For many years, I (along with many others) have been advocating the use of larger “standardized� arc flash boundaries when possible. i.e. perhaps 6 or 8 feet or more. This assumes the larger boundary is sufficient - sometimes the AFB can be quite large as most of you know.

The reason for using a large AFB is that it really only affects the person NOT performing the work - (and not properly protected). It helps with consistency of work practices by attempting to keep things more uniform.

The use of larger arc flash boundaries seems to be slowly catching on.

This week’s question is about arc flash boundaries.

What arc flash boundary do you/your company/clients use?

• Exactly what is calculated
• Something larger
• Values from NFPA 70E Hazard/Risk Tables

If you use something larger, please elaborate

As a consultant, I supply in the report the actual calculated AFB and this is what goes on the label. In the NFPA 70E training that we perform, we do discuss that the AFB is to protect the people that do not have the PPE to be inside the AFB. We discuss barricading at that boundary and that it is prudent to go larger if space permits.

In our plant floor space is at a premium making it difficult to maintain the calculated AFB let alone a lareger one. Keeping other personnel away is also tricky as we try not to shut down too many lines when working on one. I'd like to do more but it's really not an option.

wshaw - sounds like you are shutting down equipment to work on which to me implies a lockout/tagout. Once this has been done and the equipment verified as de-energized wearing the proper PPE, then the arc flash hazard has been removed and no longer applies. Seems like for the time this takes the AFHB could be maintained and protected.
Of course, if you are troubleshooting, then the equipment needs to be live.
Just to make sure, the AFHB needs to be maintained and proper PPE worn when you are interacting with the equipment in a manner that may cause an arc flash. This does not include walking by equipment while it is energized and operating.
Also, in the majority of the studies I have done in various types of industrial facilities, I have found that by the time one gets to a machine control panel, the AFH is usually <1.2 cal/cm^2.

We have a problem caused by short sightedness in the beginning which is that several machines in a line, and sometimes more that one line, is controlled/fed from a common cabinet. Thus establishing an electrical safe condition means killing several machines or lines. I wish the exception for >240 volts fed from a transformer less than 125 kva was still in 70E. I'm betting it will disappear from IEEE 1584 next version too. So now we will be suiting up for everything which doesn't make staff electricians happy.

We show the calculated value on the labels. However, we are starting to teach all nonqualified personnel to just go with 10 feet as an overall boundary distance.

wshaw - a couple of points on your last post. The exception is for <240V not >240V (I am sure that is a typo but for other readers, that needs to be clear)
The guiding document that is used to perform the calculations is IEEE 1584-2002. This contains the statement about equipment fed by a transformer less than 125 kVA and below 240 V need not be considered. Many people performing arc flash hazard analysis use this exception. In NFPA 70E-2012, Article 130.5, there is an informational note number 5 that says to see IEEE 1584 for more information regarding arc flash hazards for three phase systems rated less than 240 volts.

wbd - Thanks for picking up on the typo..... In my hand book on NFPA 70E all the changes to the various chapters are listed in the beginning of the chapter. The exception was specifically deleted from 130.5 in 70 E. There have been a couple of addendums to IEEE 1584 but I don't know what the are. The last one was in 2011. I suspect the exception is gone or soon will be there as well. Hopefully the software I use, EasyPower, will send an update to cover the changes in Ieee 1584 whenever they occur.

wshaw - the exception is still there to use and I suspect there always will be an exception for small circuits. It may turn out to be a current limit vs txf kva limit.
I use EasyPower and force the equipment that falls in that exception to a 1 cal/cm^2 so it shows up on the spreadsheet and labels can be printed.

We use standardised 'exclusion zones' which are set at roughly double the AFB for the worst case example on site. The AFB used to establish the standardised exclusion zone is determined as if the door were open (noting that there is no recognised method for establishing prospective incident energy or AFB beyond a closed door). Excessive clearance time cases for the line side of LV incomers are managed by opening the primary side MV circuit breaker before racking the LV and hence these are excluded from the worst case determination. The standard exclusion zone for LV is 3m. For MV there is a 5m zone and a 10m zone. The philosophy is that a bystander who has valid reason to be present wants to know where they will be safe, and the AFB doesn't answer that question.
Our operation is in Australia by the way and so we have greater flexibility in applying 70E. We don't post AFB's on switchboards but have posters in every switchroom that include the standard exclusion zone details and have training for every electrical worker that includes the arc flash system.

I like that idea and will give it a try. We already have numerous places with small incident energy levels that frustrate the team when they have to suit up to work on them. We have already made it a policy not to bothher with a permit unless the energy level is above 8 cal/cm^2.

Once you do an evaluation of something that can be excluded, you can't then go back and exclude it. That is straight from a discussion with one of the 70E Handbook authors.

Now that we are all using software for arc flash evaluations, where it is trivial to do the evaluation for the transformers in question, you should be doing the evaluation. If someone were to get hurt because you gamed the system so they wouldn't have to wear the PPE that the evaluation would lead to, that would definitely not be good.

I like the idea of a more standardized boundary and making it larger makes a lot of sense. However, I have a question, does anyone list the larger boundary on the label or is this just something that is part of the work practices? - Thanks.

I went ahead and put it right on the label, since there was room for it after I took several things off that we didn't really need.

I would consider the effects of modeling the circuits that meet the IEEE 1584 exception. For example it is possible to have high incident energy on the low side of a small transformer where the only protection is the protective device on the primary side. There have been papers and discussions on this, mainly saying that the small arcs will not sustain. How reasonable is it to have someone in a HRC 3 or 4 PPE to operate a 30A 208V breaker in a 200A MLO panel? And then that same individual goes to operate a 1600A 480V breaker with HRC 0 or 1 PPE? I would think the workforce would probably question the credibility of the study and maybe take short cuts on the back shifts.

The high incident energy comes from a small arc lasting a very long time. Contrast that against a very large arc lasting a very short time. The effects are not the same.

Two points -
1. More recent testing resulted in more arcs sustaining at the lower voltage, which is probably going to result in the eventual new release of the standard reducing the transformer size in the exception, if not eliminating the exception completely. The example you cited is exactly what I found when I came to the place I'm at now. There are lighting panels that are category 3 or category 4 because there is nothing after the primary fuses, and the fuses might not blow in the event of an arc fault on the secondary side. We are planning to mitigate this by installing fuses on the secondary of the transformer, which would leave the transformer itself a high category item, but drop the lighting panel to a more reasonable range.
2. Hazard/Risk Category is supposed to take into account both the incident energy available and the risk of the job. If we were able to do that with discretion ourselves, we would drop the category on the 208 V lighting panel and raise the category on the 480 V distribution breaker.

You are supposed to be doing this now

These are some of the places where 'risk' is mentioned in NFPA 70E.
110.3(F) is titled "Hazard Identification and Risk Assessment Procedure". This section requires a risk assessment be performed.
130.2 Exception; for live work permit says "...provided a risk assessment is performed...."
Annex F is completely about risk

Ok, well that is the prerogative of anyone doing the arc flash evaluation. One considers the standards that are in place and balances that against what a calculation would show. I ran into a situation recently where at the secondary metering package which contained utility meters along with a breaker on the load side of the meter. This was at 208V and the calculated AFH was a HRC 3 to operate the breaker but utility cited a study done I think by either EPRI. EEI or PGE that said the AFH would be only HRC 0. So we ended up with the client facility having to use HRC 3 PPE and the utiltiy doing any work on the meter or transformer on the 208V side, HRC 0. How does that make sense?
Same situation arises when you don't take the IEEE 1584 exception

It sounds like it might be a long time before 1584 is revised (a decade? IEEE is glacial except IEEE C2). The exception should be deleted for two reasons. First, Annex D references several methods and 70E itself has a tabular method. In the spirit of not prejudicing one method against another, it would be inappropriate to list the 1584 exception. Second, although PG&E data showed arcs don't sustain below 240 volts, unpublished IEEE/NFPA data says it can sometimes. It depends on the voltage, available fault current, electrode configuration, and size of the enclosure. A simple kva rule would not work but a volts/amps rule could, although the fault current cutoff could be raised depending on enclosure factors. The PG&E study did not capture these nuances and showed optimistic results. As a sidenote, the "fuse" for most tests was a 14 gauge bare wire. At low voltages/currents, a smaller "fuse" wire is sometimes needed to initiate or sustain an arc. This issue alone confounded a lot of earlier tests. A further issue at least currently is that on the low end, 1584 significantly overstates the incident energy with weak arcs. The low cutoff threshold for the other procedures is also quite high. I ran into this issue recently with a very large (500 kva) 208 volt distribution system where the available fault current was around 2 ka. 1584 predicts 10 cal but test data shows under 1.2 cal for similar equipment. A judgement call had to be made to blindly follow 1584, the 70E tables, or test data.

I will post an update about what is happening with IEEE 1584 soon. (long overdue). Yes, it is going to take a long time yet. Some of you may know, I became Secretary of IEEE 1584 earlier this year and the view from deep inside is quite interesting.

We are still in the testing phases. Lots of tests have been conducted, new configurations, equations etc. have been developed that we are reviewing and tweaking.

As far as the sustainability, the problem is if you tweak the configuration enough, you can sustain very low currents at low voltage. The question (and sometimes heated debate) is what is a realistic configuration. Unfortunately that question is yet to be answered. We had some tentative language (posted elsewhere on the forum) where we were thinking of a cut off of maybe 30 or 45 kVA and/or 3 or 4 kA for the exact reason that Paul states with his 500 kVA transformer - it's really a matter of short circuit current and kVA was just a method to help define fault current. My personal thought is it is better to define fault current in terms of amps. None of this has been cast in concrete, what I can say is whatever the low cut off becomes, it will NOT be 125 kVA and will be lower.

Stay tuned, I'll try to give a more complete update in a few days.