Where, exactly, is Arc Flash Required

wbdvt

Senior Member
Location
Rutland, VT, USA
Occupation
Electrical Engineer, PE
This is what I wrote to this question on the arc flash forum:

First all, outside of a specific OSHA requirement for utilities to perform an arc flash study, there is no legal requirement. That being said, studies are performed for the following reasons:

1. Arc flash is a known hazard and OSHA requires employers to protect from known hazards (General Duty Clause)
2. Employers are required by OSHA to provide the proper PPE to protect them from hazards in the workplace (1910.132)
3. Some insurance companies will require an arc flash study as a condition of insurance
4. Companies care about their employees and want to provide a safe workplace

Since arc flash is a known hazard and there are means to protect employees from the hazard, if not done and there is an accident, the employer leaves themselves exposed to lawsuits.

Outside of determining the incident energy, a good electrical study will examine short circuit, equipment duty ratings and coordination to enable the employer to have a safe electrical system.
 

paulengr

Senior Member
Outside of power limited, intrinsically safe systems, there is no such thing as a safe electrical system. The NEC and other codes exist to try to protect people from the dangers of electricity. Any time you are working on live equipment, there is an arc flash hazard. The codes deal with how to mitigate that hazard, through proper PPE, remote racking of breakers, arc energy reduction, and other such strategies. The appropriate strategy depends on the equipment and level of hazard present.

Simply not true. The 70E Committee can’t get its act together though and keeps waffling on this,

First off read the definition. A system that is functioning normally does NOT pose an arc flash hazard unless you are doing something that could cause an electric arc, and that is treated in s very limited way. So one element of this is human interaction with the equipment. Sure it can mysteriously explode on its own as you walk by but people have been killed by asteroids randomly, too. 70E is not concerned with highly improbable events, or shouldn’t be. This is the source of the waffling. So if you are just operating equipment that hasn’t tripped because of a fault for instance, the chance of an arc flash is remote. If you see improbable events as “dangerous” then EVERY safety code disagrees with you. Even the ANSI 10 definition of the signal word DANGER disagrees with you. Not all tasks are inherently dangerous, some categorically so.

Second consideration is if an arc flash can occur. Under 250-300 V you cannot get a self sustaining arc without somehow containing the arc in some way. This is the reason IEEE 1)84 previously had a cutoff below 208 V although realistically it goes up to 250-300 V as I stated. But that doesn’t mean it can never happen. In 2009 two electricians in Georgia were removing a construction panel without waiting for a lineman. One was hospitalized and the other died. You can check OSHA reports for this. But I have never seen any evidence anywhere for 120 V arc flash. Bruce Power tested up to 20 kA at 130 VDC which is a much worse case than AC with no sustained arc. Similarly there is also a current (kA) cutoff under the current 1584 edition. And even if one does occur in these conditions testing on actual equipment instead of theoretical tests done by EPRI has shown upper limits. Effectively a lot of equipment can’t produce an arc flash over 2-3 cal/cm2, so low that essentially ANY FR uniform is enough. Hence the current popular FR requirement among utilities.

This isn’t simply a matter of “power limited” cases. Single phase 130 VDC at 20 kA is 2,600 kVA, not exactly “power limited” and typical of large UPS systems and substation batteries.

In addition going with OPs question residential breakers are typically UL type B curves. They trip “instantaneously” (in 1 cycle or less) at 10x rated current. In the vast majority of residential conditions this makes them trip even without an AFCI in the event of a short or arcing fault to ground. Hence the reason that it took decades of incompetence for someone to find a way to die from a 240 VAC arc flash.
 

xptpcrewx

Power System Engineer
Location
Las Vegas, Nevada, USA
Occupation
Licensed Electrical Engineer, Licensed Electrical Contractor, Certified Master Electrician
This is what I wrote to this question on the arc flash forum:

First all, outside of a specific OSHA requirement for utilities to perform an arc flash study, there is no legal requirement. That being said, studies are performed for the following reasons:

1. Arc flash is a known hazard and OSHA requires employers to protect from known hazards (General Duty Clause)
2. Employers are required by OSHA to provide the proper PPE to protect them from hazards in the workplace (1910.132)
3. Some insurance companies will require an arc flash study as a condition of insurance
4. Companies care about their employees and want to provide a safe workplace

Since arc flash is a known hazard and there are means to protect employees from the hazard, if not done and there is an accident, the employer leaves themselves exposed to lawsuits.

Outside of determining the incident energy, a good electrical study will examine short circuit, equipment duty ratings and coordination to enable the employer to have a safe electrical system.
The OSHA General Duty Clause and NEC sections 240.67/240.87 ARE the legal requirements to perform the study. One cannot adequately perform a risk assessment or determine the available arcing current without conducting some kind of analysis or study.

Also see the following OSHA General Industry - 29 CFR 1910 requirements:
  • §1910.132(d)(2),
  • §1910.145(f)(5) and (7),
  • §1910.335(b),
 

xptpcrewx

Power System Engineer
Location
Las Vegas, Nevada, USA
Occupation
Licensed Electrical Engineer, Licensed Electrical Contractor, Certified Master Electrician
If you see improbable events as “dangerous” then EVERY safety code disagrees with you. Even the ANSI 10 definition of the signal word DANGER disagrees with you.
Can you elaborate how probability factors into the OSHA tag signal word?
 

MPdesign

Member
Location
USA
OSHA clearly states - and includes the word "Arc Flash Study" when referring to utility companies.

If they really meant for the millions of buildings out there to all have to have a $10-$20k study, wouldn't they have mentioned it by name here as well instead of referring to 70E?

No offense, but this seems a little crazy for 90% of the Office buildings out there with an 800A main gear.

Now, wait a minute.
Let's assume an 800A main gear in the typical office building. I did see that it is not required for panels fed from 175kVA and less at over 240V. Would that "typically" mean that it is only required if you had people "in-house" working 277/480V. Is that your understanding?
 

jim dungar

Moderator
Staff member
Location
Wisconsin
Occupation
Retired Electrical Engineer - Power Systems
OSHA clearly states - and includes the word "Arc Flash Study" when referring to utility companies.

If they really meant for the millions of buildings out there to all have to have a $10-$20k study, wouldn't they have mentioned it by name here as well instead of referring to 70E?

No offense, but this seems a little crazy for 90% of the Office buildings out there with an 800A main gear.

Now, wait a minute.
Let's assume an 800A main gear in the typical office building. I did see that it is not required for panels fed from 175kVA and less at over 240V. Would that "typically" mean that it is only required if you had people "in-house" working 277/480V. Is that your understanding?
Very few studies, in the 800A service size, are over a couple thousand dollars. Of course there are always exceptions, especially for older buildings

It has been several years since NFPA 70E dropped the exceptions for small transformers. Now the exceptions s is refers to available fault current.
 

kwired

Electron manager
Location
NE Nebraska
OSHA clearly states - and includes the word "Arc Flash Study" when referring to utility companies.

If they really meant for the millions of buildings out there to all have to have a $10-$20k study, wouldn't they have mentioned it by name here as well instead of referring to 70E?

No offense, but this seems a little crazy for 90% of the Office buildings out there with an 800A main gear.

Now, wait a minute.
Let's assume an 800A main gear in the typical office building. I did see that it is not required for panels fed from 175kVA and less at over 240V. Would that "typically" mean that it is only required if you had people "in-house" working 277/480V. Is that your understanding?
If you don't have any "in house" people that work on such things, then you likely don't have any issue from OSHA. But when some "outsider" comes in to work on such things they very well are still subject to OSHA and if they do things correct might need to do some studies just to be able to know what PPE is needed to perform the work they are there to do, unless maybe you don't mind having the main supply shut off while they work, but technically they still need to wear PPE while checking to make sure it is dead.
 

paulengr

Senior Member
Very few studies, in the 800A service size, are over a couple thousand dollars. Of course there are always exceptions, especially for older buildings

It has been several years since NFPA 70E dropped the exceptions for small transformers. Now the exceptions s is refers to available fault current.

70E didn’t “drop it”. They just deferred it to the arc flash engineering method. Not all arc flash methods agree on the “low end”. For instance NESC basically allows anything under 250 V based on testing actual equipment and the fact that it maintains a minimum 4 cal/cm2 PPE while old 1584 had a 125 kVA/208 V cutoff and new 1584 has a kA cutoff. So they rightly removed the cutoff.
 

hillbilly1

Senior Member
Location
Atlanta,Ga
Occupation
Field coordinator/ technical support
In essence, you are all saying that the power's that be, have decided that it would be easier to create a new study for each and every one of millions of buildings (25% of which are empty or almost bankrupt) than to have certain standard PPE that was approved for less than say 600A, 277/480V and 22kAIC?
Is that right?

That sounds about right for the way things go today.
Did I forget to ask, what section requires labeling to not eat the breakers?
I suppose it also has to be written in 4 different languages.

Sorry - long week and I digress.
There is more to it than service size and available fault current, clearing time plays a big part in it. I’ve seen 200 amp 120/208 panel boards with higher PPE requirements than a 200 amp 277/480 volt panel next to it.
 

jim dungar

Moderator
Staff member
Location
Wisconsin
Occupation
Retired Electrical Engineer - Power Systems
70E didn’t “drop it”. They just deferred it to the arc flash engineering method. Not all arc flash methods agree on the “low end”. For instance NESC basically allows anything under 250 V based on testing actual equipment and the fact that it maintains a minimum 4 cal/cm2 PPE while old 1584 had a 125 kVA/208 V cutoff and new 1584 has a kA cutoff. So they rightly removed the cutoff.
The NESC would not typically apply to a 800A service on a small building.
 

paulengr

Senior Member
The NESC would not typically apply to a 800A service on a small building.

Yes but that is why 70E “dropped” the low end.

The nominal amo rating doesn’t matter anyway. Care to guess how dangerous PT cabinets in generating plants can be on the high side of primary fuses?

What matters most is available fault current and especially its effect on the OCPD. If you are pushing 1-2 second clearing times fault current doesn’t matter much. Realistically the longest sustained arc in any of the joint NFPA/70E 208 V testing is 85 ms, not 2 seconds. It does get over 1.2 cal/cm2. So it’s a very low probability here. But then there is the 2009 fatality at 240 V…so it’s possible, however extremely unlikely. So unlikely that it took decades of fatality investigations for even one case to appear.

There is a lot of data manipulation going on too. For instance Caulder, the guy that puts out the arc flash statistics that everyone relies on, looks at the burn injury data. The trouble is OSHA lumps startle cases like the guy that falls off a ladder when a big spark (not arc flash) happens and breaks their neck. It includes electrical burns, both shock and arc flash (non shock). Caulder does strip the startle cases and other non-electrical out but counts shocks with burns as “arc flash”. So he reports 2 shocks per 100,000 workers per year and 1 arc flash. But the reality is if you search the data yourself and remove all the shock-burn cases it drops to more like 100 cases per year. And around 1 out of 14 arc flash cases are fatal. And I have only ever found a two cases not caused by human error, and one was a maybe. In one case reported by OSHA they were tugging on a cable in a vault/trench and a splice pulled apart causing an arc flash to someone technically not near anything exposed or doing that kind of work. I’ve also heard about the traveling arc case that went “around the corner” with an MCC from CBS but never found it. The last case is where the night attendant at a parking garage was switching off the lights at a breaker panel (SWD breakers) and it created a huge arc flash. Earlier that day an electrician was working on the panel and just sort of sandwiched a piece of loose metal over an open breaker slot between the other breakers and the cover. Apparently whatever the attendant did caused it to fall into the bus. So this is clearly human error but more of an innocent bystander case, one for which I have no good answer to.
 

hillbilly1

Senior Member
Location
Atlanta,Ga
Occupation
Field coordinator/ technical support
Yes but that is why 70E “dropped” the low end.

The nominal amo rating doesn’t matter anyway. Care to guess how dangerous PT cabinets in generating plants can be on the high side of primary fuses?

What matters most is available fault current and especially its effect on the OCPD. If you are pushing 1-2 second clearing times fault current doesn’t matter much. Realistically the longest sustained arc in any of the joint NFPA/70E 208 V testing is 85 ms, not 2 seconds. It does get over 1.2 cal/cm2. So it’s a very low probability here. But then there is the 2009 fatality at 240 V…so it’s possible, however extremely unlikely. So unlikely that it took decades of fatality investigations for even one case to appear.

There is a lot of data manipulation going on too. For instance Caulder, the guy that puts out the arc flash statistics that everyone relies on, looks at the burn injury data. The trouble is OSHA lumps startle cases like the guy that falls off a ladder when a big spark (not arc flash) happens and breaks their neck. It includes electrical burns, both shock and arc flash (non shock). Caulder does strip the startle cases and other non-electrical out but counts shocks with burns as “arc flash”. So he reports 2 shocks per 100,000 workers per year and 1 arc flash. But the reality is if you search the data yourself and remove all the shock-burn cases it drops to more like 100 cases per year. And around 1 out of 14 arc flash cases are fatal. And I have only ever found a two cases not caused by human error, and one was a maybe. In one case reported by OSHA they were tugging on a cable in a vault/trench and a splice pulled apart causing an arc flash to someone technically not near anything exposed or doing that kind of work. I’ve also heard about the traveling arc case that went “around the corner” with an MCC from CBS but never found it. The last case is where the night attendant at a parking garage was switching off the lights at a breaker panel (SWD breakers) and it created a huge arc flash. Earlier that day an electrician was working on the panel and just sort of sandwiched a piece of loose metal over an open breaker slot between the other breakers and the cover. Apparently whatever the attendant did caused it to fall into the bus. So this is clearly human error but more of an innocent bystander case, one for which I have no good answer to.
Probably one of those metal panel blanks. I can’t remember if it was FPE, or GE that had those back in the 60’s? I believe.
 

paulengr

Senior Member
Probably one of those metal panel blanks. I can’t remember if it was FPE, or GE that had those back in the 60’s? I believe.

From the description the guy literally just slapped a loose piece of sheet metal in there, not a factory part.

I keep plenty of sheet metal on the truck for covering all kinds of holes. But I also screw, bolt, or do something to attach them.
 
Top