Attached are some pictures of an arc flash that occurred in a 480/277V panel, I'm trying to determine what can be done in the future to prevent this.

An employee attempted to reset one of the parking lot light circuit breakers when he heard shorting/arcing from within the panel. The breaker being reset was located on right side of the panel, about 2/3 of the way down (I believe it was located where the copper is missing on the right busbar). The bulk of the damage occurred on the left side of the panel where it destroyed the load side terminals of almost all of the breakers. Based image of the panel cover charring, the bulk of the damage was done on the left side, however there appears to be quite a bit of damage behind the busbars inside of the box. Eventually the main building feeder ground fault activated (set points 1200A, 18 cycles) ending the fireworks. Thankfully the employee walked away unharmed.

Additional info:
-Available fault current was in the 6-8kA range
-Upstream fuses had incorrectly been replaced with Bussman FRS-400 (should have been 225A)
-Circuit breaker that was being reset had a short somewhere along its run
-The single pole breakers have arc shoots above the tab where where they bolt onto the panel
-Several of the 1-pole/20A breakers have been used for daily switching of overhead lights (rated SWD)
-Panel is ~16 years old
-The breaker that was being reset appears to have little/no damage

My best guess is that trying to reset the breaker into a short created an arc inside which vented out, propagating into an arc flash. Is it possible that the daily switching of the lighting circuit breakers deposited a layer of conductive carbon over time which contributed to the arc flash? Would adding wall 277V toggle switches or adding a lighting contractor panel be a good solution? Is there any sort of preventative maintenance that would have helped? Any information would be greatly appreciated.

[ATTACH]315[/ATTACH] [ATTACH]318[/ATTACH]

Several issues:
1. Breakers are NOT an effective means of switching on a regular basis. They are rated for perhaps a few thousand cycles at most. Switches are rated for 10's if not hundreds of thousands of switching operations. Depending on load, you may or may not be better off with a lighting contactor. You could add a photo cell and save a little extra electricity for the extra cost, too.
2. Molded case circuit breakers must be inspected for damage before re-energizing after EVERY trip. They also need to be exercised every year, and tested every few years if they are serving a safety purposes. NEMA AB 4 which is referenced by EVERY molded case breaker manufacturer explains this in detail and has pictures and is fairly easy to read. Look on [url='http://www.nema.org's/']www.nema.org's[/url] web site. AB 4 is free.
3. It is a huge OSHA regulatory violation to reset a breaker without testing/inspecting to determine the cause of any fault. When this occurs, make it very, very clear to unqualified personnel to always call an electrician. The electrician must always, always inspect the equipment to determine the fault, and always, always inspect the breaker before resetting it (see #2).
4. NEVER close into a fault except with a breaker rated for that duty. Breakers are usually rated to survive OPENING on a bolted fault ONE time, and if it tripped due to a fault, you might have exceeded the life of the breaker. They are not rated to CLOSE onto a fault unless they are reclosers.
5. Not sure where the carbon you are referring to would originate from. Molded case breakers do vent but not carbon. There's no source of carbon internally except for the plastic housing.

I suggest you pay careful attention to the color photos in NEMA AB 4 and then go look at other breakers in the facility. Chances are you are going to be spending some money replacing several of them because they are already no longer in serviceable condition. What you are describing is exactly the reason why Article 200 in NFPA 70E exists. NFPA 70B talks in more detail about maintenance but since your specific issue is probably related to poor breaker maintenance, start with the manufacturer recommendations in the NEMA standard.

You might also want to pull UL 489 standard for molded case breakers and pay attention to what UL tests a breaker for. It is eye opening to realize what the minimum standard for a breaker is. Design to the standard, and you won't be sorry. If you can't afford the standard, you can probably google enough details to understand what the design requirements are.

Molded case breakers listed by UL as SWD have passed life testing and switching requirements similar to light switches for regular operation of fluorescent lighting, however the time frame 'regular' is not published.
Molded case breakers are rated for 1 fault at 100% (bolted) AIC level. They are rated for 50 operations at 6X. In between values are not defined

There operation of a molded case breaker vents 'arcing gases' into the panelboard enclosure. Many breaker designs have the venting towards the panelboard busbars.
In the late 80's UL revised its testing procedures to investigate mid-level fault interruption. Evidently some breaker designs did not vent gasses from medium level fault current 'fast enough' to prevent them from causing an arcing fault on the panelboard bus.

Thanks for the responses, I incorrectly thought that there would be some sort of carbon vented which might cause lower dielectric. It sounds like the ONLY culprit here was the circuit breaker that was reset into a short.

AB 4 looks like a good resource- that's one document that I have yet to stumble across.

I just skimmed though UL 489 and it appears as if breakers 100A and smaller go though 6,000 operations under load. Doing the math assuming one on-off cycle per day it works out to be ~16 years. A bit of a surprise to me, the requirements for SWD breakers only add additional temperature tests. I guess the big question now is whether a SWD breaker that has been operated for more than 6,000 operations that is in otherwise good condition is still safe to use for switching. I know of some panels that have been used for switching that were installed around 1980.

But they also go through 4,000 no load operations, so the life of the breaker really depends on the amount of current it makes and breaks.
Statically some breakers fail early and some last extremely long, but in general it is the make/break of current >100% that causes breakers to fail.

I'm not seeing the pictures...
Are these stab-in breakers?
If not, were the breakers checked to ensure they were tight to the bus?

I suspect this is the ultimate cause - the breaker had a loose connection and caused the initial arc flash when it was operated. Earlier operators were just lucky. Equipment maintenance is VERY important, even on these small panels.

Pictures would be good to see.

Forum must have a max size on images, original ones never stuck...
Pictures are now added to the first post.

I am familiar with carbon build up on contactors. It has been my experience that carbon build up is more prevalent on lower voltages (10 - 24 VDC) when you dont use gold and/or bifurcated contacts.

If the contacts are opened under load or if they are operating with high ambient temperatures, organic gases are generated from the synthetic resin used to bond the conductive materials of the contacts. The gases, arcing, and frictional energies released during the switching of the relay contacts cause the gases to oxidize to carbon. Over time, if these contacts are not inspected on a regular PM, this carbon can collect over the surface of the contacts to cause contact failure.

Also if the contact is opened under high current, you can get carbon, though the more serious problems stem from the contacts pitting, melting and welding.

While doing field service for many years, I have seen carbon build up on small ice-cube style relays, and to a lesser extent on contactors and motor starters. The repair included burnishing or dressing the contacts, or in worse cases replacing the contacts, the relay, the contactor or the motor starter.

This is a perfect example why "switching circuit breakers" is considered "Interacting� with electrical equipment and why NFPA 70E added the Information Note No. 1. Always wear your PPE when “interactitng� with electrical equipment. You never know when the electrical equipment can fail!

From NFPA 70E – 2012 Article 100 - Definition

Arc Flash Hazard.A dangerous condition associated with the possible release of energy caused by an electric arc.

Informational Note No. 1: An arc flash hazard may exist when energized electrical conductors or circuit parts are exposed or when they are within equipment in a guarded or enclosed condition, provided a person is interacting with the equipment in such a manner that could cause an electric arc. Under normal operating conditions, enclosed energized equipment that has been properly installed and maintained is not likely to pose an arc flash hazard.

Actually there is no NFPA70E proscription about general interaction. Your quote addresses a specific condition "interacting with equipment in a manner that could cause an electric arc".

You need to consider the risk of operating that breaker.
A simple on-off-on operation would not be expected to cause an arc for maintained breakers. However an on-off-rewire the circuit-on operation has introduced unknown conditions and therefore might cause an electric arc.

NFPA70E is about identifying and managing risk.
with the equipment in such a manner that could cause an electric arc. - See more at: http://arcflashforum.brainfiller.com/threads/2962/#post-13765
with the equipment in such a manner that could cause an electric arc. - See more at: http://arcflashforum.brainfiller.com/threads/2962/#post-13765
interacting with the equipment in such a manner that could cause an electric arc. - See more at: http://arcflashforum.brainfiller.com/threads/2962/#post-13765
interacting with the equipment in such a manner that could cause an electric arc. - See more at: http://arcflashforum.brainfiller.com/threads/2962/#post-13765

By your definition (which is not supported by 70E...further sentences in the note and others state that normal operation of equipment is not likely to cause an arc), anything short of just walking by is likely to cause an arc. By this definition all electrical equipment is inherently dangerous and should be banned carte blanche by the CPSC, OSHA, and others. Especially with switching equipment. By nature/design it arcs during making/breaking. It is designed to do so in a safe and controlled manner. The only times that it does not do so safely is if it is not properly maintained, is damaged/defective, or is used improperly such as removing or inserting mcc buckets while energized. The one small exception is inserting/removing draw out gear because this action simply can't be adequately controlled with any design I know of.

70E 2015 edition in development finally takes a much stronger stand with regard to this.

If operating a circuit breaker is not an interaction that could cause an electric arc, why is this task not Cat 0 in the task tables?

Are you suggesting that anything in the task tables makes sense? 2015 draft rectifies this. It looks like the 2nd round of inputs did not change it either.

What brand panel?

PaulEngr,

I'm having a hard time corelating your Saturday post with the one prior. If breakers are not an "effective" means of switching, then doing so should not be considered a normal operation and should be considered an interaction likely to cause an arc. Please explain further. Thanks.

Under the definition of Arc Flash Hazard, Informational Note No. 1 states that "An arc flash hazard may exist...provided a person is interacting with the equipment in such a manner that could cause an electric arc. Under normal operating conditions, enclosed energized equipment that has been properly installed and maintained is not likely to pose an arc flash hazard."

So this brings up two questions. First, what interactions could cause an electric arc, and to that specifically we've got to consider an arc flash hazard in particular. Just flipping a typical residential light switch on and off, to say nothing of plugging in cord-and-plug residential equipment, certainly causes an arc every time. However these arcs and controlled and harmless. That is the reason that all switching equipment is not categorically banned by the various consumer product safety organizations. The same thing happens in industrial settings except when performing certain activities. The question is, which activities? On this basis, the 70E Technical Committee has been less than helpful to date. In several responses ROP's in the past, they have repeatedly stated that "just walking by" doesn't constitute a likely case but the question is then what does?

Informational Note #2 in 130.7 states, "It is the collective experience of the Technical Committee on Electrical Safety in the Workplace that normal operation of enclosed electrical equipment, operating at 600 volts or less, that has been properly installed and maintained by qualified persons is not likey to expose the employee to an electrical hazard."

In this case, we've narrowed it down to "normal operation of enclosed electrical equipment" but have not defined "abnormal operation". Again, we're left with everyone having an opinion about exactly what this means. Claiming that all switching devices are inherently dangerous whenever they switch is clearly ridiculous. Some cases of switching such as numerous very public cases of racking breakers in and out of cells is clearly inherently dangerous and we have the battle scars to prove it. This is where the current (2012) edition of 70E is no help at all. Some entries in the tables have a "0" for arc flash PPE while having a higher rating for other activities but this is again not much help.

The new tables that are currently being proposed are listed in the draft comments and help quite a bit to elucidate what the 70E Technical Committe considers "normal operation". Following the latest table in SR45b from the NFPA 70E Committee's second draft, "Normal operation of a circuit breaker, switch, contactor, or starter" does not require arc flash PPE if it is properly installed, maintained, all doors are closed and secured, covers are in place and secured, and there is no evidence of impending failure. At the bottom of the table, it even states, "When this table indicates that arc flash PPE is not required, an arc flash is not likely to occur". Well, shazam! Finally, we have a definitive opinion. It's out there in a draft document but there it is, the 70E Technical Committee finally stepped up to the plate.

So moving on, we then have to consider each of the elements "properly installed, properly maintained, no evidence of an impending failure...".

NEMA AB-4 is the "go to" reference for molded case circuit breaker maintenance. I'm going to just suggest that right now it requires annual exercising (daily in this case), annual inspection, periodic testing, and inspection after every trip. I'm going to go out on a limb here and assume that all of this is being done, so "properly maintained" is checked off. So is having the doors on. I'll even go so far as to suggest that there might not have been any externally visible indication of an impending failure. So far, so good.

That leaves the design issue. With ANY switching device except solid state devices, the number of acceptable switching operations is given in the device data. You just have to look for it really hard sometimes because it may be in the manufacturer's data, or it might be buried in a UL/ANSI/NEMA standard. In the case of breakers it is common to see switching cycles of around 5K operations while contactors and switches are rated for anywhere between a low end of 30K operations to as high as 10 million, depending on the type of device. In this case switching a standard (not switching rated) breaker once a day with a 5 K design limit means it is good for 5,000 days, or almost 14 years. Since the equipment is generally designed to work for around 25-30 years for a typical design life, this is way too often. So we failed the "properly installed" criteria.

I would even go one step beyond what the 70E Technical Committee has done so far. Given especially that with molded case circuit breakers that OSHA requires that the fault must first be determined before re-energizing and that NEMA AB-4 requires a breaker visual inspection before re-energizing after a trip, I would consider an undiagnosed trip by itself to be considered "evidence of an impending failure" because if you don't find and fix the fault in the first place, it is almost a certainty that it will happen again. 70E doesn't go that far but I will.

Now, rectifying the differences in my posts. I take exception to this:



And also to this:



I agree in principle that there are NO absolutes in life. However, I don't accept two premises here with either one of these statements. The first is that you can't ever tell when equipment is going to fail. That goes directly against the idea of predictive maintenance. I agree that in most cases this is true with electrical equipment but you can't make a blanket statement about the predictive nature of failures with electrical equipment, just as the predictive maintenance folks can't give us a 100% guarantee that unexpected failures will never occur.

However the standard is not that an arc will NEVER occur. Car accidents occur all the time but we don't ban automobiles. Similarly once in a while you will hear about a breaker or even a simple light switch "exploding" but we don't ban those either. The standard is that an arc is not LIKELY to occur. I will even go so far as to suggest that based on actual industry standards (ISO 14121 Part 2), the term "likely" should mean that failure rates are around 1 in 100,000 per year or less. "Likely" in this case is a pretty small probability. The wording is perhaps not the best but the intent is pretty clear.

So any activity which has better than a very small chance (1 in 100,000 per year) of causing an uncontrolled arcing fault requires arc flash PPE. Everything else does not. And for the general non-electrically inclined public, this means that the equipment design requirement is that failures need to be kept below that threshold.

That is the standard that I consider when evaluating the idea of "interacting with equipment that is likely to cause an arc". In this case "likely" is a very poor choice of wording. It would be incorrect to suggest that an arc can NEVER occur with any energized equipment (exception: intrinsically safe devices), or that only equipment that meets this criteria should be acceptable. It would also be incorrect to suggest that as long as it isn't happening multiple times a year, that the potential for failure is acceptable from a safety point of view.

Hence the reason that I completely disagree with the absolutist point of view, just as I completely disagree with the idea that you can use any piece of electrical equipment any way you want (anything goes).

General Electric, A Series