Do you model the 120/208 panels? Where do you end your study?

Not currently. IEEE 1584 currently states that there is no reason to do this because the arc is unlikely to be self sustaining, if the transformer is under 125 kVA. Lab studies more recently have confirmed that sustained arcing faults are at least POSSIBLE. The 1584 committee is under active discussion and this section is likely to change.

Even if it does change you can plug in some pretty mild assumptions and look at whether the incident energy even gets to 1.2 cal/cm^2 at 2 seconds to find the minimum required arcing fault current as well as short circuit current, and it's pretty high in most cases. So for that reason by making some very mild assumptions you can categorically eliminate a lot of cases just based on transformer sizes.

so what kind of sticker should i put on those panels?

I have a feeling that 125kVA and under still produces some incident energies as well.

Arc physics is the problem here. In the case of DC, sustained arcs are widely studied because if for no other reason than the fact that it is important to make neon lighting work. Below about 150 volts DC depending on the available current and the electrode spacing, an arc may not be self-sustaining. In the world of AC what actually happens is that the arc extinguishes when it gets to a zero crossing and then reignites again when the AC voltage gets high enough to restrike the arc. This results in the characteristic "square wave" appearance of an actual AC arc. The restriking voltage depends on not only the electrode spacing and current in DC but also the temperature of the air (breakdown voltage of air is temperature dependent) and the amount of conductive vapor (melted metals) floating around the area. The "hotter" the arc is, the lower the restriking voltage becomes. This is also the reason that as the voltage lowers, the amount of current increases and it does so in a nonlinear fashion since especially at say 120 VAC, you have very little time to create enough heat to restrike on the next cycle.

Some test data that was presented a few years ago that is on the internet clearly demonstrates that although arcs can form below 250 VAC depending on available fault current, they don't last for more than a couple cycles in some cases. And this phenomena even extends up to 480 VAC as well. However more recent and careful test data seems to have muddied the waters and made it clear that enclosure size and many other factors may play a role as well. The result is that to date, nobody has come up with a hard-and-fast rule or threshold.

In response, 70E deleted the rule in the 2012 edition. The IEEE 1584 committee charged with this issue is struggling with it as I understand it from private E-mails.

I've done some calculations on my own. If I plug in extreme assumptions into the existing IEEE 1584 calculation, I came up with my own "low voltage" and "low current" thresholds for 1.2 cal/cm^2 for control circuitry. This approach ASSUMES that self-sustaining arcs can and do occur regardless.

For low voltage/low available fault currents you can also use Hertha Ayrnton's DC equation (remember to multiply by 1.87!) to predict if it is even possible to have a self sustaining arc at DC since in an AC scenario, the requirements are even more rigorous.

Does your company have an Electrical Safety Program, use this Safety Management System to address your company's specific requirements. Don't let an equipment manufacturer or engineer consultant talk you into spending hundreds of thousands of dollars that you don't need to spend by modelling every lighting panel. Use "Method 3" of arc flash hazard analysis which is by policy you stipulate what arc rated clothing is required for 120/208V, in the Electrical Safety Programs I build, minimum 8 cal/cm2 and up to 225kVA transformer size.

Regards;
Terry Becker, P.Eng.
[url="http://www.esps.ca"]www.esps.ca[/url]

This is a great topic, I have two comments/questions. I do technically agree with what you are saying, but, the legal aspect is one where I can not side with you:

1. Where does NFPA 70E say we can exclude the 208 volt circuits within a study and wasn't that section regarding the the exclusion of <125 kVA circuits removed?

2. Where in 1584 does it say "there is no reason to do this because the arc is unlikely to be self sustaining".

I do not like the rules, but, we must play by them and I do not believe you can circumvent NFPA 70E. On the other hand I also believe as you do that most 120/208V faults are not self sustaining, however, I did serve as an expert witness in a case where the the source was a small dry type 120/208V transformer and the electrician did receive a 50% body burn. I have it well documented and would like to write a white paper on the events if I can get a venue.

Even a bit more interesting is a study that I just did. The project had approximately 100 dry type transformers that were < 125 kVA. I ended up modeling the 45, 75 and 112 kVA transformers. I ran studies on a large portion of the units and i also ran studies on the secondary feeder circuits. Every transformer secondary bus and every circuit 100 amperes and greater on the secondary side was a less than 4 cal/sq cm. I do not think this will vary too much on any site. In the future I will spot check these smaller transformer as I do circuit of less than 70 amperes (that is my general cut-of), however, I will start designating all these transformers as a 4 cal/sq cm node.

Yes, it was removed. This is technically correct. The only risk assessment that NFPA 70E has is the tables. It references no less than 8 different methods for calculating an arc flash hazard as well as a single (but woefully incomplete) risk assessment method in the annexes. Since only one of the 8 hazard calculation methods currently specifies the exception, it would be inappropriate to put it in NFPA 70E, irrespective of the reason that it was purportedly inserted and then later removed.



It is buried in a couple paragraphs. And yes, it is so darned convoluted and weakly stated and given so with absolutely no justification whatsoever that the reliance on it is tenuous at best to say the least. The way folks talk about it, you'd think it's much stronger than it is. If it was acceptable prior to NFPA 70E-2012, it is just as acceptable in IEEE 1584. But if you actually bother to read the justification for it, you won't feel so comfortable with your position.



There's a space on this forum for articles. That being said, you're not circumventing 70E. 70E is similar to NESC in this respect. Both give a relatively complete risk assessment method based on a set of tables. Neither specifically attributes any particular source to where the values came from. In the case of 70E, the Technical Committee used IEEE 1584 and then made some adjustments ("credits") without documenting the logic behind those decisions, nor mentioning IEEE 1584. In the case of IEEE C2 (NESC), they used ArcPro for the calculations and then gave an extensive list of adjustments for various tasks which at least in the draft version I've seen (I have not seen what made into the official NESC-2012 document), it was heavily footnoted with sources. ArcPro is not mentioned anywhere in IEEE C2 either...you just have to know about it.



Be careful here. So what specific PPE does the secretary or janitor need when they enter the building in the morning, never mind go into the broom closet and use a snap switch? We really do need not only a "0" category but a "00" category, or else need to get serious about doing risk assessments and not just hazard assessments and call them risk assessments. Right now the proverbial secretary/janitor needs (with your cutoff of 4 cal/cm^2) long sleeve FR pants and shirts, gloves, an arc flash resistant face shield, hard hat, safety glasses, and leather work boots. Although there are some men and women who dress far more casually than is appropriate for the business environment, this does seem kind of ridiculous doesn't it?