Several arc fault breakers in panels

Status
Not open for further replies.
earshavewalls said:
I tried to post photos of two panels side by side that are 42 circuit panels with 36 spaces, all bunched together, no spacing, full of AFCI breakers.........but the file is too large to get on here.
Even if they separate the breakers as best as they can, it still would have several grouped together.
A Cutler Hammer representative is planning a trip out to take some measurements and other readings on what is going on. Hopefully this will help the development of these things so that we won't have this problem in the future. I believe it was definitely a design flaw here......but that's evolution.
(Design flaw on both the manufacture and the engineering of this system)
Click to expand...
afci.jpg
 
Pierre C Belarge said:
al hildenbrand said:
Given that the heat concentrates at the connections and internal breaker contacts, the end-of-life scenarios will most likely include the catastrophic failure of those connections / contacts with load current sustained series arcs, that the AFCI electronics, we are assured by UL, will NOT detect because the arc current is not exceeding the 50 to 70 amps trigger threshhold.

That's assuming the electronics are, in fact, still viable at the end of this heat degradation cycle.
Click to expand...
Al
Where did you get that information from?
Click to expand...
Pierre,

The information I base my opinion on is the result of my personal experience, not exclusively with AFCIs, but also with the existing installations that have failed or are failing, that I have repaired, replaced or examined.

Most recently, I have removed a 1978 Gould 100 A 2-pole service disconnect breaker in a 16-pole panel that was at the erratic end of being consumed by a burning lug and an internal contact burning.

The homeowner called me to his dwelling to troubleshoot a GFI receptacle that he had just installed that he was certain was messing up the house. I got there, and as he described the extent of the sporadic and spread out lack of power, I realized I wanted to see the panel rather than look at the GFI and its j-box. As we went through the house and down to the service center, the HO realized that the power was out again (partially, as it had been many times before).

When we got to the panel, before I could react, he reached up to the 100 A disconnect, shut it off, and turned it back on. The basement light behind us came on for the first time.

In this particular instance, the #2 Al SE conductor on the left pole of the disconnect breaker had been heating badly enough to have over two inches of #2 Al stranded conductor melted clean of insulation. The heat, created in the resistance of the failing terminal lug, had also been conducted into the breaker and was adding to the heat normally created in the internal breaker contacts. Over the many months that this heating and cooling, under varying load, had been happening, the contacts inside the breaker on its left side had degraded and the contact resistance was rising.

When I closely examined the breaker, the bakelite case was starting to discolor and crack like a drying out riverbed mud field. This cracking was local at the internal contacts of the breaker. Other bakelite damage was occurring by the terminal lug.

The resistance of a lug or terminal or contacts, once it starts to rise enough, will cause small incremental increases in the resistance of the splice. With each increase in resistance, the amount of heat created will increase. This can be a gradual process until the amount of heat created exceeds the heat that can be conducted, convected and/or radiated away. The heat concentrates, with attendant temperature spiking and catastrophic effects ensue.

Pierre, if this same process of heat degradation of contacts and splices occurs in a breaker with AFCI electronics, in my opinion, the electronics will stand an excellent chance of having been turned into green goo by the heat, long before the runaway temperature spikes start occuring.

But, if the electronics happen to survive to the end, they won't help because the heating and burning of the contacts, under load, are going to be only at load, not short circuit currents. The branch/feeder AFCI will ignore currents below 50 to 70 Amps.

If the occupant of a building with a breaker starting into the catastrophic end cycles of heat concentrations exceeding "black body radiation limits" is content to keep resetting the breaker, rather than call in a professional, then the conditions are present for property loss, injury or death.
 
al hildenbrand said:
In this particular instance, the #2 Al SE conductor on the left pole of the disconnect breaker had been heating badly enough to have over two inches of #2 Al stranded conductor melted clean of insulation.
Click to expand...
Isn't it amazing the way aluminum seems to disintegrate insulation? No puddles or ashes; it's just gone. :-?

The resistance of a lug or terminal or contacts, once it starts to rise enough, will cause small incremental increases in the resistance of the splice. With each increase in resistance, the amount of heat created will increase.
Click to expand...
True, especially with aluminum. This is at least partially similar to the problems with smaller-gauge aluminum. Because of both its relative softness and large coefficiency of expansion, cycling of load creates a heat-distort-cool-loosen-repeat snowball effect.
 
4x4inspector said:
afci.jpg
Click to expand...

Looking at the panel photo it appears that the breakers are connected to 4-wire multiwire circuits. Also, the breakers look to be two pole. Is the panel 3 phase? I'm confused on how the wiring and neutrals are interconnected on this MWBC, if that's what it's showing. Am I just having a brain fart?
 
Yes, these are three-phase 120/208, 42 circuit panels with about 36 spaces taken up by AFCI breakers. They are 2-pole breakers but have been properly wired and the proper number of neutrals have been used and distributed properly.
Everything is correctly installed per the manufacturer's installation instructions and to Code.
Go figure.......
 
If you follow the colors downward and ignore the whites, they go black-red-blue, etc. First breaker: black and red, second: blue and black, third: red and blue, etc. Each breaker gets one load-side white, and the breaker line-side pigtails are bundled.

I am curious about the premises; hotel?
 
Last edited:
Larry,

Back at the opening post the description of the premises is:
earshavewalls said:
We have a condition at a new Motel with arc fault breakers and heat. The electrical engineer is insisting on using arc fault breakers for all guest rooms, . . .
Click to expand...
 
Well, the latest from Cutler Hammer is that they claim that this is not anything to be concerned about. They will probably not be sending a representative to run any tests on the systems.
This bugs me a bit. If they DID send someone, and if that someone DID run a series of tests on this arrangement, and if those tests DID show that there is an issue with performance and/or safety, they would have to make these results public.
We're talking about a LOT of $$$$ here. Wouldn't it be a mess if, just before they were to release the 2008 NEC with ALL circuits needing arc fault protection, they discovered that you cannot put that many of these breakers together in a panel?????
Talk about the brakes going on!
Well, now I have an email in to U.L. I want to see if THEY might take issue with this situation. Probably not, but it doesn't hurt to check.
I'll keep y'all posted.
 
problem? what problem?

problem? what problem?

earshavewalls said:
Well, the latest from Cutler Hammer is that they claim that this is not anything to be concerned about. They will probably not be sending a representative to run any tests on the systems.
This bugs me a bit. If they DID send someone, and if that someone DID run a series of tests on this arrangement, and if those tests DID show that there is an issue with performance and/or safety, they would have to make these results public.
We're talking about a LOT of $$$$ here. Wouldn't it be a mess if, just before they were to release the 2008 NEC with ALL circuits needing arc fault protection, they discovered that you cannot put that many of these breakers together in a panel?????
Talk about the brakes going on!
Well, now I have an email in to U.L. I want to see if THEY might take issue with this situation. Probably not, but it doesn't hurt to check.
I'll keep y'all posted.
Click to expand...

:) sounds to me that the legal department got wind of the issue....

Situation:
"if those tests DID show that there is an issue with performance and/or safety",... "We're talking about a LOT of $$$$ here"....

result: manufacturer stock answer
"that this is not anything to be concerned about"

translation:
legal dept: "stay the hell away from it"

:)
"
 
IMHO it's time to call in every inspectional authority that will come and take a look. CH won't help, so blow the whistle. The word needs to get out. Call the newspapers and construction magazines. Has any state or local inspector looked at this problem? There are people who frequent this site that may be able to put a bug in someones ear. It just isn't right that this kind of install has to be tolerated.:mad:
 
yanici said:
It might pay to call the manufacturer of your panel and see what they have to say. They should be informed about field problems. It may be necessary for you to install another panel or panels so that you can distribute the load and get the heat down. Ventilate the electric room also.[Dennis types faster]
Click to expand...
I hereby claim initial disclosure (patent pending) for a special muffin fan to be installed in the spare 2 1/2" knockout on the bottom of the panel, along with two screened knockouts covers on the top.

The fan will be operated off a special breaker with a thermostat that plugs into the panel between two AFCI breakers. The fan kit will include special plastic baffles to direct air flow for optimum cooling.

It will be furnished both as part of the panel (installed by OEMs) and as a retrofit kit (only $179.95 plus $19.95 shipping).

I will submit a request to the NFPA proposing a restriction on all panels with more than 3 AFCI breakers to a 60 degree C rating on all breaker terminals unless the cooling kit is installed.

It will of course be UL listed.
 
Well, here's the latest now.
I have an email into U.L. and am awaiting a response.
The project engineer has admitted that he was mistaken in his interpretation of the code as to specifying AFCI breakers for the guest rooms. He is now in negotiations with Cutler Hammer trying to work out a 'deal' for them to credit the AFCI's and provide standard breakers.
So, I guess the problem may be aproaching solution on THIS project, but what about the next one? and the next?
I think it's time we started making a little noise before the 2008 NEC adoption brings this situation into an everyday occurance.
 
Ears, would it be possible to get an inspector to look at it and write up an official opinion before the dirt gets buried under the rug?
 
Status
Not open for further replies.
Top