Arc fault breakers – on every circuit?
Well, on 12/23, we had quite a scare. We had a receptacle start to smoke in our bedroom. Luckily my son noticed the smoke and alerted everyone. I immediately shut the breaker off while calling 911. Electrician came today and thinks the problem was: 1) loose neutral wire combined with 2) high draw in 15 amp circuit due to blow dryer and recently purchased electric space heater (1500 watts). Heater was in bathroom. 2 bedrooms and bathroom share the same 15 amp circuit. Stupid me. I knew all about the 80% rule on circuits but didn’t think to check the wattage on the space heater. Plan now is: 1) Wire the bathroom on its own (this was planned as part of current bathroom renovation but has moved up in priority). 2) Install arc fault breakers to cover all of the bedrooms. QUESTION – even though code doesn’t require it, is there any reason why I should NOT replace all breakers with arc fault breakers? Cost is not a concern. Also, any idea why the standard circuit breaker didn’t trip due to the high load from the blow dryer and space heater? Is it possible for a copper wire connection (hooked around the screw) to work its way loose over time? Thanks and hope everyone has/had a safe and Merry Christmas. Life has once again shown me what is really important in life…a safe family.
Replies
The breaker did not trip because you can make a lot of smoke and not overload the breaker.
I don't know where you are located, but where I live, I think new construction would have all the bedrooms on arc fault breakers.
Find a reputable local elecrician you are comfortable with, and they will help you out.
H
Arc fault detectors have a fairly high failure rate and can be tripped by certain "normal" conditions, so they probably shouldn't be used for all circuits (especially, eg, refrigerator, freezer, furnace). But not a bad idea for bedroom circuits.
emaxxman--
Good to hear that you caught the arcing receptacle. Although it may have been a bad connection on a terminal screw, I'd bet that it was a spring-type, poke-in backwire connection. They're notorious for failing, and of the many I've seen that've failed, it's more often the neutral that goes bad.
ANSWER 1: You certainly could put all 15- and 20-amp branch circuits on AFCI breakers. There are some constraints (shared neutral circuits is a big one).
You may find when retrofitting circuits w. AFCIs that there are latent problems like crossed neutrals and grounds, inadvertantly shared neutrals, high-impedance ground faults, and so on.
There's a proposal to the NEC (for 2011) that if adopted in June 2010, will require almost all outlets (lighting, receptacle, and other) on residential 15 & 20 amp branch circuits to be AFCI-protected (exceptions: receptacles that are GFCI-protected, like kitchen counter and bath receptacles).
ANSWER 2: A standard breaker is intended to trip on overload, or if there's a short circuit. There may have been an overload in your case, but a breaker doesn't trip instantly on overload at the rated ampacity of the breaker. A common 15-amp GE breaker will allow 75 amps to pass for 10 seconds before tripping. See this time-current curve: http://www.geindustrial.com/publibrary/checkout/Time%20Current%20Curves%7CGES-6119D%7Cgeneric. Big suprise to most people...
So, the bad connection at the receptacle generated a lot of heat at the loose connection, but because it was probably an intermittant arcing fault, not enough at the breaker to trip it.
Q3: The loose connection at the receptacle was probably a spring-type poke-in backwire termination. They ought to be banned for recpetacles, IMHO. In contrast, a back wire connection that uses a clamp-type terminal (like on GFCI receptacles) is very good; if done right it's as good as wrapping a wire around a screw terminal.
It's interesting, people think that electrical wiring ought to last the life of a house. If the installation is done right, and by that I mean "best practices" and not Code-minimum techniques like spring type backwiring, the wiring ought to not need any maintenance for a long time. But it's rare that best practices were used (at least post WWII) because of cost pressures and work done by unqualified people. Most work I've seen that was done by professional electricians, from the early 1900s (1905) to about WWII, is uniformly in good shape.
About where to go from here, short of opening every receptacle and switch box and checking the connections... there is a way you can sleep a little better. Do a voltage drop test of every circuit. Ideal Industries has a voltage drop tester (the Sure Test) as does Tasco Industries (the Inspector III). These are fantastic instruments, they're like a cardiac treadmill test for your wiring.
The tester puts a heavy current on the circuit for a very short time and measures the voltage, and calculats percentage voltage drop. Every connection causes some small bit of voltage drop, and a weak connection causes a lot more voltage drop.
A voltage drop tester, properly used, with correct intepretation of the results, will find and help locate failing (and close to failing) connections. And it'll do that quickly and efficiently. I couldn't work without the testers I have, whether it's troubleshooting, assessing the condition of wiring before a remodel, or checking my own work.
Best of luck on the remodel, and Merry Christmas...
Cliff
Hi Cliff - lots of great info. Thanks.
The connections were definitely made wrapped clockwise around the terminal screws. I replaced all of the receptacles in the house when I moved in 10 years ago. The old ones were all crusted over with paint and some did not hold plugs securely.
Based on tips on FHB, I went with with wrapping the wire around the screw for the same reasons you mentioned above.
Is it possible for a tight/solid connection, under high load, slowly melt away the plastic on the receptacle and then "create" a loose connection?
The electrician that came to the house said no but he also has never installed any AFCI's. I'm sure he's right but he's lack of experience with AFCI's gives me some reason to doubt him.
As others have stated, the better option is to spend the extra money for "spec grade" "back wired" outlets and switches. They're generally about 4x the cost of your basic outlet, but just the time saved wiring is worth the cost, plus you have a lot more confidence in the connection when you're done.
Generally a connection shouldn't "melt" -- the connection itself is all metal -- but repeated heating/cooling cycles can cause a connection to work loose, especially if it wasn't quite tight enough to begin with.
A little odd that your sparky would have no experience with AFCIs, since code has required them in some circuits for 2-3 years.
I found it strange too that he didn't have experience with them. The fire marshall, upon exit from the house, strongly I suggested get an electrician to check out the house (which I strongly agreed with). He gave me the name of the electrician and said that he heard he was good.
The electrician stated that in our area (NW NJ) started to require them in October. Maybe that's why he had never installed one.
I'll ask him for an estimate for the work I want done but will also get other estimates. A few GC's in the area have provided recommendations as well.
Back to the question on the AFCI's on all circuits...I'll look into putting them into all circuits except for those with appliances that may cause repeated trips (ie fridge, furnace, a/c). Sounds like those circuits would have similar issues with GFCI breakers.
Cliff wrote: "About where to go from here, short of opening every receptacle and switch box and checking the connections... there is a way you can sleep a little better. Do a voltage drop test of every circuit. Ideal Industries has a voltage drop tester (the Sure Test) as does Tasco Industries (the Inspector III). These are fantastic instruments, they're like a cardiac treadmill test for your wiring."
--------
I looked these tools up and they are pretty pricey. More importantly, they seem geared for the professional electrician and not a trained amatuer homeowner like myself. I would wonder if I could read the readings correctly and understand what they meant.
Are there any other options that would allow me to determine if there is an issue or not without opening all of the suspect boxes (which I have no hesitation in doing)?
Also, how would you test light fixtures?
"Are there any other options that would allow me to determine if there is an issue or not without opening all of the suspect boxes (which I have no hesitation in doing)?"
===========
You can certainly test voltage drop yourself with just a DVM and a known load. What you want is a known resistive load that draws about 10 A consistently. Plug it into a receptacle you want to test, measure the voltage at the service panel, then measure the voltage at the receptacle. The difference is the voltage drop.
So what is normal voltage drop? It can be calculated using Ohm's law (E=IR). In this case, the expected drop is the 10A load current times the resistance in the wire. This site
http://www.cirris.com/testing/resistance/wire.html
has a calculator. Assuming you have 20 feet of wire, the resistance is .032 ohms, so the expected voltage drop is 0.32 volts.
Put a simpler way, if you see a voltage drop of more than 1V to 2V for 100' or less of 12 gauge wire with 10A load, something is wrong.
The receptacle meltdown could have started at one of several points--the contact point(s) between the plug blade(s) and the receptacle, inside the receptacle, or the contact point between the wire and the receptacle terminal. If the receptacle wasn't destroyed, it might be possible to determine the point of failure.
A connection of a wire to a screw terminal usually doesn't fail. That depends, of course, on whether connection was made up right during installation. Even something apparently minor like a nick in the conductor (due to a bad stripping technique or tool) can eventually lead to a wire break and an arcing fault.
Wow, I had no idea that there were areas 8 years behind the NEC (AFCI breakers were first required in the NEC in the 1999 version, effective 1/1/2002, for bedroom receptacles only.)
There's going to be a whole lot of consternation in your area as electricians start installing AFCIs in new construction. There is a learning curve; sloppy wiring like inadvertently shared neutrals, or grounding conductor/neutral conductor contact, will trip an AFCI.
And retrofitting AFCIs, as you're considering? It'll be a real can of worms.
Good luck,
Cliff
Yep, it's common for an
Yep, it's common for an outlet to fail due to poor connections at the plug/socket junction causing overheating of the entire outlet, including the screw connections. Another reason why it's a good idea to not buy the 39 cent devices.
Wow, I had no idea that there were areas 8 years behind the NEC (AFCI breakers were first required in the NEC in the 1999 version, effective 1/1/2002, for bedroom receptacles only.)
From from start there where a large number of locals that questioned the effectiveness of the AFCI and while adopting current versions of the NEC and/or IRC they excepted that requirement.
Don't know how many are still doing it.
What is interesting is that until about 2 or 3 years ago the AFCI's only detected parallel faults. Saw a site that was used to promote AFCI's and showed a failure similar to the one that this person had.
The receptacle/box/wiring caught on fire before the AFCI tripped.
It was "hopped" that after the AFCI tripped due to ground fault, parallel fault or overload from the now insulationless wires that the fire would go out before spreading.
I have not seen anything on the newer combo AFCI which are suppose to also trip on series arcing.
The SureTest is about $250, that's not cheap, but what's it worth in your time to open every box?
Also, the SureTest can provide an indication of a bad connection in a junction box that's not apparent (in an attic, basement/crawlspace, or a buried splice). It'll tell you the impedance separately of the hot, neutral, and groudning conductors...so you can tell if the voltage drop is attributable to one more than the others. That's very useful info to have when tracking down a bad connection.
In response to your question--for fixtures, you remove the fixture; that usually means that you re-make the splice.
I have no connection to Ideal or Tasco, BTW.
When you consider the cost of an electrical fire, either buying a tester or finding an electrician with one is pretty cheap. As to interpreting the results, the tester instructions give you a basic handle on that. Of course, you get more out of it when you've used it a lot and seen how the results translate into actual conditions.
Best,
Cliff
Also, any idea why the standard circuit breaker didn't trip due to the high load from the blow dryer and space heater?
=======================
As others have indicated, a conventional 15A circuit breaker trips only on sustained moderate overcurrent (e.g. 100A for minutes) or huge overcurrent (e.g. 500A for 10 seconds).
That's why the AFCI breakers were introduced and mandated by NEC. The AFCI is a very reliable device that will consistently detect shorter duration and magnitude overcurrents and intermittent overcurrents. These devices contain a computer that does some fairly sophisticated analysis to differentiate normal surges (e.g. starting a motor) from hazardous surges caused by loose wires, etc.
In some cases, they do tend to nuisance trip, so it's probably best not to put them on circuits like the refrigerator circuit.
Regarding the cause of loose wiring, the most common cause in my experience is connections that were not torqued to the proper spec when the receptacle was installed. I've never seen anyone (other than me :-) tighten the screws with a torque screwdriver to ensure proper torque. However, if the screws are significantly under- or over-torqued, they do tend to loosen over time.
The connections on heavily loaded circuits are slightly more likely to loosen, due to the thermal expansion and contraction.
Inexpensive or spec-grade receptacles are often blamed, but I've seen no evidence to support this claim. All receptacles should meet UL requirements. The primary benefit of pricier receptacles is that they are more durable when used in applications where plugs will be installed and removed a LOT. I've always used the lower cost receptables for residential wiring and have never had or seen a problem with them for the last 4 decades or so.
As others have indicated, a conventional 15A circuit breaker trips only on sustained moderate overcurrent (e.g. 100A for minutes) or huge overcurrent (e.g. 500A for 10 seconds).
You got a reference for that?
Well, you could simply read the spec sheet for any circuit breaker. Look for the time/current curve - here is one for reference
http://www.eaton.com/ecm/idcplg?IdcService=GET_FILE&dID=78934
That shows the breaker tripping in about a minute at 2x breaker rating, which is about what I remember. About a second at 10x breaker rating.
For a 15A breaker, this chart shows that this particular brand and type of breaker will trip in 1 minute or less with a sustained overcurrent of 300% of nominal, or 45A, at 25C (nominally room temperature). It will trip in 1 second or less at an overcurrent around 1300% or 195A.
Different brands and models of breakers will have somewhat different specifications, but these are in the range. The point is that sustained moderate overcurrent - such as 3 times the rated current for a minute - is required to trip the breaker, and that significant overcurrent, 10 times or more, is required to trip the breaker within a second.
Most conventional breakers use a bimetallic strip or similar technology to implement the trip mechanism. Thus they integrate the current until a threshold is reached. The time required to reach the threshold is determined by a combination of ambient temperature, current, and time.
An intermittent or lose connection can thus generate overcurrent of 3 times the breaker rating, for example, and if the current drops to the normal range 10% of the time, the breaker may not trip.
One other interesting observation from the chart is that some samples of this breaker may trip if they draw rated current for an hour. In other words, the minimum specified on the chart is that the breaker will trip if the rated current is supplied for hours. Another good reason to design with a little margin.
So show me where it says a 15A breaker will pass 100A "for minutes".
"I've always used the lower
"I've always used the lower cost receptables for residential wiring and have never had or seen a problem with them for the last 4 decades or so."
You is one lucky dude! Did you notice on the OP's photo that the outlet failure was due to internal soft prong contacts, not on the external attachment? One of the big advantages of spec grade is the prong contact pressure, which prevents the type faiure seen by the OP.
I dont have a single cheap non-spec grade outlet in own house - not even in the junk piles, wont even pick themup whenin a 'free pile' at garage sales.
"Did you notice on the OP's
"Did you notice on the OP's photo that the outlet failure was due to internal soft prong contacts, not on the external attachment? One of the big advantages of spec grade is the prong contact pressure, which prevents the type faiure seen by the OP. "
It does appear from the photo that the "hot spot" was the blade/socket contact point. However, the OP also indicated he was inadvertently overloading the circuit. I can't conclude by looking at that photo that socket contact pressure contributed to the problem.
I do agree with you that the upgraded receptacles have higher contact pressure. The lower cost sockets (which some people refer to as "spec grade", but I don't think that is what you meant when you used that term) that I use are plenty "stiff" when new, to the point that it requires a good bit of pressure to insert the plug the first time. In an application where plugs will be installed and removed a lot, I would opt for the better receptacles. However, I personally think the lower cost receptacles are fine for most home applications, where plugs may be installed 20 times or less per receptacle over the lifetime. Heck, I have receptacles in my house that have never had anything other than a tester plugged into them :-)
The lower cost sockets (which some people refer to as "spec grade"
The lowest grade is Residential. It is the only grade available that has backstab connects available. Also, AFAIK they are only available in 15 amp version. At least in the past there have been versions with only backstab and no screw terminals.
The next grade is Spec or Specification. They are available in 15 and 20 amp versions with or without Back Wire (Clamp) versions.
The next grades up are Commercial and Industrial.
This are industry standard terms and recognized by UL which has different test for each of those grades.
Thanks for the info, Bill. Unfortunately, this seems to be one of the areas where local vernacular is sometimes preferred over code terminology.
Stopped at HD last night to compare. The Leviton receptacles at HD (at least in this area) are labelled "standard", "Preferred" and "pro", apparently corresponding to residential, spec, and commercial grades per NEC.
I don't use the back stab contacts on the residential grade receptacles - I have heard of problems with them and am not comfortable with them. I always use the screw terminals.
I compared the "standard" and "preferred" receptacles and found the following differences:
1. cover plate screw hole is threaded plastic on std, embedded nut on pref.
2. pref has a spring contact on one mounting screw to make better ground contact with a metal cover plate
3. mounting ears are attached to the plastic body of the receptable on std, but part of a continuous metal strap on the pref.
4. Screws and metal thickness for the exposed contacts appeared to be the same for the hot and neutral. Ground terminal was incorporated into the metal strap on the pref, attached to the ground terminal on the std.
5. both grades are available as tamper resistant receptacles (they incorporate a shutter that blocks the tab holes when a plug is not installed)
6. I was not able to evaluate strength or thickness of the blade contact (without disassembling the receptable).
7. The pref. comes in a box with instructions, the std is sold in bulk form.
8. The pref costs about 4 times the cost of the std.
So it appears that the "preferred" (presumably spec) grade sockets would be more durable in situations where they are subjected to abuse or very frequent use. I still think that for typical light residential use, the standard or residential grade would be fine, provided the screw terminals are used. While I have not been involved in committees developing that particular code section, it seems nearly certain NEC would not allow use of these sockets if they were unsafe.
Having said that, you guys have convinced me to upgrade some of my receptacles, such as the receptacle where I plug in my computer, since it is plugged and unplugged at least once a day.
I have never opened then up to compare them, but my impression is that the wiping contacts on the spec are stronger than the residential grade.
Looking very closely the wipers in the spec do look wider than in the residential grade. I know that one of the UL test is for repeated insertion/withdraws and the required number is higher for spec than residential.
Also the spec grade uses nylon for the face and case which is less likely to crack than the rigid plastic use on the cheapies.
BTW, here Lowes as the residential in both bulk and packaged for 15 cents more. Don't know why.
I agree with you, residential are fine in places in living rooms and most bedroom.
But I like spec grades in kitchens, baths, and shop areas. Also probably a good idea for garage and outdoor where you often have extension cords which get pulled on.
BTW, all of the spec grade receptacle that I have in the spare/junk box had Spec or Spec Grade stamped on the mounting shaft next to the ears.
And AFAIK all GFCI receptacles are spec grade or higher.
duh, still have not figured out the edit on the new format....
Arc-Fault breakers
I am an electrician with almost 30 years of experience. Arc Fault breakers are notorious for nucience tripping. Tread mills and vacuum cleaners are the main problems. We frequently get calls about these appliances tripping AF breakers. I have never found a problem with the appliance. I have been told by breaker manufacturers that these motors generate a radio frequency that the internal parts of AF breakers dont like. The breaker trips when in use. You can try a surge protector. Sometimesthis will help. Most breaker manufacturers make a device that acts as a filter to "trap" the noise before it reaches the breaker. They want you to try several other remedies before they send this device. Other instances I have seen that trip AF breakers are, Grounded neutrals, neutrals from different ckts being tied together, Blown light bulbs, pinched wires, staples being too tight, and AF breakers being installed ahead of ground fault devices. The strangest call I have had was when a paper shredder was being used it tripped the AF breaker on another ckt. The only thing that I could find in common was The 2 ckts were comming thru the floor in the same hole they were stacked and stapled up the wall to the second floor. I rerouted the wires so that they were separated. The staples were not too tight. This leads me to believe that a voltage or noise was being induced on the AF ckt causing it to trip. I have probably had close to 100 calls on Tripping AF breakers in approx to 10 yrs and have yet to find and arcing fault to be the problem. I believe the public has been mandated to use a product that has not been properly tested.
breakerman1
You've replied to a couple yr old thread, but...............
There was a much more recent question involving the same trip problem.
I'll try to find it, but if you have the time you might try the search up above-it might generate a lead that shows more recent threads first.
Thanks for the detailed answer.
EDIT: Here you go- http://forums.finehomebuilding.com/breaktime/general-discussion/afci-circuit-issue