scope of arc flash study

[malachi constant]

I started a thread a few weeks ago in the estimating forum (see here: http://forums.mikeholt.com/showthread.php?t=151588)
I have since purchased the Jim Phillips "Complete Guide to Arc Flash Hazard Calculation Studies" book. About a quarter of my way through it. So far it is seems like an excellent resource.

Thought I'd bring the thread over here as my question is now more about safety (the initial post was more about design time / fees). I also understand the situation a little more.

We have a client, a school district, they own more than a dozen buildings. A couple years ago we educated them about arc flash requirements and they decided to put an arc flash study in the queue, but only for the services at each facility. That engineer has left our company, and two years later, I am prepping to do the study. I know that NFPA 72E requires that equipment such as downstream three-phase panelboards be included in an arc flash study (is this a newer requirement?). I am concerned that by studying ONLY the service entrance the client (and myself) will be exposed to liability should an accident occur downstream somewhere. (And of course, to heck with liability, I am concerned that a serious accident might occur at all.)

What is the right thing to do? We do not want to recommend our client spend a lot of money on a full-blown study if slapping some chart-generated labels on the equipment is sufficient for safety. Nor do we want to take a client that has bought into arc flash and talk them out of it into something less safe. Nor do we want to do something half-baked that results in part of a building's system being appropriately labeled and the rest of the facility not labeled at all (or not labeled helpfully, or whatever).

Can you help me understand what are the practical hazards to be avoided while working live in a K-12 facility - like, can it be assumed that with proper training and labeling it can be reasonable to expect nothing ever be worked on live except for the main service? I would think if all panels are labeled "MUST BE PROPERLY SHUT DOWN AT MAIN SERVICE PRIOR TO OPENING" that might cover it. At least just as much as saying "MUST WEAR such and such PPE EQUIPMENT" - you can label it with reasonable rules, but you can't make everyone always follow the rules. At some point, if you've educated them and clearly labeled the equipment, it's their problem, not yours.

Anyway, I'm leaning towards doing a study of all service entrances as a phase 1. That would get me into each building, get utility data, major motors and basic riser diagrams documented, etc. Then as part of this "service entrance only" study I could analyze what I now know about the building electrical systems and recommend, on a case-by-case basis, in each building they do either a full-blown study or basic chart-based labeling.

I have pulled plenty of wire / bent plenty of conduit in my day, and worked on many branch circuits live, but even with that experience I am more or less clueless when it comes to how things (or rather electricians) work in the real world. Can you help me strike the prudent balance here? Thank you.

 

[wbdvt]

First off, I would be concern about your qualifications to do the study if you are just reading the book. Please don't take offense but it is a complicated area. Are you a degreed electrical engineer? Do you have a P.E. license/registration? How many of these studies have you done? What software are you using? These are just a few questions that a company/person should be asking of the company/individual doing the study.

This statement in particular bothers me....."I have pulled plenty of wire / bent plenty of conduit in my day, and worked on many branch circuits live, but even with that experience I am more or less clueless when it comes to how things (or rather electricians) work in the real world" as this type of work is best left in realm of degreed, experienced engineers.

My answer to this..." Can you help me strike the prudent balance here? " would be to hire someone experienced!

I don't want to be harsh here but someone who has done many studies would know the answers to the questions you are asking? Are you being mentored by someone experienced in this field of study?

All that being said:
1. The right thing to do is to identify all the pieces of equipment that can be interactive with in a manner that could cause an arc flash. This would be circuit breakers, disconnect switches, fused disconnect switches, control panels where personnel might have to troubleshoot energized, etc.
2. Perform an arc flash hazard analysis to those pieces of equipment.
3. Perform a short circuit analysis to determine if any pieces of equipment (breakers, panels, switchgear, disconnect switches, etc.) are overdutied.
4. Determine if a manufacturer's or industry good practices maintenance program is in place.

 

[malachi constant]

First off, I would be concern about your qualifications to do the study if you are just reading the book. Please don't take offense but it is a complicated area. Are you a degreed electrical engineer? Do you have a P.E. license/registration? How many of these studies have you done? What software are you using? These are just a few questions that a company/person should be asking of the company/individual doing the study.

Licensed PE, thirteen years engineering experience.

I've done what seems like hundreds (probably more like 50) short circuit studies using SKM software. (Related - I received a week of training from one of their Houston classes.)
When I was an EIT I performed a few arc flash studies - around 2005 or so - under the oversight of another engineer. He didn't know what he was doing. We just set up the short circuit study, added overcurrent protection, clicked over to the arc flash module, and gave the info to an electrician to make the stickers. I know a lot more now and plan to do better. Unfortunately none of the other engineers I have worked with have ever performed arc flash studies.

When I was in college I worked a summer for an electrician. That's where I "pulled lots of wire, bent lots of conduit." Among other things did a service changeover for a school. I've since rewired my house (legal in my state to do without an electrician's license). (Side note - the original electrical system on my 1923 house was installed with conduit! so was able to pull new wire including ground to almost every receptacle in the house. How cool is that?) My strippers and linesmen have more notches in them from nicking them while working live than I care to admit. I don't pretend to be an electrician but I am sympathetic to the effort - the stimulation and the tedium, the fun and the frustrations.

Hope that makes you feel a little better about dialoguing. I wish my answer was "I've done what seems like hundreds of arc flash studies", but I haven't. You have to start somewhere. I am trying my best to take this MORE seriously and be MORE conservative in my design approach than I should. But I don't want to waste our clients' money by overselling them something. Some people make a good career on overselling products and services, good for them. But I value long term relationships and don't want any client of mine to feel like they were duped into buying a service they didn't need. They're willing to pay, the trust me to make the right decision, that's what I'm trying to figure out for them with this post.

Sorry for getting all defensive. I completely understand why you asked the question, and it was probably prudent of you to do so. Thanks!

 

[charlie b]

We just set up the short circuit study, added overcurrent protection, clicked over to the arc flash module, and gave the info to an electrician to make the stickers.

That is essentially all that is needed. There are, of course, a couple tricky aspects of the process. The first is selecting the overcurrent protection devices, and if appropriate choosing the settings, for this has a huge influence on the final result. Another is evaluating the results to determine whether they are reasonable, and this is where experience comes into play. Finally, there is the question of what to do if you (or the owenr) does not like the results. There are many things that can be done to reduce the available arc flash energy, as calculated by SKM or another software package. But any single thing you change (like a breaker setting or a conductor length) will have an impact on other aspects of the analysis (such as selective coordination or voltage drop).

 

[malachi constant]

All that being said:
1. The right thing to do is to identify all the pieces of equipment that can be interactive with in a manner that could cause an arc flash. This would be circuit breakers, disconnect switches, fused disconnect switches, control panels where personnel might have to troubleshoot energized, etc.
2. Perform an arc flash hazard analysis to those pieces of equipment.
3. Perform a short circuit analysis to determine if any pieces of equipment (breakers, panels, switchgear, disconnect switches, etc.) are overdutied.
4. Determine if a manufacturer's or industry good practices maintenance program is in place.

I guess #1 gets at the nature of my question. There are panels all over a school. This is a school, not a manufacturing line. Is it reasonable to advise them they could set up a maintenance plan that says:
* The only thing they are allowed to work in live is the main switchgear. (Maybe add to that other switchboards & large distribution panels.)
* Everything else needs to be shut down at the source and locked out / tagged out.
* Label everything to this effect.

Thinking it through now, the hole I see in that plan is opening up a switch is not a guarantee that the circuit opened - correct? I assume you could open up a switch/breaker but the internal mechanisms fail and one or more phases remains live. Then you open up your panel and have a potential arc flash hazard. You should of course put a meter on each of the phases before you start your work, but the fact remains you could be exposed to a live part even if you lock out / tag out at the source. So if this is correct, it would be unsafe to not define an appropriate level of PPE at each downstream panel, MCC etc.

If that is correct then the next question is, is it reasonable to run an arc flash study at the service entrance (taking into consideration utilities & large motors), but save money on labeling downstream equipment per NFPA 72E tables?

I really have my client's interests at heart here. I do not want to recommend that they triple their short circuit fee (my fee) just because it is convenient for me and could be construed as "the right thing to do". There may be a perfectly reasonable approach that costs much less than a full blown study of every piece of equipment in each building.

Thanks!

 

[malachi constant]

That is essentially all that is needed. There are, of course, a couple tricky aspects of the process. The first is selecting the overcurrent protection devices, and if appropriate choosing the settings, for this has a huge influence on the final result. Another is evaluating the results to determine whether they are reasonable, and this is where experience comes into play. Finally, there is the question of what to do if you (or the owenr) does not like the results. There are many things that can be done to reduce the available arc flash energy, as calculated by SKM or another software package. But any single thing you change (like a breaker setting or a conductor length) will have an impact on other aspects of the analysis (such as selective coordination or voltage drop).

I'd read an article that suggested a good study involved running and comparing a variety of simulations. The crux of the argument being that engineers tend to be conservative, for instance assuming worst-case values that result in maximum fault currents. But (the argument goes) worst-case fault current does not always create the worst case arc flash, because a smaller fault may last longer without tripping, allowing more net energy to be released into the flash. So at a minimum you should run two scenarios (say, worst-case and "realistic-case"), compare results with your trip curves, see if there is a chance something bigger could happen in a realistic scenario.

Doesn't sound like a terrible amount of work. Sounds kind of fun, actually. (Real engineering - not just plugging in numbers!) Fun or not, is it something you would recommend doing?

 

[zog]

First off, glad you like the book, Jim did a great job, I may have helped a little

Second, it is 70E, not 72E, once is a typo, twice makes me think you have not read it? You also need to read IEEE 1584, both documents cover many of the questions you are asking.

Yes you have to run at least 2 senarios and yes it is realistic that lower arcing currents can result in the highest incident energies.

As far as your SC study, even if you use the 70E tables you need to know available fault current and clearing time of the OCPD to know if the tables can be used so you already have most of the data you need to complete the analysis. There will be some point where you stop the study, <240V and <112.5 kVA is a common cut off point, that used to be in the 70E but was removed in the 2009? cycle due to some recent testing that shows that cutoff may not be realistic as once thought.

 

[mayanees]

good questions mc

good questions mc

I think you're asking all the right questions mc, AND exercising practical engineering judgment by attempting to provide your client with a minimum essential service.
I can't say I'm a fan of a full blown analysis for every panel in a facility, but 70E doesn't leave many options. As zog wrote, there USED TO BE a cutoff for transformers under 125 kVA, but that's gone, so the right thing to do is include all panels.
The main caution I have - before you change instantaneous or stpu levels to reduce the HRC, is to make sure you know the characterisitics of the load, in terms on the inrush requirements, and allow a tolerance of 15-20% higher with your settings.
Before arc flash, if you did a Study and were ever called back to a facilty to re-adjust a breaker because it tripped, and you could do it, then you didn't provide the proper settings to start with. Nowadays there's a tendency to try to get everything to HRC 2, and that could compromise reliability, such that your adjustments may cause a trip. In that case you might get called back, and could adjust things up to tolerate the load, which isn't desirable. It looks like you didn't know what you were doing.
In my opinion, if a system is analyzed properly, knowing load and ocpd characteristics, and it comes up to be HRC Dangerous, it's not a ticking time bomb that needs remediation. An info note in 70E 2012 states that equipment, when used as intended, is not likely to experience an arc flash (papraphrased). You just have to treat it as such.
Good luck.
John M