'Proof' that AFCI devices really locate arcs.

[templdl]

Unless you have an AFCI that tells you what the trip was, everyone is guessing. It could be an overload, a short circuit or high current ground fault, a low current ground fault, an arcing fault (assuming they even exist at dwelling unit voltages) or is just tripped because it wanted to.

That being said, I have had some discussions with AFCI design engineers and they believe that 80% or more of the trips are ground fault trips. Many of those from poor wiring practices.

I totally agree. The problem is the we are responding to an incident and then apply our best electrical knowledge, experience, and expertise to diagnose what the caused of trip was. As I have said before I am a believer in the AFCI technology as my home was BETA for them just before their release as a knew one of the design engineers personally and regularly communicated with many others. Since their release some 15+ years ago I have been very disappointed that the documentation from the field has been slim to none. I would have though that the insurance companies would have given then great accolades but silence. I'm not trying to be sarcastic but where does that 80% come from since the first documentation appears to be slim to none at best?
Make no mistake that I am a supporter of AFCI technology but have been extremely disappointed that there hasn't been more positive feedback from the field. This is frustrating.

 

[growler]

We used some newer CH AFCIs last year on a project. Help went there this morning and via discussion with homw owner determined that the fault was a series arc. They found the guilty extension cord with the series arc.

Proof #1 for us.

I don't think there is any doubt that an AFCI can detect an arc. The question is, how many arcs are out there for them to detect. I have not seen any so far except for the arcing of electric motors.

 

[macmikeman]

Nobody will produce proof , because there is none. When they take out the gfi function, you are left with a plain old style breaker that has a coiled tail and a trip button. $$$

 

[don_resqcapt19]

I totally agree. The problem is the we are responding to an incident and then apply our best electrical knowledge, experience, and expertise to diagnose what the caused of trip was. As I have said before I am a believer in the AFCI technology as my home was BETA for them just before their release as a knew one of the design engineers personally and regularly communicated with many others. Since their release some 15+ years ago I have been very disappointed that the documentation from the field has been slim to none. I would have though that the insurance companies would have given then great accolades but silence. I'm not trying to be sarcastic but where does that 80% come from since the first documentation appears to be slim to none at best?
Make no mistake that I am a supporter of AFCI technology but have been extremely disappointed that there hasn't been more positive feedback from the field. This is frustrating.

So you believe that there is evidence that self sustaining arcs exist at dwelling unit voltages. I just can't find any thing to document that.

As far as the 80% being GFP, just opinions of some of the design engineers that I have talked with.

I don't see the insurance companies looking at this for a long long time, if ever. Based on the idea that at least 85% of the dwelling unit fires that are said to be of electrical origin occur in units at least 20 years old, it will be many years before there would be any statistical data to show the effectiveness of AFCIs, and even if there is a statistical reduction of electrical fires over time, will they really be able to say that the AFCI caused that reduction?

 

[templdl]

So you believe that there is evidence that self sustaining arcs exist at dwelling unit voltages. I just can't find any thing to document that.

As far as the 80% being GFP, just opinions of some of the design engineers that I have talked with.

I don't see the insurance companies looking at this for a long long time, if ever. Based on the idea that at least 85% of the dwelling unit fires that are said to be of electrical origin occur in units at least 20 years old, it will be many years before there would be any statistical data to show the effectiveness of AFCIs, and even if there is a statistical reduction of electrical fires over time, will they really be able to say that the AFCI caused that reduction?

I may have been misunderstood as my believing there is field evidence of arc issues as that is what is disappointing to me that there appears to be no evidence what so ever. You are right regarding the lack of documentation which is the point that I failed to make as it has been slim to none if there is any documentation at all. Other than exposing ground faults of various types arc fault (other than those to ground that is) examples are conspicuously missing and this has been very discouraging to say the least. I would have expected to have seen at least a few examples.

 

[ptonsparky]

I don't think there is any doubt that an AFCI can detect an arc. The question is, how many arcs are out there for them to detect. I have not seen any so far except for the arcing of electric motors.

Well, that was the first series arc, that the CB diagnosed, for us. The rest were ground faults. One of which, same house, the CB diagnosed. We dont do a lot of houses so we may have sold 40 afci over the years. One series arc out of forty installed, but of those having problems, maybe one of ten. WAG at best.

 

[cowboyjwc]

It's like I've said for years, we have no idea how many lives are saved by GFCI's, because no one reports that they almost got shocked, but we can tell you how many died because they didn't have a GFCI.

So with AFCI's you're not going to see a report that it tripped because someone sent a sofa on the lamp cord and cut the hot wire in half. You will see a report because the lamp cord was damaged and started a fire. May have even started a fire with an AFCI installed, because once the arc has happed it is possible that even if the AFCI tripped that it could have set a material on fire.

 

[growler]

So with AFCI's you're not going to see a report that it tripped because someone sent a sofa on the lamp cord and cut the hot wire in half. You will see a report because the lamp cord was damaged and started a fire. May have even started a fire with an AFCI installed, because once the arc has happed it is possible that even if the AFCI tripped that it could have set a material on fire.

They don't even test the AFCI for that scenario. The lamp cord (extension cord).

Think about it, a lamp would normally have a couple of 60-75 watt bulbs (lamps) for a current draw of around 1-1.5 amps. They test the AFCI at 5-10-15-20 amps of current draw. Even in the lab what will it do with a 1 amp load.

 

[ActionDave]

Could be, but do you know for sure or are you just guessing?

I feel very comfortable taking a bet for how ever much money you are willing to put up that it was GF trips.

 

[cowboyjwc]

They don't even test the AFCI for that scenario. The lamp cord (extension cord).

Think about it, a lamp would normally have a couple of 60-75 watt bulbs (lamps) for a current draw of around 1-1.5 amps. They test the AFCI at 5-10-15-20 amps of current draw. Even in the lab what will it do with a 1 amp load.

Well then with no more information than that I would say that they are not worth the plastic they are made of, because that's exactly what they were supposed to test for. Not just arc faults in the wiring system, but anywhere on the system, because people were running cords all over the place and they were getting cut and smashed and even bad appliance cords.

 

[ActionDave]

Well then with no more information than that I would say that they are not worth the plastic they are made of, because that's exactly what they were supposed to test for. Not just arc faults in the wiring system, but anywhere on the system, because people were running cords all over the place and they were getting cut and smashed and even bad appliance cords.

That's why so many of us hate putting them in, extra cost, no extra safety, add in the nuisance trip headaches.....there is no upside for the EC or the HO.

 

[big john]

It's like I've said for years, we have no idea how many lives are saved by GFCI's, because no one reports that they almost got shocked, but we can tell you how many died because they didn't have GFCIs...

But we couldmake the exact same claim about a normal thermal-magnetic breaker or just about any other protective device.

The proof of operation comes from reproducible testing and field reports of witnessed operation: We have all seen GFCIs operate to reduce a hazard. It should raise some serious questions that we are a decade-and-a-half into widespread AFCI use and stilll have very little evidence of either of those to bolster the necessity of AFCI use.

Thatt we as professionals--the ones with the most experience with these devices--are still widely unconvinced after years of use, speaks to a real problem.

 

[don_resqcapt19]

Well then with no more information than that I would say that they are not worth the plastic they are made of, because that's exactly what they were supposed to test for. Not just arc faults in the wiring system, but anywhere on the system, because people were running cords all over the place and they were getting cut and smashed and even bad appliance cords.

The original branch circuit and feeder type AFCIs did not look for an arc signature unless the current was 75 amps or more. The combination type we are now required to use looks for a series arc if the current is 5 amps or more and looks for a parallel arc if the current is 75 amps or more.

 

[romex jockey]

Reducing the issue anecdotal evidence is on par with us opining on global warming

Holding the manufacturers feet to the fire would be far more revealing of afci capability

The exact testing (UL1699 40.4) procedure juxtaposed with the standard repeated via multiple NRTL's would have my attention

Facts, not fiction please....

~RJ~

 

[mbrooke]

I don't think there is any doubt that an AFCI can detect an arc. The question is, how many arcs are out there for them to detect. I have not seen any so far except for the arcing of electric motors.

For me there is. There is simply not enough logic inside an AFCI to discriminate correctly between normal and abnormal arcing. This is why they trip on UL listed appliances. Whose to they the wont do the opposite on a real arc?

 

[templdl]

For me there is. There is simply not enough logic inside an AFCI to discriminate correctly between normal and abnormal arcing. This is why they trip on UL listed appliances. Whose to they the wont do the opposite on a real arc?

Interesting, how did you conclude that there is not logic? Don certainly appears t have a good handle on the function of the AFCI as he seems to have comments from good authority from the manufacture.
Is your comment regarding the lack of logic from consulting with the manufacturer or is it based upon your assumption? In your opinion how much more logic would be required?
If you have read my previous posts on the subject you shoud be familiar weith my excperience snd opinions of he AFCI.. Thousands have been installed over the past 15 years and you would think that there would be some instances reported other that L-G faults.

 

[mbrooke]

Interesting, how did you conclude that there is not logic? Don certainly appears t have a good handle on the function of the AFCI as he seems to have comments from good authority from the manufacture.
Is your comment regarding the lack of logic from consulting with the manufacturer or is it based upon your assumption? In your opinion how much more logic would be required?
If you have read my previous posts on the subject you shoud be familiar weith my excperience snd opinions of he AFCI.. Thousands have been installed over the past 15 years and you would think that there would be some instances reported other that L-G faults.

It comes from research, and the opinions of those who research and make utility grade arc fault relays. Its not that there is no logic, just not enough. Discrimination between normal and abnormal current ripples is an entire science involving logic, math, data processing ect that has been researched for decades, let alone what a dangerous arc looks like in a dwelling. Simply applying flow chart logic on peaks and shoulders is like doing pediatric neuro surgery with a hammer, it looks good to laymen but in reality does little if anything. For starters a lot of the concepts used to reliably detect arcs actually involve artificial neural networks...

 

[user 100]

But we couldmake the exact same claim about a normal thermal-magnetic breaker or just about any other protective device.

The proof of operation comes from reproducible testing and field reports of witnessed operation: We have all seen GFCIs operate to reduce a hazard. It should raise some serious questions that we are a decade-and-a-half into widespread AFCI use and stilll have very little evidence of either of those to bolster the necessity of AFCI use.

Thatt we as professionals--the ones with the most experience with these devices--are still widely unconvinced after years of use, speaks to a real problem.

Absolutely. Proof is all we're really asking for. That is what one of the "true believers" (a scion of afci marketing) on another afci thread (which seems to have been closed for some "mysterious" reason) on this forum couldn't seem to grasp and when pressed he only offered what amounted to speculatory conjecture that cannot be replicated at residential voltages. In this trade most of us try to do our best and you have literally tens of millions of safe residential installations without afci technology that have not, nor will ever, have an electrical fire due to negligence on behalf of a licensed qualified electrician. Considering how deadly even one loose joint can be (which an afci won't detect), that alone should speak volumes about necessity of afcis for fixed premises wiring. The claims about afcis preventing fires from damaged cords-very noble, possibly useful idea there, but prove it works. We don't need these sham devices "protecting" homes if there is no proof they protect anything.

 

[templdl]

For me there is. There is simply not enough logic inside an AFCI to discriminate correctly between normal and abnormal arcing. This is why they trip on UL listed appliances. Whose to they the wont do the opposite on a real arc?

Interesting, how did you conclude that there is not logic? Don certainly appears t have a good handle on the function of the AFCI as he seems to have comments from good authority from the manufacture.
Is your comment regarding the lack of logic from consulting with the manufacturer or is it based upon your assumption? In your opinion how much more logic would be required?
If you have read my previous posts on the subject you shoud be familiar weith my excperience snd opinions of he AFCI.. Thousands have been installed over the past 15 years and one would think that there would be some instances reported other that L-G faults. I'll be the first in line to see some relevant events is response to series and L-N arcing. As a berliever in the AFCI technology I have been very disappointed.

 

[mbrooke]

Interesting, how did you conclude that there is not logic? Don certainly appears t have a good handle on the function of the AFCI as he seems to have comments from good authority from the manufacture.
Is your comment regarding the lack of logic from consulting with the manufacturer or is it based upon your assumption? In your opinion how much more logic would be required?
If you have read my previous posts on the subject you shoud be familiar weith my excperience snd opinions of he AFCI.. Thousands have been installed over the past 15 years and one would think that there would be some instances reported other that L-G faults. I'll be the first in line to see some relevant events is response to series and L-N arcing. As a berliever in the AFCI technology I have been very disappointed.

Uh, same post:

http://forums.mikeholt.com/showthread.php?t=170744&page=4&p=1661572#post1661572