Author Topic: AFCI (Arc-Fault) Breakers Ineffective?  (Read 68 times)
MVMH_99
Member
**
Offline

Gender: Male
View Posts
View Gallery

Nothin' Like Clear MV...


AFCI (Arc-Fault) Breakers Ineffective? « on: May 18, 2026, 04:14:50 PM » Author: MVMH_99
Hello everyone,

   I have had some recent experiences with AFCI (Arc-Fault Circuit Interrupter) breakers not functioning as advertised or claimed, and I thus wanted to share this.  I found this both disconcerting and interesting at the same time.  I have two recent experiences, collectively, that I believe highlight a slightly concerning manner in which these breakers may not actually perform as intended when “called upon.”  This post will probably be a little bit long, but I have a good amount of stuff to cover.

   For background, in late 2022, we had most of our house re-wired, and, per code, had AFCI breakers installed on all of the new circuits.  We also used dual-function AFCI/GFCI breakers on the bathroom circuits, although AFCI is still not quite yet required in this location.  We thought it might be nice to have the added protection, not yet knowing just how effective (or not) these AFCI breakers really were.  Just for reference, all of them are Square D Homeline breakers, all manufactured around the same time.  They are also combination-type ones, and not the older branch/feeder type (the earlier version that could only detect parallel arcing).  So, in theory, they are supposed to respond to series arcs.

   My first encounter with the breakers not performing as intended was when my mom’s hair dryer started malfunctioning and arcing inside, presumably due to a loose internal collection.  She reported that there were beautiful blue and green sparks inside (not from the motor itself, but presumably due to a loose connection that had gradually developed), and yet, the dual-function AFCI/GFCI did NOT detect this.

   Most recently, I had a cheap extension cord with an inferior internal contact start arcing with my Chromebook charger.  You could hear it arcing and crackling inside, and the arcing was even strong enough to cause the lights (on a different circuit, mind you) to start flickering slightly!  This, to me, was particularly concerning, as this is EXACTLY the type of fault that these breakers were INTENDED to be able to detect and shut off, and yet, NOTHING happened here!  The breaker just stayed on, and the arcing continued, getting progressively worse.  If no one had been around, this could have turned ugly, needless to say!

   As a result of my experiences above, it is my conclusion that the AFCI breakers do not really function as intended (at least on series arcing), and that they would not prevent an electrical fire were a “true” series arc fault to actually occur.  This is even “backed up” by a video I found a while back, where someone plugged in a toaster oven to a newly-connected AFCI circuit, loosened one of the receptacle screws, and allowed it to freely arc.  Even in this instance, and under a circuit with a much higher load than a small laptop charger, the AFCI still did NOT trip.  I therefore believe that the protection provided by these AFCI breakers, even if somewhat better than a regular circuit breaker, is still inferior compared to how they are advertised and what they were actually intended for.

   In closing, I do believe the theory of the AFCI was good, and that they had great potential.  However, when they were a new product, they were lobbied into code, largely influenced by the manufacturers.  And, even then, they weren’t ready for “prime time,” as they had tons of nuisance tripping problems that caused countless headaches for homeowners and electricians alike.  In addition, perhaps the testing methods used to certify them per UL requirements are not “real-world” enough, based on what I’ve described above.  Either that, or perhaps they just aren’t sensitive enough anymore so as to prevent the major nuisance tripping issues from returning.  In any case, however, I now seriously doubt as to whether or not these breakers would actually prevent a fire from starting, and it seems to me right now like they may be an unjustified expense.

   If anyone has any thoughts or input on this topic, I’d love to hear!
Logged

If it ain't broke, don't fix it - let old lighting LIVE!

fluorescent lover 40
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: AFCI (Arc-Fault) Breakers Ineffective? « Reply #1 on: Today at 12:31:02 AM » Author: fluorescent lover 40
My house was built in the mid 2000s and has a Square D Homeline panel that is original and unmodified, and the only AFCI breakers it has are for the bedrooms, which still work great, no problems with any kind of nuisance tripping.

I've tripped the one for my room quite a few times over the years, with a few direct shorts too.  ::) A few experiences related to this topic that I've had with the breaker for my room in the past:

1) I got an old C9 light set without a fuse from a thrift store, and one of the lamps had popped internally with a small light show as soon as I plugged the set in, and that tripped the breaker.

2) I was doing some work with the light in my room (hardwired) and had taken it off the ceiling. I had the switch off and when I was putting the light back on my ceiling, I accidentally touched the hot and neutral (didn't cap them off, yeah dumb, I know), and that tripped the breaker.

3) I wired up a 70w HPS yard light to test, it tripped the breaker twice. I was checking connections and when I got to the ground wire connection it felt pretty good, but not good enough, so I made sure it was extra secure and after that the breaker didn't trip.

I personally have not dealt with newer AFCI or dual AFCI/GFCI breakers or receptacles, or others than mine, so I can't comment on them. Mine seem pretty effective for being a little over 20 years old, and while I haven't experienced a serious arcing situation like yours, I'm pretty confident my AFCI breakers can detect the serious fault and trip.
« Last Edit: Today at 12:34:58 AM by fluorescent lover 40 » Logged
Laurens
Member
*****
Offline

View Posts
View Gallery

Re: AFCI (Arc-Fault) Breakers Ineffective? « Reply #2 on: Today at 01:20:15 AM » Author: Laurens
MVMH, all these faults sound like things in devices, and not in your installation. Usually breakers are there to protect the installation, not the devices themselves. If a device arcs you usually notice that pretty quickly, if it starts smoking a smoke detector will pick it up in minutes at worst.
If on the other hand that stuff happens in the walls, it'll take a long time for the smoke to get out somewhere, and by that time things are already veeeeery toasty in there.

Your devices might have filters and suppressor caps in them, and the total length of wire is pretty long so any RF signal the arc makes may be attentuated enough for the breaker not to notice it.

Either way, ask the manufacturer if these breakers are supposed to respond in the scenarios you sketch.
Logged
Medved
Member
*****
Online

Gender: Male
View Posts
View Gallery

Re: AFCI (Arc-Fault) Breakers Ineffective? « Reply #3 on: Today at 01:44:13 AM » Author: Medved
It is important to remember devices like AFCI just can never work 100%, that is just physically impossible.
Problem is, many devices are internally arcing just as a way how they normally work (universal motor brushes, switches,...), so the breakers must not get triggered by that.

The main problem with arcs in 120V systems were parallel arcs, where the just very brief temporary shorts on longer thinner wires (e.g. low current extension cords) cause the peak current limited by the resistance with the few 10's V of the arc drop to be below the insant electromagnetic trigger level (5..10x the nominal breaker rating, so 80..160A for a "16A" breaker), because the pulsed nature the total rms current to be still below the regular breaker trip point (or at least in the area of the characteristics where the response time takes too long), while a lot of heat gets dissipated at a point of the arc, so quickly reaching temperatures igniting fires.
These were the main arcs these breakers were designed to detect distinguish from normal operating currents and so trip. But the signature of these arcs is really such it is easy to differentiate, it is way off even what the craziest loads present. So these work quite reliably.


With arcs in series with some impedance the situation is way more difficult, it is completely different game.
With the working impedance in series there is no way the current could exceed wiring design limits in any form, but the breaker has no other information than the current to work with.
At first there are many devices where the arc in series with an impedance is the way how the device normally works, so you likely don't want your breakers to trip on every fluorescent or HID light, when the physics is exactly the same as an arc on a loose screw...
And if you say "fault arcs usually contain irregularities and frequency rich spikes", well yes, but so do all the universal motors, mainly when the commutator is not 100% clean. Like in an vacuum cleaner, where the air with quite a lot of microscopic dust that has passed the main filter just blows around it. You don't want that to trip the breaker either.

So series arc detection will never be reliable, it is just a probability game, where "catching" some significant percentage of them means reducing the fire risks at least somehow.

This is also the reason, why AFCI's arrived into 230V world later: 230V is high enough any even slight parallel spark immediately turns into fat arc and reaching currents triggering the fast electromagnetic trigger so quickly there is no time to accumulate any dangerous heat, so the parallel arc is not that much a concern as it is on 120V circuits.
The series arcs are the main concern (the higher voltage makes them more persistent), but the difficulty/impossibility to differentiate between fault arc and normal system arcing means there was just no way to have it reliable enough to make sense to use it in normal installations.
Over the years the R&D was able to make the DSP in the breaker differentiate between normal and fault a bit more reliably, plus the regulations on how devices are allowed to behave during normal operation (EMC and harmonics limits getting stricter) means there are less devices that have the tendency to generate mess that could be falsely interpreted as a fault by the breaker, so the AFCI's are becoming a thing, but still they are just a small improvement.

What it is threatening to do is to force people to regularly upgrade and replace the breakers: Rumors say the AFCI's are equipped with an internal timer that makes them to refuse to turn ON (or just artificially fail the self-test) after 10 years, so force their replacement with something more sensitive that becomes available over that time.
So I'm looking with a big worry to the tendencies to "require reliable series arc detection" in electrical codes - it means a ton of false trips mainly when older devices are used and need to regularly buy new set of breakers...
Logged

No more selfballasted c***

Print 
© 2005-2026 Lighting-Gallery.net | SMF 2.0.19 | SMF © 2021, Simple Machines | Terms and Policies