Combination AFCI/GFCI Circuit Breakers?

The FIRE-GUARD=99 Arc Fault Circuit Interrupter (AFCI) is a residential circuit breaker with an integrated processor which recognizes the unique current and/or voltage signatures associated with arcing faults, and acts to interrupt the circuit to reduce the likelihood of an electrical fire. With the Cutler-Hammer FIRE-GUARD AFCI, protection from arcing faults is combined with conventional thermal and magnetic overloads as found in standard residential circuit breakers protecting wiring from excessive heat or damage due to overloading or short circuits. FIRE-GUARD AFCI can also be equipped with 5 mA ground fault protection to protect from personal shock hazards. Now, there is a residential circuit breaker that provides protection from arcing faults, conductor damage due to thermal overloads and short circuits, as well as 5 mA ground fault protection in one integrated design.

from:

When will someone develop the toddler with a paper clip detector circuit interrupter.

[8~{} Uncle Monster

A coworker (engineer) installed the combo units (5ma trip) in his addition. One was tripping and we did some investigations to figure out why. Turns out there was a staple that had nicked the insulation of the neutral conductor and was shorting it to the grounding conductor. We duplicated this fault in my lab and tested a regular breaker, a GFCI, an AFCI, and a combo AFCI/GFCI (5ma). As expected, the regular breaker did not trip. Neither did the AFCI. The GFCI tripped as expected (half cycle or so). The combo tripped, but took much longer. I don't have the data in front of me, but it was something like 5 to 10 cycles. Now, is this a big difference? I don't know from a human survival standpoint. However, if my wife throws a toaster in my bathtub, I want the traditional GFCI that trips in half a cycle, NOT the slower AFCI/GFCI combo unit.

I am still not convinced of the usefulness of the new breakers and question the required investment verses the limited events that may be detected.

Just my personal opinion.

Charles Perry P.E.

It is called the tampre resistand receptacle and that is in the 2008 code too.

Tee hee!

As the line goes, "Look for the Money."

Conventional CBs in the popular sizes (15/20) are just dirt cheap. They also last "forever."

But an AFCI or GFCI has a bunch of "accident waiting to happen" electronics just waiting for a surge or excess humidity. Even if we "assume" the electronics are good for 30 years there's a significant replacement market out there.

Oh my God! I'll bet the things will be like the so called child proof safety caps on pill bottles. Impossible for older folks to use.

[8~{} Uncle Monster

The following information was interpreted using information from the Lawerence Livermore Labs. Charles Dalziel invented the GFCI in 1961. It is my understanding that he worked at these labs or nearby.

Experience has shown 5 mA can cause uncomfortable shock, and in cases of children or where persons are on an operating table, a current of 5 mA may cause serious discomfort and complications. By Figure 1 the UL Standard 943 performance curve requires the GFCI to open the circuit in 7 seconds at 4 mA, in 5 seconds at 5 mA, and in 0.04 seconds at 110 mA fault current. . However, the performance curve for the LM 1851 IC has .4 seconds at 4 mA, .2 seconds at 5 mA, and .02 seconds at 110 mA. Ventricular fibrillation that alters the heart's normal rhythmic pumping action can be initiated by a current flow of 75 mA or greater for 5 seconds (5-s) or more through the chest cavity.

To calculate the probability of ventricular fibrillation perform the following: To determine the 5-s current flow (in mA) necessary to cause a 0.5% probability of ventricular-fibrillation, multiply a person's weight (in lb) by 0.49. To determine the 5-s current flow (in mA) necessary to cause a 99.5% probability of ventricular fibrillation, multiply a person's weight (in lb) by 1.47. (Reference: Lawrence Livermore National Laboratory, University of California.)

The 99.5% probability for a 6 pound infant is 1.47 x 6 equals 8.82 mA for 5 seconds. This is within the safe limits of the operational curve of a GFCI by a slim margin. However, the .5% probability is 0.49 x 6 =

2.94 mA. for 5 seconds. The UL and National Semiconductor operational curves do not require a GFCI to operate for less than 4 mA of fault current. Based on these formulae, operational curves, and calculations the operation of a GFCI can incur a possible 0.5% probability of electrocution by ventricular fibrillation for infants that weigh less than 8.2 pounds (4mA / .49). It should be remembered that for circuits protected by a GFCI circuit breaker there will be leakage current and this leakage current increases as the circuit length increases. If the leakage current is above 2.06 mA then the fault current required to trip the GFCI circuit breaker would only be 2.94 mA for an average GFCI. These conclusions are only possibilities. The trip time verses fault current curves allow for many other conclusions.

Poster's note: This calculation is not verifiable since we cannot experimnet with infants. It is very possible that the calculation is too liberal and in fact, 4 ma may cause ventricular fibrillation in infants.

Ref:

The 2008 NEC rule to require tamper proof receptacles protects children, but we need GFCI's to protect adults. Can old people use these tamper proof receptacles? Good question. I guess we will find out.

That makes you afeared dunnit? :-]

We didn't need a primer, idiot.

Some of my older friends have problems understanding those newfangled GFCI outlets. I've had to show them how to reset the darn things and explain how they work. It's very hard for many old folks to see or push that little reset button. If the design of electrical outlets changes, I'm afraid many people, not just old people, blond women perhaps, will have a lot of trouble with the new devices.

[8~{} Uncle Monster

I think the AFCI combined with the GFCI combined with the tamper proof receptacles is going to bring high tech to old people's receptacles and maybe become like operating a DVD player.

The real problem with rushing all of this into the code is this is an evolving technology that still can't detect a series arc (the one from a loose plug, cracked wire and all of those "aluminium" problems). They already have millions of obsolete AFCIs out there mandated in the

1999 code and enforced in the 2002 code that can't detect the bad power cord they scared us with when these were first proposed. The original AFCI only detects shorts in the wall. Finding a short in a power cord is what the "combination" unit is supposed to do. None can find a loose connection or partially broken wire. Then you have a bunch of Square D units that don't work at all and have been recalled but there is no tracking on this recall so most still remain in the customer's home. The technology is so proprietary that the industry can't agree on what a tester should inject in the line to see if the product works. An arcing short that trips a Square D might not be seen by a Cutler Hammer and vice versa. It is clear this is a product that is not ready for the public yet but has been jammed down our throat by NFPA and Cutler Hammer, the inventor and who proposed in in 1997. It is being tested in our homes, by force of law and the customer is paying for it.

The problem is there are already "insertion force" problems with the ridiculously expensive tamper proof receptacles out there now. Imagine how bad they will be when we get down to the ones builders use to hit a price point.

If what you say is true, I think we may have a problem adopting the

2008 NEC. The AFCI requirement is not going to be accepted without further investigation. We need additional information.

I think they are mostly 30 ma.

Interesting. Why would the AFCI/GFCI was slower. A new topic of investigation for EPRI.

As you probably know, a GFCI includes a function that detects downstream N-G shorts with no load. When an AFCI has a load there will be a voltage drop that can produce the N-G fault current to trip the AFCI.

[Are you suggesting the code should add 'wife malice' protection? An exciting new field.]

IMHO it would have been appropriate for the extension of AFCIs beyond the previous bedrooms to be based on the record of the AFCIs already installed in bedrooms. It wasn't.

The existing AFCIs that detect parallel arcs should detect arcs in power cords from wire-to-wire. H-N arcs are supposed to produce enough current to trip the AFCI. And if there is a ground wire, I think the logic was that the arc would rapidly involve a ground fault so ground fault detection was included.

Are there any commercial "combination" (series/parallel) AFCIs on the market yet?

I agree it was stupid for the 2008 NEC to vastly expand the use of "combination" AFCIs, required only in bedrooms in 2008 by the 2002 NEC, when the devices didn't exist and did not have field-use experience.

Will be interesting what AHJs do with AFCIs in adopting the 2008 NEC.

Keep in mind that we tested using a single make/model of each type. We did test multiple units of the same make/model of the combo unit just to make sure we were seeing what we thought we were seeing. We also done some testing of tripping of AFCI breakers due to EMI/EMC problems. Just think, your neighbor installs a new heat pump with an adjustable speed drive and your breakers starts tripping. Could happen. I know they say the filtering is great, etc but we make a living cleaning up after things that don't exactly work as predicted.

Now if I could just get several utilities to fund some AFCI testing work ;-)

Hahaha. I have to be careful what I say right now. I am recovering from ACL reconstruction (replacement actually) surgery that happend one week ago today. I am VERY dependant on my wonderful spouse (she might be reading over my shoulder) during this recovery!

Agreed. It was based on undue influence by those with GREED. IMO of course.

Charles

"Combination" refers to btranch circuit wiring plus line cord protection. The first swing at this required so much fault current (60-70a) that you could not reliably get that much in a lamp cord.

I did hear several panel members discuss the ground fault trip rating for AFCI's. They said that the first AFCI's were made to trip at 5 ma ground fault current, but his "competed with" the GFCI's so they run the ground fault trip up to 50 ma. Now what they meant is a mystery. Did they mean the market was competing or did they mean the AFCI wasn't working right when a GFCI receptacle was installed on the AFCI circuit?