AFCI Breakers

[growler]

My thoughts was it was likely the "glowing connection" that caused this fire.

Would AFCI have prevented this? Most likely not.

I'm sure the manufacturers test labs know if the AFCI will detect a "glowing connection".

If it does you would think they would shout it from the mountain tops.

 

[mbrooke]

I feel its been a very productive discussion. While I sincerely hoped to answer the questions posed to the best of my abilities, it's really the questions that are more interesting than the answers. If you sift through all the rhetoric and banter (mine included) and get the discussion down to it's most basic elements, there are obviously major gaps in the messaging that AFCI advocates are providing and the information you guys actually want and need. I know your frustration and skepticism is real. In my travels to the various events and meetings that I attend, I get many of the same inquires and sense many of the same doubts. [/CODE]

I agree, a lot of good info

I think understanding theory is essential.

The history, purpose and details of the 15+ year old UL 1699 standard is apparently not well understood. While the cost to own a copy is certainly a barrier, it's surprising to me how many electrical professionals have never read it or seen the tests performed. Industry needs to do a better job getting this information out to the general electrical workforce that are dealing with AFCIs day-in and day-out. This isn't generally true for most of the other listed products. For the most part, electrical professionals are more interested in the code requirements and do not have much interest in the product standards. Since AFCIs seem to be the exception, this information should probably be added to presentations and in AFCI literature.

My understanding is that the price, copy right, confidentiality, and limited accesses to witness AFCI testing prevents electricians and other experts from knowing much about it. Most are just left in the dark. In fact I cant even find a You Tube video of 1699 testing. Why is information so hard to get? Redacting information only fuels conspiracy theories.


Also, correct me if I am wrong, but UL tests for non combo branch feeder AFCIs could be passed via GFP and low (75amp) magnetic trip?

Unwanted and nuisance tripping of AFCI devices and the methods for troubleshooting the protected circuit seems to be the number one complaint from the field. Interestingly, the reports that I have received indicate essentially an "all-or-none" phenomenon. That is, it seems like you either have nearly no problems with AFCIs or you have nothing but trouble with AFCIs. This is not easy to explain.

I believe that may come from the inability for AFCIs to differentiate between normal and abnormal arcs. Manufacturers are left with either making the device more sensitive to dangerous arcs and have it nuisance trip or less sensitive to dangerous arcs so it can hold. There is simply not enough computing power to differentiate between arcs reliably. Assuming dangerous arcing is a real concern to start with this is one of the biggest reasons why AFCIs are a gimmick.

Far more computing power is needed for volts 120 AFCIs to be effective. Arc logic used where profit truly depends on quality rather than mandatory requirements leads to substantially more computing power.

I have a network of electrical contractors in all of my (11) states in the south that voluntarily report to me their experiences with AFCIs and this holds true for them as well. Most are really not having or have had very few problems to report. A few are reporting problems on nearly every installation. The free training program from UL Knowledge Services does provide some good procedures and tips for troubleshooting AFCI protected circuits, but more needs to be offered.

A lot more needs to be offered.

The last significant matter I am taking away from this discussion is the issue with reported fire data. It's quite clear you all want solid evidence that shows as the installation of AFCI devices increase, the number of residential fires from electrical faults are decreasing. We have really good correlating data of this nature for GFCIs, but that data goes back to around 1975. Electrocution deaths from the use of a consumer product is easier to report and track. As GFCIs got expanded to various locations throughout a building, the number of deaths in those areas started to dramatically decline. It's a no-brainer. Fire data is a bit more tricky. More work and effort needs to be dedicated to this task.

That is because most portable metal frame appliances did not have an EGC back then. GFCIs came about to solve an existing well known problem hence why the statics fell. Requiring EGCs would have done the same.

FWIW, in the UK since the 50s all appliances with metal frame appliances were required to have an EGCs. Even a plastic shell toaster with metal lined slots needed to have those connected to an EGC. This allowed RCDs (GFCIs) to be delayed until the 90s.


AFCIs on the other hand are chasing big foot or aliens. We dont even know if they exist, or what they look like.

Anyway, I will be at the South Carolina BCC - Code Study Committee Meetings this week in Columbia. Next week is the IAEI Georgia Chapter Meeting being held just outside of Atlanta. The week after I will be instructing a 2-hour presentation on Retrofit Kits for Lighting and Signs at (6) locations throughout the state of Alabama. And for the last week of the month I will be in Nashville, TN for the NASFM annual conference. If any of you happen to be in the area or plan to attend these events, please track me down and introduce yourself. I would love to meet you in person. I swear I do not have horns, red scaly skin, and a spiky tail. Nor will I try to serve you any "kool-aid".

Any info you have you are welcome to share

 

[epimetheus]

Again, your hanging your hat on the opinion and analysis of one person. The author is wrong. The arcing fault current is in series. This is the exact exhibit from the standard: (Follow the flow of current - it has but one path)

First, I think part of the confusion with 40.4 test comes from how the damage to the prepared sample would occur in reality. The image you linked from the standard shows the the prepared damage across the strands of the SPT-2 lamp cord. Normally there would be a load at the end of that cord, thus making this fault, should it occur in reality, a parallel arc. I don't disagree that the arc is in series with the AFCI in the exhibit, but there's no load included in the circuit. I think this is a terminology issue. Series arc implies that the arc is in series with a load. Parallel arc implies that the arc is in parallel with the load. When there's no load, a series arc (per my implied definition) cannot exist, but a parallel arc (per my implied definition) is technically in series with the AFCI.

Based on the way I see this test procedure, I think using 40.4 to claim series arc protection is a bit of a misnomer. I would be satisfied if the test included a varying load on the other end of the SPT-2 sample and the sample was prepared with a cut or gap in a single conductor that was then pyrolized (sp?) per the existing test procedure.

Second, regarding the fire data that prompted the development of AFCI in the first place, as with others, I don't see how there is any conclusive evidence that arcing is the major cause of home electrical fires. Of course it's the most logical assumption. Anyone who has been in this industry for five minutes, or has been through a primary school physics class, can logically assume that electrical arcing and sparks cause fires. That's my guess how all this got started. It's a perfectly logical assumption. Unfortunately, mandated requirements like AFCI's on every circuit, at significant additional cost and with questionable performance capabilities, should not be based on assumptions without conclusive evidence, not matter how logical.

I'm an electrical engineer in the power industry by day and I've been exposed to the effort to identify arcing faults on medium voltage distribution circuits. Based on that exposure, and the logic behind it, I would want that protection in my home. I was even considering installing AFCI's in my 1974 house until I started reading up on the history and controversy surrounding them. Just like the effort to reliably identify medium voltage arcing faults, AFCI detection technology on it's own does not appear to be mature enough to reliably detect only dangerous arcing. The originally included GFPE functionality offered something in addition to the arc sensing. Without that, I don't see any value in the arc sensing technology alone. That, and the lack of any reasonable ability to to field test the arc sensing circuits prevent me from supporting AFCI in it's current form. FWIW, medium voltage arc sensing is done with complex algorithms in a multi-thousand dollar relay with powerful microprocessor and sensitive current and voltage sensing elements. If that technology has not gained widespread adoption, how can one expect a tiny molded case breaker to do the same reliably?

Finally, why aren't UL standards available for public viewing? If most items we use on a daily basis are listed to some UL standard, why shouldn't we be able to look up that read that standard? Why should a conscientious home over have to pay hundreds of dollars to see how a protective device is their home is tested?

 

[don_resqcapt19]

...

Second, regarding the fire data that prompted the development of AFCI in the first place, as with others, I don't see how there is any conclusive evidence that arcing is the major cause of home electrical fires. Of course it's the most logical assumption. Anyone who has been in this industry for five minutes, or has been through a primary school physics class, can logically assume that electrical arcing and sparks cause fires. That's my guess how all this got started. It's a perfectly logical assumption. Unfortunately, mandated requirements like AFCI's on every circuit, at significant additional cost and with questionable performance capabilities, should not be based on assumptions without conclusive evidence, not matter how logical. ...

...

Is it? I have seen no evidence that self sustaining arcs even exist at dwelling unit voltages. The most common cause of electrical fires is joule heating at points of connection.

 

[epimetheus]

Is it? I have seen no evidence that self sustaining arcs even exist at dwelling unit voltages. The most common cause of electrical fires is joule heating at points of connection.

Ok, maybe I should have said "a logical assumption", not the most logical assumption. I don't disagree with you, but when most people see an electrical arc or sparks, they think "Wow, that could start a fire!" Most people don't know that arcs don't sustain at dwelling unit voltages. For the record, I don't think sustained arcs are the concern. Again, I think terminology is a problem here. It's intermittent arcs and sparks that AFCI's are trying to protect against. It's been well established in this thread that joule heating and glowing connections are not prevented by AFCI technology.

 

[PetrosA]

I think the foundation of the backlash from electrical professionals regarding AFCI technology has been pretty well described and it seems that Bryan is beginning to understand it.

I'll just recap the arguments here and add a few more:

- The fact that AFCI tech was introduced with misleading talk from a lot of parties
- The fact that they create a lot of nuisance trips for many installers and troubleshooters
- The fact that field testing them is not possible by any quantifiable methods
- The fact that the standards and test methods are difficult/expensive to obtain
- The fact that AFCI requirements have been one of the most expensive additions to the code in the last two decades, and the hardest to justify/explain to customers
- The fact that unlike GFCI technology, there is no way to prove to a customer that they actually protect anything
- The fact that, in spite of the manufacturers' claims, no insurance companies appear to be giving discounts for a device that supposedly protects against fires, nor do they refuse to underwrite homes without the technology installed

Any fraction of these arguments is enough to make a lot of people suspicious, especially ones who deal with AFCIs on a regular basis. The manufacturers have a lot of work to do yet. In any other context we'd just say that it's an example of horrendous customer service, but when it's required by code, it becomes more than that. It becomes a bone of contention.

 

[ActionDave]

This thread opened on 6-2-2015 and by 6-30-2015 it had 298 posts but was not included in the Code Forum Recap - June 2015 email newsletter. What's up with that?

 

[GoldDigger]

Conspiracy theory?

 

[GeorgeB]

I have seen no evidence that self sustaining arcs even exist at dwelling unit voltages.

Don, it's not wiring components, but when I was a youngster, we removed the carbon positive electrode from D cells, used salt water connecting 2 nails in a container as a resistor, and held arcs for minutes on 115V house electricals. I was excited when a local theater owner gave me several of his projector carbons; didn't work any better, but looked a lot nicer.

 

[macmikeman]

Conspiracy theory?

I could run that thru my filter and publish the results at ET if you wish......

 

[GoldDigger]

Don, it's not wiring components, but when I was a youngster, we removed the carbon positive electrode from D cells, used salt water connecting 2 nails in a container as a resistor, and held arcs for minutes on 115V house electricals. I was excited when a local theater owner gave me several of his projector carbons; didn't work any better, but looked a lot nicer.

As mentioned earlier in the thread, and in other threads, physics tells us that the conditions for sustaining an arc(reigniting every half cycle) are very different for metals than for carbon.
Hence the need to manufacture a carbonized path as the first step in UL testing.
Now if we used fused carbon fiber for wiring....

 

[mbrooke]

First, I think part of the confusion with 40.4 test comes from how the damage to the prepared sample would occur in reality. The image you linked from the standard shows the the prepared damage across the strands of the SPT-2 lamp cord. Normally there would be a load at the end of that cord, thus making this fault, should it occur in reality, a parallel arc. I don't disagree that the arc is in series with the AFCI in the exhibit, but there's no load included in the circuit. I think this is a terminology issue. Series arc implies that the arc is in series with a load. Parallel arc implies that the arc is in parallel with the load. When there's no load, a series arc (per my implied definition) cannot exist, but a parallel arc (per my implied definition) is technically in series with the AFCI.

I agree, this in reality is a parallel fault. To me a series fault would be where only one side of the zip conductor is cut and the two are making contact.

In the real world such a fault can be taken care without AFCI. The original concern was a parallel fault may not reach the breaker's magnetic trip curve. Requiring a fuse within each cord cap would mitigate that problem.

Yet somehow the most complex solution was choosen.

Based on the way I see this test procedure, I think using 40.4 to claim series arc protection is a bit of a misnomer. I would be satisfied if the test included a varying load on the other end of the SPT-2 sample and the sample was prepared with a cut or gap in a single conductor that was then pyrolized (sp?) per the existing test procedure.

I agree, if the true intent is a series arc it must be carried out differently to reflect the real world.

Second, regarding the fire data that prompted the development of AFCI in the first place, as with others, I don't see how there is any conclusive evidence that arcing is the major cause of home electrical fires.

To this day not a single shred of evidence exists that 30,000 home fires a year come from arc faults.

Of course it's the most logical assumption. Anyone who has been in this industry for five minutes, or has been through a primary school physics class, can logically assume that electrical arcing and sparks cause fires. That's my guess how all this got started. It's a perfectly logical assumption. Unfortunately, mandated requirements like AFCI's on every circuit, at significant additional cost and with questionable performance capabilities, should not be based on assumptions without conclusive evidence, not matter how logical.

In my eyes its clever marketing. Its a logical assumption to none professionals. Since kids we have seen electricity depicted in books, media, cartoons and the like as lightening bolts. In school we are taught lightening during a thunderstorm is electricity. Curriculum often start with lightening, moving to difference in charge potentials, and then how conventional electricity works. Pull a plug under load out of the wall and people see a spark. Guy who gets hit with lightening in movies looks like toast. Its becomes engrained into everyone mind electricity automatically equals arcing, and arcing equals destruction.

Once respected institutions come out claiming arcing is behind dwelling fires and here is a device to stop that, absolutly no one has any reason to question it. The hand fits perfectly into the already molded glove.

However, just like in cartoons where a rectangular iceblock appears around a frozen person, in real life science and experience proves that not to be the case.

I'm an electrical engineer in the power industry by day and I've been exposed to the effort to identify arcing faults on medium voltage distribution circuits. Based on that exposure, and the logic behind it, I would want that protection in my home. I was even considering installing AFCI's in my 1974 house until I started reading up on the history and controversy surrounding them. Just like the effort to reliably identify medium voltage arcing faults, AFCI detection technology on it's own does not appear to be mature enough to reliably detect only dangerous arcing. The originally included GFPE functionality offered something in addition to the arc sensing. Without that, I don't see any value in the arc sensing technology alone. That, and the lack of any reasonable ability to to field test the arc sensing circuits prevent me from supporting AFCI in it's current form. FWIW, medium voltage arc sensing is done with complex algorithms in a multi-thousand dollar relay with powerful microprocessor and sensitive current and voltage sensing elements. If that technology has not gained widespread adoption, how can one expect a tiny molded case breaker to do the same reliably?

This is profound. Something few know about, but key. Arc sensing in medium voltage systems is far more mature and much more reliable, coming only from exceptional computing power and decades of complex research both of which lack in residential AFCIs. If anything this clearly demonstrates AFCIs are a gimmick. Simple flow chart logic looking for peaks and shoulders is not enough, its a joke. You need extraordinary computing power. Joe Engels was the one who envisioned a central processor, and now I know why.

And there is an irony. One of the reasons why arc logic was created for medium voltage distribution systems came from the inability to set zero sequence (GFCI) relays down to low values. The reason being multi grounded neutral distribution systems, where downed conductor would go unnoticed. Fault current simply looked like normal neutral current, hence the need to create another relay involving discrimination of downed conductors.

Where all loads are connected phase to phase, it is possible to set the ground pick up relays to only a few amps with a definite/instantaneous time current curve. This actually works out very well indicating nearly all down conductor and high impedance faults.

In residential wiring the EGC does not carry neutral current, so there is nothing stopping the US of GFP, especially when UL own testing determined over driven staples always involve the EGC.

Finally, why aren't UL standards available for public viewing? If most items we use on a daily basis are listed to some UL standard, why shouldn't we be able to look up that read that standard? Why should a conscientious home over have to pay hundreds of dollars to see how a protective device is their home is tested?

I would say there are to many inconvenient truths. BPH was able to go into great detail about the cord specimen leaving out nothing, but quietly left out 15,000 volts. This is how the industry gets away with it. They hide evidence both passively and actively hoping we take their word. Any working in the legal justice system sees criminals pull this trick on a daily bases.

 

[epimetheus]

Arc sensing in medium voltage systems is far more mature and much more reliable, coming only from exceptional computing power and decades of complex research both of which lack in residential AFCIs. If anything this clearly demonstrates AFCIs are a gimmick. Simple flow chart logic looking for peaks and shoulders is not enough, its a joke. You need extraordinary computing power. Joe Engels was the one who envisioned a central processor, and now I know why.

I would fully support arc detection and tripping at the main breaker level if I knew the detection methodology was mature enough to be reliable and secure. In the world of smart meters, maybe we need a microprocessor based relay on the main. This relay (I'm thinking something along the lines of a SEL-751A) could perform arc detection. It could adjust it's detection based on each consumer's usage characteristics and installed wiring. Maybe you could teach the relay what is a false detection. Of course, all this is well above the average consumer's head, so probably not something that most consumer's will want to bother with.

 

[mbrooke]

I would fully support arc detection and tripping at the main breaker level if I knew the detection methodology was mature enough to be reliable and secure. In the world of smart meters, maybe we need a microprocessor based relay on the main. This relay (I'm thinking something along the lines of a SEL-751A) could perform arc detection. It could adjust it's detection based on each consumer's usage characteristics and installed wiring. Maybe you could teach the relay what is a false detection. Of course, all this is well above the average consumer's head, so probably not something that most consumer's will want to bother with.

That is what it actually takes. Utility relays have the luxury to be programed to do just about anything, in addition to being able to plug in a laptop to tell you exactly what tripped it.

Arent there arc fault relays that even tall you how far down the line the event took place?


BUT, I will be as pessimistic as usual . Electronics can fail, where as an electromechanical RCD can survive.

 

[kwired]

Ok, maybe I should have said "a logical assumption", not the most logical assumption. I don't disagree with you, but when most people see an electrical arc or sparks, they think "Wow, that could start a fire!" Most people don't know that arcs don't sustain at dwelling unit voltages. For the record, I don't think sustained arcs are the concern. Again, I think terminology is a problem here. It's intermittent arcs and sparks that AFCI's are trying to protect against. It's been well established in this thread that joule heating and glowing connections are not prevented by AFCI technology.

But there are intermittent arcs in equipment that are "normal operational" characteristics. Flip on a typical lighting switch and there is an intermittent arc that takes place. These devices need to be capable of being very discriminating at determining what is acceptable and what isn't and that is causing a lot of trouble because they currently can not do this to user expectations.

Don, it's not wiring components, but when I was a youngster, we removed the carbon positive electrode from D cells, used salt water connecting 2 nails in a container as a resistor, and held arcs for minutes on 115V house electricals. I was excited when a local theater owner gave me several of his projector carbons; didn't work any better, but looked a lot nicer.

How likely is it for such a carbon rich path to develop in a typical dwelling?
How likely is it for there to be an intentional path through a commutator/carbon brush in a motor operated appliance in a typical dwelling?

 

[don_resqcapt19]

Don, it's not wiring components, but when I was a youngster, we removed the carbon positive electrode from D cells, used salt water connecting 2 nails in a container as a resistor, and held arcs for minutes on 115V house electricals. I was excited when a local theater owner gave me several of his projector carbons; didn't work any better, but looked a lot nicer.

An arc as in power passing through an air gap?

 

[GeorgeB]

An arc as in power passing through an air gap?

We touched them together, then pulled them back; I didn't know then, but now think I know it is plasma, ionized something or other.

Other comments on carbon and metal being different ... arc welders use steel electrodes and steel parts being welded; Do AC arc welders do more than attain a plasma "bubble"? Is it metal ions that do the conductive part? I don't know what's special about a "buzz-box" AC welder transformer, but is there something to establish a low power factor keeping current and voltage out of phase?

 

[kwired]

We touched them together, then pulled them back; I didn't know then, but now think I know it is plasma, ionized something or other.

Other comments on carbon and metal being different ... arc welders use steel electrodes and steel parts being welded; Do AC arc welders do more than attain a plasma "bubble"? Is it metal ions that do the conductive part? I don't know what's special about a "buzz-box" AC welder transformer, but is there something to establish a low power factor keeping current and voltage out of phase?

I really have no idea but am going to guess when welding steel you can get away with using AC current because of the carbon in the steel, and maybe other reasons on top of that. When welding most other metals you need to use DC current - Aluminum is one exception and generally must be welded with AC current.

 

[gadfly56]

We touched them together, then pulled them back; I didn't know then, but now think I know it is plasma, ionized something or other.

Other comments on carbon and metal being different ... arc welders use steel electrodes and steel parts being welded; Do AC arc welders do more than attain a plasma "bubble"? Is it metal ions that do the conductive part? I don't know what's special about a "buzz-box" AC welder transformer, but is there something to establish a low power factor keeping current and voltage out of phase?

When carbon arc lamps were more popular, this was exactly the procedure for getting them started. It's called "striking the arc". Since the carbon is consumed in the process, you have to keep adjusting the gap to maintain the arc. I recall reading somewhere about automatic mechanisms that were invented for that purpose.

 

[mbrooke]

FWIW as usual it looks like the Chinese are copying our business skills:


http://www.afcimaster.com/En/


And again, more proof you need a heavily carbonized gap in order to sustain an arc:


https://www.youtube.com/watch?v=tTdCU0GeWBE