current at which the device operates. Table 1 shows what type of test and the current level. Table 2 only shows whether or not it operates as an AFCI, but not which scenario it operates under.???
Re: AFCIs...my turn#14236 09/17/0208:09 AM09/17/0208:09 AM
Your question centers in on one of the black holes in the available information.
To my knowledge, the technical info on the contents of the black box "sophisticated technology" contained within the manufactured branch feeder AFCI units is not available to us, the end installers.
As I look at the UL1699 AFCI Test Scenarios that Ed posts above, I note that the low current (below 75 A) tests rely on enough physical damage occurring as a result of the failure, that a connection with the egc is established.
I presume that the "AFCI" then operates because of the add-on feature of also being built with a 50 mA (or thereabout) GFP. The manufacturers have stated that the AFCI portion ignores events below 75 A magnitude.
The branch feeder AFCI fails the low current tests when the wiring method is ungrounded (knob & tube, 2-wire NM, 2-wire zip cord).
The silence of the manufacturers about the construction of the microprocessor(s) and memory(ies) inside the branch feeder AFCI, and the silence about the progamming executed by the same, really concerns me.