AFCI's Explained by Mike Holt & Eaton Engineer

Check out my VIDEO page of and visit video # 4 which is a recorded segiment of explaining how AFCI’s work and why Parallel Arcs are MUCH more important than Series Arcs.

This video was recorded back in 2003 so they have advanced even MORE.

Very Nice thanks Paul
Roy Cooke

I didn’t hear them talk about series arcs. Maybe it stopped early or something.
Just remember virtually every aluminum wire fire was started with a series fault.

Series arcs are covered as they speak about the switch and so on but Series arcs greg are simply not as problematic as Parrallel arcs in that series are limited by the current flowing through them before the breaker will trip on its own( Note: that is NOT to say a series arc of a low current series arc is not dangerous…yes it is…greatly so but they are addressing that )…with a parrallel arc the OCPD simply will not measure this heating up over a longer period and this is where the majority of fires come from…not the series arc.

He talked about the fact they know the waveform of a series and parrallel arc and the AFCI is programmed to understand this arc and protect against it but it is also programmed in a way to identify these arcs and this was in 2003…Eaton released the latest versions in 2005-2006 that even take it to the next level…

I have many questions for them but I feel confident the AFCI is moving in the right direction…atleast to me it is BUT we have many other.

Greg with no disrespect do you have any documentation on that statement about Aluminum series faults, remember the amount of current through a series arc is limited to the OCPD that controls the circuit…

On a parrallel arc you have two points of LOW arc that can go without notice on the OCPD because it is simply an ARC that can represent itself as a load and the OCPD simply never clears until which point it becomes a fire issue.

While Series can present a problem at the terminals…what you end up getting is a high heat point that usually melts down to BECOME a parrallel arc to which the AFCI would react as the melting point of the devices is rather low compared to a non-combustible fire point issue.

I believe they are reading this Series ARC…and today even more but the main focus still remains on the parrallel arc issue.

However…I am not defending them greg…lol…I can only provide what I know about the subject based on what I have seen. I am ALWAYS open to understanding anyone elses views on it.

It quite possible could be smoke and mirrors…however, based on the explanations of the engineers at Eaton it still amazes me what they are doing to detect this…but it is all based on the signature of a normal sine wave versus a known arc wave and the frequency of the arc that aids in detection.

greg…I am not negating the problems of Series Arcs…as I have seen tests where this is a problem as well…I am only refering to the AFCI itself and it’s concept…again from my understanding the largest problem arc they wanted to deal with was the parrellel and then they delt with the series in a more detailed nature…I intend to ask them point blank and see what they say on it


Think on this

An arc has a electronic foot print — This can be spotted and trip off all power to the house!!

This idea of protecting each branch is very $$ intensive

Your thoughts??


I have also found out greg and please correct me if I am wrong but the newer AFCIs are actually AFCI/GFI and AFCI/GFCI combination units ( atleast Eaton/Cutler Hammer models are ) and offer the advantage of ARC detection on the parrelle level and to some point the series level but the actual low current " GLOWING" arc resulting from series arcs was a problem but with the dual combination and earth leak features of the newere AFCI’s this has been tackled…I look forward to being able to see what this involves.

Check out this article from Mike Holt and his experience with EATON when they had him come out to their facility and test it…


I think me and greg are just hashing over ideas.....I most certainly am learning more and more about the AFCI every day as I made it a project of mine but it is still a engineering mystery as to HOW they develope this footprint and is it always a CORRECT footprint.

On the expense, well I am again a electrician who believes the bedroom is where it should stay simply because that is the location we are the least aware as we sleep…I put LIFE well above any material thing but it is simply not my call.

Yes the key thing here is an ARC has a defined and known waveform to which it made it rather easy I think for them to start with Parallel Arcs first but as greg so correctly stated they lacked in Series arcs.

I think the conventional thinking was the Series arc ( unless it is something like a standed conductor with only partial strands broken yet still touching that this arc could turn into something more…sure it could but in reality it is limited by the circuit itself where a parallel arc may be " HIDDEN" from the OCPD until which time it is too late…but they are coming together now.

I am like greg ( please dont think I was arguing with ya greg, I know what you are saying ) but as I start to look deeper in the development they are addressing this with other advents…like most people do not know that GF protection is built into AFCI’s…but for property not personal protection…it is usually about 30MA…but usually this would also aid in detection of a low current series arc but possible not so in a parrallel arc…

Is it going to be costly IF the all branch circuits end up being AFCI protected…yep for the builders and their clients but not for the electricians…:slight_smile:

But in my talks with Eaton they have expressed that realistically they would like to see costs for AFCI breakers down to 7-10 dollars a breaker as they move into the " ALL" branch circuit protection takes place but at this point it is not known.

I find this funny…read how (1) month can change Mr. Holts opinion on AFCI’s as I remember his HARSH newsletter blasting them to the point Eaton was about to come after him…then he witnessed some tests and well in 30 days he changed his mind…

Check out this report…kinda interesting if you are into the AFCI thing-

The purpose of this investigation was to assess the ability of a Branch/Feeder
AFCI incorporating ground fault protection to respond to a glowing connection
occurring at a wiring device termination. The investigation considered cases of
glowing connections occurring at either the interface between the insulated
(black) ungrounded circuit conductor and the wiring device terminal; or between
the insulated (white) grounded circuit conductor and the wiring device terminal.
By test design, a physical separation was maintained between the two insulated
conductors and between each insulated conductor and ground. By virtue of this
worst case configuration*, it was demonstrated that a Branch/Feeder AFCI
incorporating equipment ground fault protection (30 mA trip) is capable of
terminating a glowing connection and the associated heating effects by sensing
the differential current associated with thermal degradation of the wiring device
insulating material(s).
(* Worst case configuration refers to the deliberate prevention of tripping by the
Branch/Feeder AFCI incorporating ground fault protection due to contact
between overheated circuit conductors, or between either of the heated circuit
conductors and ground.)

The CPSC report titled What Causes Wiring Fires in Residences?7, examined fire
statistics for fires originating in the electrical distribution system. As part of this
investigation, the failure modes involved in 149 residential electrical distribution
system fires were examined by component. It was found that receptacle outlets
and switches accounted for 29 (19%) of the fires. Twenty-six of the 29
investigated fires were attributed to electrical system components associated
with receptacle outlets. Where a cause could be determined, the most frequent
problem cited was “loose connection,” which was either a loose fit between the
outlet contacts or a loosening of the wire connection to the receptacle. In some
cases it was reported that copper wires were inadequately secured during initial
installation so as to cause subsequent deterioration and overheating.
Lighting fixtures and lamps accounted for an additional 19 (13%) of the fires.
The most common problem noted in lighting fixtures and lamps was poor
workmanship, as evidenced by loose or otherwise defective connections that
caused wires and insulation to overheat.

Recently, there have been a number of questions from the field about the response of arc fault circuit interrupter (AFCI) circuit breakers to commercially available AFCI indicators. AFCI indicators operate by producing a waveform similar to an arc fault. However, because they cannot produce an actual arc fault, an AFCI indicator may not trip every available AFCI.
Therefore, if an AFCI indicator plugged into a receptacle protected by an AFCI does not trip the AFCI, it does not mean that the AFCI protecting the circuit is defective and needs to be replaced. When this situation occurs, you should push the “Test” button provided as an integral part of the AFCI itself. If the integral test button does not trip the AFCI circuit breaker, it should be replaced.
To notify users of this product limitation, Underwriters Laboratories Inc. requires AFCI indicators to be marked or be provided with instruction manuals that state the following or equivalent:
CAUTION: AFCIs recognize characteristics unique to arcing, and AFCI indicators produce characteristics that mimic some forms of arcing. Because of this the indicator may give a false indication that the AFCI is not functioning properly. If this occurs, recheck the operation of the AFCI using the test and reset buttons. The AFCI button test function will demonstrate proper operation.
For more information on AFCI indicators, contact Steve Brown in Melville, N.Y., by phone at +1-631-271-6200, ext. 22420; or by e-mail at
If you find a defective AFCI, please alert the product’s manufacturer and UL through UL’s AHJ Product Report Form available online at
Click here for more Mike Holt AFCI newsletters

Electrical fires are often caused by loose connections, broken wires, or shorted wires in the power distribution system. Arcing at the point of faulty contact may cause sparks that can jump several inches to ignite combustible material. Furthermore, heat generated by loose contacts can deteriorate the insulation thus exposing the wires and resulting in intermittent short circuits between conductors.
There are two types of arcing in electrical circuits:** Parallel** and Series.

Parallel arcing occurs when there is a direct short between two power wires and the current is limited only by the resistance of the wire in the distribution circuit. Usually the magnitude of the parallel arc current is much higher than the breaker rating. Even though it only flows intermittently, the average current may be sufficient to eventually trip a conventional breaker due to heating of the bimetal strip or the peak current may be large enough to trigger the magnetic sensor. This makes the thermal/magnetic breaker reasonably effective in protecting against parallel arcing when the peak current is a few hundred amps. Unfortunately, the parallel arc current can be limited if the circuit is implemented with wire that is too small for the application. Extension cords often aggravate the situation by introducing additional resistance in the circuit so the average value of the arc current is below the thermal breaker’s rating and the peak current is below the sensitivity of the magnetic sensor.

Parallel arcing is generally more hazardous than series arcing because the energy in the sparks is much higher and the hot metal that is ejected is more likely to come in contact with combustible material. Since parallel arcing usually results in peak currents well above the handle rating of the breaker, electronic circuits can detect it easily and trip the breaker in a fraction of a second.

Series arcing is caused by a loose connection in series with the load circuit. Series arc current is limited to a moderate value by the resistance of the device that is connected to the circuit, such as an appliance or lighting system. The amount of energy in the sparks from series arcing is less than in the case of parallel arcing but only a few amps are enough to be a fire hazard. Series arcing is particularly insidious because the arc current remains well below the rating of the thermal breaker and the magnetic sensor will never respond to such low amplitude currents. Since the peak current is never greater than the steady-state load current, series arcing is much more difficult to detect than parallel arcing.
Hazardous Conditions

NOTE: this is where we HOPE the GFP at 30MA kicks in to add the elevated level of protection in the AFCI…even if the NEC does not require the AFCI to have GFP built in…It probably SHOULD !

  • Loose wiring connections at outlets or switches.

  • Loose wires twisted together and held by a wirenut.

  • Frayed cords on appliances.
    The detection of series arcing is complicated by natural occurences of arcing when switches are activated or appliances are plugged into the socket with the power turned on. These short-term arcing situations are generally not hazardous and should not cause nuisance tripping.
    Non-Hazardous Conditions

  • Toggling a power switch.

  • Plugging in an appliance.

  • Replacing a light bulb.

  • Impending failure of a light bulb.
    *]Switching of a motor relay.
    When the normal load current being supplied by the circuit is below the breaker rating, circuit breakers in common use today do not react to an intermittent condition. For example, if a hair dryer normally draws 10 amps but the wall outlet has a loose screw terminal so it makes contact only half the time, the average current is 5 amps and the thermal circuit breaker thinks the current is still well below its rating of 15 or 20 amps. In this case the magnitude of the arc current is limited by the heating element in the hair dryer and the maximum value is never more than 10 amps. New arc fault circuit interrupters (AFCI’s) have the ability to detect the distinctive difference between normal current and intermittent arc current. In 1994 an insurance company survey of 660 fires of electrical origin indicated that more than one-third of them were due to arcing conditions. By opening the circuit when a hazardous condition is sensed, electrical safety can be significantly increased. AFCI’s in compact and economical packages have been made possible by advances in integrated electronic components.

Aluminum wire failures are series faults. The terminals arc open.
Mike pointed out in the video that it is actually the GF protection that will detect most faults the AFCI devlce will find.
I suspect they are just cooking the books to make these things seem like a panacea but rushing them in the code the way they did insures there are a lot of marginally capable devices installed instead of waiting until the combo device was perfected. The idea that the old style that are installed will get upgraded is fantasy. I still wonder what percentage of the defective SqDs got replaced.


Excellent point…I was thinking you were speaking more so of series arcs at actual wire points and not termination points as much which is most certainly a series arc…yes indeed…However

The UL did an study and tested under controlled conditions this very exact problem with regards to devices and terminals and found that a large percentage simply burn through opening the circuit while the remainder was detected by the AFCI and yes probably the GF portion but since they all know this it is a valid safety enhancement on that end and the true AFCI protection for the Parallel portion is protected on the AFCI technology end so in the end it is serving both purposes…in a dual role which is probably why they put it in the units without HAVING to put it in.

Probably because a series arc is much harder for them to detect so they figure lets not re-create the wheel. In 2008 they are PUT to the WALL so to speak to have a combination unit which will go hand in hand with the AFCI updates in the 2008 NEC so it should clear up that problem but again from what I am told CH has this solved already in the updated AFCI model currently being sold.

FYI…it is SUCH a relief to be able to CHAT with someone about something without the other person berating someone…thank you greg for always keeping it professional…

I am so facinated by the concept of the AFCI I can see where I might get jaded…thanks for keeping me “Grounded”

I get so into electrical chats…


I too am trying to stay up on the learning curve on AFCI’s. As HI’s we are going to have to know how to deal with this subject.

Quick question – Trip to test or not?? – Then should we check the bed rooms to see that all power is off? – I think so

I also think that if they do not reset we better make sure that some important device is brought back into service with an extension cord

This is sort of like tripping any of the other breakers in the panel and then finding out that the one to the freezer will not reset two day later

Got to be careful out there


Agreed on the added risk. However, it is why HI’s have insurance and we do this all the time when testing GFCI’s which all SOP’s require.

So basically if the AFCI is tripped and wont reset then you have done your job and you just might have helped locate a potential problem as this is not for us to make that guess.

In the future as the industry developes we will see this issue be wide spread as some say not to test the GFCI and some say test the GFCI and some say dont test the AFCI and some say test it.

My personal opinion ( and AGAIN it is MY opinion ) i believe all GFCI’s and AFCI’s should be tested at the device ( breaker ) and if they function you have done your task. Now the use of a external testing device is an added bonus but not mandatory.

We kinda hve to take the manufactures at face value here and establish a trust in them and the technology but out testing them CAN find problematic issues.

Kinda like GFCI’s in a high prone lightning area and surge area…the MOV on the GFCI can fail to do it’s job and the GFCI remains closed and still works and appears to be fine unless someone TRIES to reset…and then you notice the defective GFCI…I think we simply have a duty to test them when they are available and present.

recommending AFCI’s to clients is totally a choice the HI has to make…it is not required and the HI again will have to make that personal choice.

Here is my personal method for testing AFCI’s in newer homes i inspect. I actually test the bedrooms as normal and do not TRIP the AFCI on them until which point I get to the panel for it’s inspection. once I am done I TRIP the AFCI’s at the panel and then visit the bedrooms with a ticker to simply verify they are dead, and to make sure the lights on that room are dead and so on.

I am not doing this to check anything other than that as I checked the circuit wiring itself in the room with my tester when I was actually in the bedroom before.

Anti-Bashing Disclaimer -
This method isolates the AFCI and only takes an extra few minutes to do this…again for those who WISH to make a BASHING deal out of my method…again it is MY method…people are FREE to do it any way they wish…I have my ways and my ways will NEVER change unless I am PROVEN they are not a proper method.

Yes Sir


Richard the GREAT thing is that we have people LIKE yourself who are interested in knowing about AFCI’s and their future.

I am as skeptic as anyone and they STILL have alot of questions but in the end it is a safety enhancement that should prove a breakthru some time down the line.

Hi’s who learn it and understand it will be ahead of the curve when explaining it down the road.

Here is the LATEST draft of the 2008 NEC in regards to Section 210.12

ENJOY- for those that like keeping up to date with changes

**210.12 Arc-Fault Circuit-Interrupter Protection.
(A) Definition: Arc-Fault Circuit Interrupter (AFCI).

An arc-fault circuit interrupter is a device intended to provide
protection from the effects of arc faults by recognizing
characteristics unique to arcing and by functioning to deenergize
the circuit when an arc fault is detected.

(B) Dwelling Units.All 120-volt, single phase, 15- and
20-ampere branch circuits installed in dwelling units shall
be protected by a listed arc-fault circuit interrupter, combination
type installed to provide protection of the branch
circuit. [ROP 2–105, 2–142, 2–111]

FPN No. 1: For information on types of arc-fault circuit
interrupters, see UL 1699-1999,

Standard for Arc-Fault
Circuit Interrupters.
FPN No. 2: See 11.6.3(5) of NFPA 72®-2007, National
Fire Alarm Code® for information related to secondary
power supply requirements for smoke alarms installed in
dwelling units. [ROP 2–118a]

FPN No. 3: See 760.41 and 760.121 for power supply
requirements for fire alarm systems.

[ROP 2–143]

*Exception: The location of the arc-fault circuit interrupter
shall be permitted to be at other than the origination of the
branch circuit where the arc-fault circuit interrupter is installed
within 1.8 m (6 ft) of the branch circuit overcurrent
device as measured along the branch circuit conductors.

**[ROP 2–147, 2–137]