ECN Forum Forums Code Related Discussion NEC & other Code issues Arc Fault Breakers

This is in response to Al, I will try to address Don's issues later today or tomorrow.

Arcing faults are not really hard to understand. The thermal and magnetic elements in circuit breakers protect are there for one thing: to protect the wire from being damaged by excessive current.

The thermal element is a bimetal that provides overload protection to keep the amount of continuous current below the level that the wire can safely carry. #12 = 20 A, etc.

The electromagnet protects against short-circuits and is commonly called the instantaneous trip. However, the use of the word "instantaneous" is a bit misleading. Breakers must allow for motor inrush currents, which can be 6-10 times full load. So, a 20 amp breaker needs to be able to ride through temporary high currents up to 120 or so amperes so that they do not nuisance trip. At extremely high current levels 800% of handle rating and greater, the breaker trips in less than a second.

In a typical home, the available fault current downstream from the loadcenter is very low ~ maybe 100 amps. So, what happens when you when you get a break in the insulation is that you get short, low-current level arcs. The arc is not present long enough to trip on thermal, and it is not of a high enough magnitude to trip on instantaneous. However, the plasma arc can exceed 6000 deg. C., and explode molten copper as it arcs.

In order to see that this is happening, we analyze the current waveform and look for the spikes that indicate that there is an arc. We sample for about 8 half cycles, which is a long enough period of time that we ride through normally occurring arcs. So basically, we will see the arc, then keep sampling for a given period of time, and if we see the arc there again, we de-energize the circuit.

Hope this helps explain what the product does. Let me know if you have any other questions.

-Brendan

Quick note to Sparky:

I feel pretty certain that there is not a 60 amp, 2 pole AFCI. There are 60 amp, two pole GFCI breakers, but no AFCI of which I am aware.I believe that Cutler-Hammer is the only manufacturer with a two-pole, and we stop at 20 amps.

There are actually 2 different two pole AFCI available. One is a standard 240 volt, 2-pole, and the other is what we call a 2 pole independent trip (AFIT). It allows for AFCI protection on circuits that contractors refer to as “multi-wire,” “shared neutral,” or “home run.” These are two individual 120-volt circuits that share a neutral. The AFIT is a two-pole breaker with two independent 120-volt circuits. The breakers trip independently on overloads or short circuits, but trip in common on an arcing fault. For contractors that use 3-wire circuits this is a huge labor savings over wiring two individual 1 pole breakers, and it will allow them to continue this wiring practice.

Don,

I agree with wanting to know how they work and how reliable they are. I will be on the 'Front Lines' myself trying to explain to the customer what they are and why they are required. I will also be explaining that nothing is 100% guaranteed and it is by no means meant as a substitute for safe wiring practices.

I don't think that there should be any question as to whether they should be required at the present time. We have many safety and potentially life saving devices that are only partially effective. We still use them though in the hopes that it may work this time. As far as the testing goes, it's important that the public be educated about that.

Bill

[This message has been edited by Bill Addiss (edited 12-05-2001).]

Bill,
I think there are a lot of reasons to question the installation of AFCIs at this time.
1) I'm not convinced that they will function to stop the fires that the manufacturers say they will. They say they only work effectively on parallel arc faults. I've always been taught that 90% of electrical failures are at connections. This would be a series fault, not a parallel fault.
2) The fire statistics used to get these items into the code are faulty.
3) If they aren't fail safe they shouldn't be permitted on the market.
4) If they don't self test they won't work when needed. The average homeowner won't test them.
6) I'm glad I don't do any residential. I see a lot of potential liability for the contractor when these things don't work and there is a fire.
Of course once the code is adopted in my area I will support the use of AFCIs. Until that time I will continue to question the requirement until I'm convinced that they really work as advertised.
Don(resqcapt19)

Brendan,

Thanks. . .this whets my appetite. 8 half cycles, all one side of zero volts, or 8 half cycles equal to 4 full cycles?

Is that half cycles of just current? Are you sensing with a current transformer?

Is a snapshot or other representation taken of the suspect arc to compare with the next arc that would initiate the Interrupter? Is the "memory" longer than 8 half cycles, and if so, how much longer?

Is there a threshold level of arc activity that is tolerated, and what are the parameter(s) of that level?

Please don't hold back, gimme both barrels. The information I have been able to glean has been helpful, but not specific enough. I entreat all the hard tech you can post!!

After looking at a website of the manufacturer of a competitive product, I'm even more skeptical that the AFCIs will prevent the fires that they are advertised to. The other company says that most of these fires are caused by "glowing connections" and that AFCIs do not detect "glowing connections". Their device is a temperature sensor type item that installs at switches and receptacles and opens the circuit when the temperature exceeds a preset point. They say that these "glowing connections" are caused by high resistance at a point of connections and have studies that show that the copper can glow red hot with as little as 0.3 amp load. They claim that these connections do not create an arc signature that the AFCI can detect. http://www.fire-fighter-products.com/index.html

Sparky,

Thanks for the reference. I noted it earlier when you provided it in another thread, printed it out and went over it. That too whets my appetite. ZLAN's site contains the most tech I've seen yet, but it is just to the side of clear for me. . .I may be thick, but it seems that the core of an AFCI is waveform recognition, either current, voltage or both. What I don't understand is the implementation of the hardware solution.

Brendan,

Maybe you have access to CH's waveform analysis? What are the identifiable characteristics of the class of faults the AFCI triggers on, and how does the (presumed) logic in the trigger discriminate between arc waveforms of varied "flavors". Am I even asking the right question here? Or is the answer to the question a trade secret that each manufacturer is keeping under wraps for the time being?

Respectfully,

Al

In recent posts someone installed AFCIs in a brand new residence, and had problems right off, and finally he said he had to remove them because of the nuisance(sp) tripping.
My question is this if in a 3 bedroom house, figure 5 receptacles each, lighting outlets etc and there is a problem which happens to intermittant how do you find the problem, short of replacing or rewire-nutting all the units, checking all switches for arching, and if switches can cause an AFCI to trip, I am not saying it can, but if its suppose to trip on arching, then it would make sense.
I know backwired receptacles are notorious for arching and then welding themselves back together again, or until the wire burns itself off completely.
Also does anyone know why Smokes are to be wired to the AFCIs ?

Don,

I'll try to answer your numerous questions, although I'm not holding out any hope at convincing you...
______________________________________
So you are telling us that you have developed an electronic device that has a 50 to 100 year operational life!! I don't believe that is even possible. What is the failure mode of an AFCI?
Don(resqcapt19)
________________________________________

I don't believe I ever claimed that we had developed an electronic device with a 100 year operational life. What I said was that homes do not stay new, and that the panels in them are there for longer than the wiring is safe. Of course all electronics in homes would last longer with proper surge protection as IEEE recommends. Until it becomes code however, I don't expect that the industry will embrace surge protection as it should.

When you ask about the failure mode, I assume that you are asking what happens if the electronics go bad from surges. If this happens, the breaker will not trip when the test button is pressed, and the instructions are to replace the unit. This is the same as GFI breakers and receptacles, and obviously the key is to educate consumers on proper electrical maintenance. We maintain furnaces, gutters, woodwork etc. It's not too much to expect homeowners to push the test button. All that needs to be done is to give them the instructions that come with the unit rather than throwing them away. If they don't maintain the product, at least you've done what you can to provide a safe electrical system.

There is a bit of contradiction in your asking for a self-testing device while asking for an inexpensive unit. The reality is that we know that we cannot make these too expensive, or they will not be used. A self-testing AFCI or GFCI may be down the road, but it is cost-prohibitive at this time. If contractors are unwilling to spend $30 for an AFCI, I can pretty much assure you that they won't spend $60.

__________________________________________
1) I'm not convinced that they will function to stop the fires that the manufacturers say they will. They say they only work effectively on parallel arc faults. I've always been taught that 90% of electrical failures are at connections. This would be a series fault, not a parallel fault.
___________________________________________
I'm not sure what we will have to do to convince you that these are viable devices that will stop fires. There are a great number of folks that have studied the technology and have determined that it will save lives. The code making panel consists of every facet of the industry, contractors, builders, academia, etc. There is only one member from NEMA (National Electrical Manufacturer's Association). The process is there to keep any one body from dictating what gets through to the code. It is an arduous process to say the least, and a very important one. Without an electrical code, there would be virtually no new safety products developed. Manufacturers would not spend the money to develop new products unless the benefits could be mandated in some form or fashion. UL has a standard, CPSC and others have come out in favor of the technology, obviously it has been proven to folks that have analyzed it, so at some point you'll need to have some trust in the experts in your industry that represent you. For more on arcing faults from UL, go to http://www.ul.com/about/otm/otmv5n3/labdata.html.

_________________________________________
2) The fire statistics used to get these items into the code are faulty.
_________________________________________
The statistics that were used were from CPSC, the insurance industry and fire authorities. I'm not sure where you believe that better statistics exist, but if you know of any, I'd be happy to take a look at them.

_________________________________________
3) If they aren't fail safe they shouldn't be permitted on the market.
__________________________________________
Then you would continue this argument by stating that we shouldn't have GFCI, smoke detectors or sprinkler systems? All are not fail-safe.

Another issue that I would like to address is the glowing contact issue, and your use of Fire-Fighter products as a reason to question AFCI technology. First, our AFCI breakers are not only UL1699, but they are also UL 1053 for equipment ground fault protection, and will prevent the glowing contact phenomena. I find it ironic that you support the position of an unknown company like Fire-Fighter, but will not support the position of the company that invented the circuit breaker, and has more Ph.D. electrical engineers designing true world-class products.


Someone mentioned doing it the European way. Europe does not have whole house AFCI, they have earth leakage. Actually, some of the IEC manufacturers are looking at AFCI technology because of the problem they have with burning appliances. Europe does not have appliances that are as fire-resistant as the U.S. due to the environmentalist influence keeping chemicals out of the designs. They have a real problem, and they may soon turn to our technology to solve it.

I will try to answer Al and Motor-T over the weekend or late next week. I have to go try to sell some Arc Fault Circuit Interrupters next week.