AFCI to rescue?? maybe not??

[peter d]

Cost more money, and make perfectly good appliances work intermittently. AFCI also cannot be tested by any meter I'm aware of; if they pass a self/push button test, they are "good", if not, they aren't.

At this point we can only hope they work better, or get removed in entirety by the 2020/2023 codes. I bet there are many electricians that would gladly remove them for free just to get the enjoyment of hearing them "THUD" in the bottom of a dumpster. and MH forum topic/post count drops 5% overnight lol.

Some people aren't that patient and simply don't install them at all (or remove them after inspection), and avoid the cost of call backs and nuisance tripping. It's a calculated risk, but this is what happens.

 

[peter d]

Honestly, at this point it seems completely inexcusable to me that they don't.

It's been almost two decades. In that time computing power has gone from a 20 pound beast that sat on a desktop, to a far more functional device that fits in your pocket. But for some reason the processing power and algorithms in AFCIs still routinely screw up after 18 years of field trials?

I'm sorry the technology us junk, but it apparently is. At what point do we admit thar 30mA GFP would be a far safer and less nuisance prone technology?

But then again this is the same dinosaur code that still insists on calling a bond wire an equipment ground, so my hopes are pretty slim that they'll realise the error of their ways.

Well said. :thumbsup::thumbsup:

 

[hbiss]

It's been almost two decades. In that time computing power has gone from a 20 pound beast that sat on a desktop, to a far more functional device that fits in your pocket. But for some reason the processing power and algorithms in AFCIs still routinely screw up after 18 years of field trials?

Probably because after all those years promising that this is something that will save thousands of lives and dumping millions into R&D nobody want's to admit that it sounded good in theory but it just doesn't work. I also wonder what the legal ramifications would be. At the very least there would be class action suits with funds set up to compensate customers who were forced to install them. So there is good reason for manufacturers, UL and the NEC to put on their best poker face and hope some slick attorney doesn't catch wind of their game. Throwing in the towel would be an admission of guilt that could cost them millions and might put some of them out of business.

I think the only way for them to exit gracefully and save face is if, somehow it could be shown that AFCIs are no longer needed.

-Hal

 

[JFletcher]

Some people aren't that patient and simply don't install them at all (or remove them after inspection), and avoid the cost of call backs and nuisance tripping. It's a calculated risk, but this is what happens.

I agree. If AFCI worked, they would be embraced as GFCI are, which, to my knowledge, NO electrician has any objection to, because they work and save lives.

If AFCI breakers had a port, like a phone charger, that could be used to update fault conditions, or to field program them NOT to trip because the HO has a certain type of fridge or coffee maker or vacuum, they *might* be worthwhile. or the ability to learn, like a pseudo AI, and only recognize true arcing, or overly high current draw (e.g., a circuit usually draws no more than 5A max for the last year but now is drawing 9A, and spiking here and there from 0-26A), then they'd be worthwhile. but that isnt an AFCI, it's a "Smart Breaker". I'm sure it could be designed and work, but would need an on board computer of sorts to store, process, and analyze the circuit's information.

 

[mbrooke]

Honestly, at this point it seems completely inexcusable to me that they don't.

It's been almost two decades. In that time computing power has gone from a 20 pound beast that sat on a desktop, to a far more functional device that fits in your pocket. But for some reason the processing power and algorithms in AFCIs still routinely screw up after 18 years of field trials?

I'm sorry the technology us junk, but it apparently is. At what point do we admit thar 30mA GFP would be a far safer and less nuisance prone technology?

But then again this is the same dinosaur code that still insists on calling a bond wire an equipment ground, so my hopes are pretty slim that they'll realise the error of their ways.

But you have to keep in mind that beast which fits in your pocket is over $100. If AFCIs had the computing power necessary to discriminate waveforms correctly they would just start at $250 a breaker. The chintziest arc fault relays for POCOs start at several thousand dollars and for good reason:

https://selinc.com/products/451/

https://selinc.com/products/651R/

What arc sense does:

https://selinc.com/solutions/technologies/#arc-sense-technology


And in the least they let you down load the arcing signature to know what really happened:

https://cdn.selinc.com/assets/Liter...lyers/Arc-Sense_PF00160.pdf?v=20161031-073656

At what point do we admit thar 30mA GFP would be a far safer and less nuisance prone technology?

The industry will not admit it, because they would loose a price point. AFCI is technically different then GFCI, so legally they can charge for that feature.

 

[mbrooke]

I agree. If AFCI worked, they would be embraced as GFCI are, which, to my knowledge, NO electrician has any objection to, because they work and save lives.

If AFCI breakers had a port, like a phone charger, that could be used to update fault conditions, or to field program them NOT to trip because the HO has a certain type of fridge or coffee maker or vacuum, they *might* be worthwhile. or the ability to learn, like a pseudo AI, and only recognize true arcing, or overly high current draw (e.g., a circuit usually draws no more than 5A max for the last year but now is drawing 9A, and spiking here and there from 0-26A), then they'd be worthwhile. but that isnt an AFCI, it's a "Smart Breaker". I'm sure it could be designed and work, but would need an on board computer of sorts to store, process, and analyze the circuit's information.

Bzzzzt, wrong. That would be a real AFCI breaker. :thumbsup: Over current relays with arc fault technology not only come with several USB and serial ports to extract the waveform after a trip event, but also the ability to change the arc analysis logic structure. Further, these relays will actually use rate-of change and other algorithms in addition to signature analysis to further aid discrimination. Do you really believe POCOs will put up with hundreds of feeders carrying thousands if not tens of thousands of customers or critical facilities repeatedly tripping out at random times through out their service territory?

 

[macmikeman]

During the last year I have had to install light fixtures with UL stamps on them (legit) purchased by homeowners at the big box stores, that were twenty times more likely to catch on fire (led drivers built into the fixtures) than those rare nail or overdriven staple issues we see once in a decade. Electrical is the most insane and twisted logic industry there is outside of Hollywood.

 

[kwired]

It would be shocking if we knew the actual $$ value of the money spent on call backs for nuisance tripping and replacement of faulty AFCI's.

Well I was mostly just considering the initial investment, but yes, and a majority of that money on callbacks is likely eaten by the EC's who are stuck in the middle on this.

Probably because after all those years promising that this is something that will save thousands of lives and dumping millions into R&D nobody want's to admit that it sounded good in theory but it just doesn't work. I also wonder what the legal ramifications would be. At the very least there would be class action suits with funds set up to compensate customers who were forced to install them. So there is good reason for manufacturers, UL and the NEC to put on their best poker face and hope some slick attorney doesn't catch wind of their game. Throwing in the towel would be an admission of guilt that could cost them millions and might put some of them out of business.

I think the only way for them to exit gracefully and save face is if, somehow it could be shown that AFCIs are no longer needed.

-Hal

What makes a class action suit less likely is the fact that nobody is being injured or killed because of their existence. Unlike the class actions suits for defective medical materials, prescription drug complications, asbestos related health issues, etc. where people are suffering more then just monetary damages from the item involved with the class action suit. If your house burns down and you had AFCI's, it likely would have burned down without them as well.

During the last year I have had to install light fixtures with UL stamps on them (legit) purchased by homeowners at the big box stores, that were twenty times more likely to catch on fire (led drivers built into the fixtures) than those rare nail or overdriven staple issues we see once in a decade. Electrical is the most insane and twisted logic industry there is outside of Hollywood.

don't forget insurance, pharmaceutical, and health care industries. Health care alone may not be so twisted by itself, but the insurance and pharmaceutical industries like to sleep in the same bed and have some influence on it.