Hi everyone

I left university 3 years ago and I was quickly involved in arc flash studies.

So I worked with several senior eng. and had to accept their personnal opinions on arc flash tricky topics, for example the hazard risk category for operating a D.S. with door closed. There is other threads for this one, but here are 2 questions I'd like to have your opinion on.

1.My first question is about fault propagation. In a LV swgr, is it possible for an arc flash on a feeder breaker to move to the main breaker primary, exposing a worker to the usually category 4 or higher hazard risk category?

I agree that the swgr back panels should be labelled with the highest energy level if removing them expose you to the main breaker primary. But should it be the same for the swgr front operations?

2. The bus bars between a transformer secondary and the LV swgr always have a very high energy level that can leads to flash boundaries of 100ft or more.

NFPA 70E 2009 changed the arc flash hazard definition and added "...provided a person is interacting with the equipment in such a manner that could cause an electric arc."

Some say that you should respect the flash boundary at all time, even if you are just walking near the substation. They say they have seen arc flash events that occured without any people interacting with any equipment, and that we should see NFPA 70E as minimum safety guidelines.

I say that that way of thinking leads to overprotect everything.

What do you guys think?

Thank you for your opinions!

1st off, welcome to the forum, good questions, and the answers you will get are based on opioning and interpretation. IMO, I think the probality of that hapening are slim to none, if we dont assume that our OCPD's do thier job the whole arc flash study is worthless. So I would say no to your question.




Right, an arc could happen without any interaction at all, but that would be rare, and I agree that you can overprotect. The 70E has a sub commitee looking at risk vs hazard and this situation is a prime topic. The commitee head on this group had a guy injured walking past a MCC coming back from lunch when it failed. Yes the 70E are minimum guidelines, but it is clear as you noted on the definition of an arc flash hazard, I wouldnt require PPE to walk past the switchgear, but if I were you I would leave that decision up to the owner.

Interested to see what others think about these 2 very good questions.

I label the Main Breaker and the bus separately. I think the chances of a fault propagating from the bus section, through the Main, to the Xfmr output are very, very slim.



I agree with your assessment, and we have modified our procedures as such. Only if you are "interacting" with the equipment are you required to dress out, or, if you know that operations are going on such that the breakers may be operated remotely. This is mainly for our some of our MV starters and breakers with high energy levels and remote operation at plant startups and stops.
I'd hate to see everyone walking around in bear suits all day!
If you feel that uncomfortable around electrical equipment, you should avoid it.

Opening doors on switchgear would be with the PPE required for the calculated IE. Operation with doors closed is difficult to answer. I think the 2009 edition is making it clear that PPE is required just for switching.

As for the nature of regulations. I've been doing this a long time, not Arc Flash, just working with NEC, OSHA, ANSI, etc. What you have is business of regulation that never ends. To say a standard is complete means the revenue stream will stop. My feeling is that we are now at the point that even meeting the regulation requires great effort. I have no desire to try to surpass it.

I guess my rational is different than some. While I never want to see anyone hurt I realize that to absolutely guard against it is impossible. Like car accidents - the only way to absolutely positively guarantee that 100% of the driving public will never be hurt in a car accident - is to prohibit all driving and all cars. Well maybe you can sit in them in your driveway and remember what the open road used to feel like. Somewhere there has to be a balance, we all drive our cars knowing we could be hurt - but it doesn't stop us from driving. What you want to do is make a goal of 100% safety in your plant. More can be achieved through frequent training, supervision, and planning by a competent engineer than any book of codes.

I was reading this again and since english is not my first language, I'd like to be sure my first question was well understood.

If the main breaker line side and the bus (load side) are not isolated, separated from each other, is it possible that an arc flash on the bus move through the air to the main breaker line side?

In other words that the heat and gas and everything from the first arc flash event on the bus initiate an arc flash event on the main breaker line side?

Relative to your question on the propagation to the main breaker:

1. The magnetic forces push the flash away from the source, but:

2. The problem is, if the enclosure fills with the ionized gas the fault will cause the short to bridge the the line side terminals. Take a look at the original IEEE arc flash testing video, the test with open panels, specifically the 6 cycle breaker test with the wireway above the panel. I have performed forensics on multiple events, one that I did recently was on a relatively small enclosure, the event generated 150 cu ft of copper gas. That event pushed back into the enclosure and bridged line side. This was only a single disconnect with an open door, it were switchgear there is no doubt it would have progressed further.

In addition, I have reviewed several additional events that started on the load side of the branch/feeder circuit ocpd and have been pushed back to the line side. When that happens, the ocpd does not operate because the voltage drops on the load side and the fault maintains on the line side.

3. Lastly, with 600V gear, unless it is recent LVPCB's type construction the manufacturer will not put in writing that the incoming ocpd limits the incident energy. There may be some manufacturers aboard and they can add to this. I have multiple clients that have been told by manufacturers that they CAN use the incoming device as part of the mitigation, however, when the manufacturer is asked to put this into writing the answer changes.

I recently received one of those letter and could post it if needed.