Arc Flash - Where do you stop?

[JensenDrive]

How far down in the electrical system do you carry arc flash analysis? It seems one has to go to every load hanging off of a panel that an electrician might work on live. Carried to an extreme, that means that every 277V light fixture and every load downstream of a 480V distribution panel needs an arc flash analysis.

 

[rbulsara]

Per the existing guidlines of IEEE/ NFPA 70, Arc Flash analysis is not required for systems fed by tranformers rated 125kVA or less at 240V or less. The hazard catergory is 0.

 

[Zogzog]

Sorry, I have to disagree with you there. While your statement is stated in the 1584 as the point where an arc will not be self sustaining and therefore not require a protective device to clear it the 70E conflicts with this.

If you do not do an analysis below this point you would need to default to the 70E tables. For panelboards 240 and below some tasks are HRC 1.

There was a proposal for the 2009 70E that was accepted that will no longer allow generic labels, the new labeling requirement will require either the Ei or HRC.

 

[Eleceng01]

ZogZog,

DO you have a link to the proposed 2009 70E standards. I also agree with you that there is a difference on the requirements of 70E and IEEE1584. Thanks.

 

[rbulsara]

We are still in 2007.

 

[rbulsara]

If some paranoid people get their way, I can see requiring all residents happily going around their houses in PPE suits, in very near future.

 

[Zogzog]

It should be on the NFPA site, I was given a copy. I just finished about 16 hours of reading every proposal, 232 pages at about a 4pt font.

Starting next week all of my courses will include notes for each part of 70E that will be affected by these changes.

 

[resqcapt19]

You can download the 70E ROP here:

http://www.nfpa.org/Assets/Files/PDF/ROP/70E-A2008-ROP.pdf

Also you can call the NFPA and ask for a hard copy...they are free. Comments on the proposed changes will be accepted until 8/31/07. The final document will be voted on at the June 2008 meeting.
Don

 

[JensenDrive]

Interesting conversation, but I still do not know the answer. I have a large facility with maybe 1000 MCC / panelboard breakers at 480V. Many of them are 20A breakers going out to some control valve, and any of them might be worked on live. Many of them feed panels that may have 20 circuits hanging off of them to various small loads, such as heaters and temporary welding recepticles. Some feed lighting panels, with 20-40 lighting breakers, and each lighting breaker may have a dozen lights on it.

To do this right, it seems you need to do arc flash analysis at every one of these elements. I do not see too much discussion of the magnitude of the data research and data entry involved. Or maybe I am missing a general concept that a rule is developed for the facility, "For this list of equipment, where you might have considered working on it live in the past, that is not allowed any longer."

 

[JensenDrive]

Something I said in another posting that gets to my problem is:

My messy problem is that a guy pays me to provide him with arc flash analysis to meet code requirements, and without clear guidance on where to stop, there is an order of magnitude difference on what this means. The scope needs to be in the contract up front I guess, but it would be nice if there was guidance beyond, "Well, no one seems interested in that normally, but I have no technical reason why it should not be included, so I guess I could include it in the calc if you want me too." Sounds pretty unprofessional an answer.

 

[alehman]

70E doesn't provide a minimum circuit size for arc flash hazard analysis. You can view the current version on line at

http://www.nfpa.org/freecodes/free_access_document.asp?id=70E04

 

[JensenDrive]

But do you stop at the panel, or do another arc flash analysis at the end load? I mean a guy is trying to figure out why the valve is not operating. That very well may mean at least putting meters into the final end valve motor, with it live. Do you put an arc flash label on the junction box at the valve's motor?

 

[alehman]

NFPA 70E says you perform the arc hazard analysis where the hazard exists. If a worker is in the junction box at the valve, you run the calculation at the valve.

70E does not address flash hazard labeling requirements. NFPA 70 (NEC) requires labeling indicating presence of arc flash hazard (but not specific energy levels) for switchboards, panelboards, industrial control panels, meter sockets and motor control centers in other than residential occupancies that are likely to require maintenance while energized.

 

[rbulsara]

A lot of discussion has taken place on the subject at this site. Just search for arc flash.

This one may be more relavent: thread238-171682

NFPA 70E is based on IEEE 1584 and recognizes its use.

 

[Zogzog]

You can decide where to stop. The <240V and <125 kVA is probally the way to go. Above that point, do an analysis, below that point use the tables and put a label on the equipment that states the HRC and PPE required (from the tables) to work on that equipment.

 

[JensenDrive]

I will call it quits at this point, but the thing that surprises me is the low significance people give to the extra effort involved in arc flash analysis. If someone asks me to do a fault withstand and coordination study, I go to a panel, find the slowest breaker, and lowest Isc rated breaker, and the analysis ends there.

Now add arc flash, I need to find the data on every breaker in the panel, every feed off of it, and then find the fault current at the end load and how long it will take to clear that fault.

I have a 10 fold increase in the work level to analyze the panel. WOW. Billable hours rock. :-D

 

[DanDel]

As long as they're quoted correctly in the first place!

 

[jraef]

One person's P.I.T A. is another's gold mine!

I recently visited a local automotive plant, they have identified 5420 panels that need Arc Flash evals. Lots of billable hours for lots of people.... but expect your next pickup truck to cost just a little more.