It is currently Thu May 28, 2020 12:39 am



Post new topic Reply to topic
Author Message
ekstra   ara
 Post subject: MCC has proper segregation?
PostPosted: Mon Jul 17, 2017 10:07 pm 
Sparks Level

Joined: Sat Apr 04, 2009 9:50 pm
Posts: 104
Location: San Antonio, TX
IEEE 1584.1 proposes that proper segregation should be checked between the section containing the main OCPD and the adjacent sections.

I had a conversation with a customer today who desires me to affix two labels for an MCC. One with the label of the AFIE at the line side of the main OCPD and one with the AFIE at the load side of the main OCPD.

The MCC's are not tested for proper segregation and they have metal barriers between different sections.

Question:

Do MCC manufacturers guarantee proper segregation of line side section with the adjacent load side section for non-arc resistant MCC's?

If not, do you feel comfortable that the arc flash energy generated on the line side section will not propagate and invade the adjacent load side section? That the adjacent load side section will withstand the blast?

I had the argument that ALL NFPA 70E committee members agree that is safe to place two labels in MCC's. Do you agree with that?

I was also told that all the main and most respectable engineering companies performing arc flash studies provide two labels, line side and load side. Do you think this is true?

I was told that anybody that requires only line side labels (without providing load side labels) is against the industry standard and imposing an unreasonable restriction to the customer. Is that what you think?

Now, I am concerned about his issue, because I am only placing arc flash labels in the line side of any equipment (including MCC's) unles they are tested by the manufacturer and the manufacturer guarantees in writing the proper segregation.

Your answers will make me change my standard policy if it is shown good evidence that what I am doing is not right.

I have being wrong many times before.

Have a great day!

Thanks so much for all your answers.

RECS


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: MCC has proper segregation?
PostPosted: Tue Jul 18, 2017 7:54 am 
Arc Level

Joined: Mon Jan 18, 2010 11:35 am
Posts: 514
Location: Wisconsin
RECS wrote:
... the manufacturer guarantees in writing the proper segregation.


I doubt that you will ever get a guarantee, even on Arc Rated equipment, after all, if you go to any extreme measurement nothing is 100%.

The issue is if the equipment construction will 'slow down' the arc movement (plasma) giving the integral main device enough time to clear before becoming engulfed in the event.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: MCC has proper segregation?
PostPosted: Tue Jul 18, 2017 11:04 am 
Plasma Level
User avatar

Joined: Tue Oct 26, 2010 9:08 am
Posts: 2174
Location: North Carolina
There are no guarantees of any kind.

With panelboards it is pretty obvious that you need to treat line and load sides identical.

With isolated switchgear sections and even isolated enclosures (provided that the adjacent wall hasn't been removed) it should be obvious that isolation exists.

When it comes to MCC's, it's not so clear. Testing has shown that it can go either way since within a bucket it is FAIRLY isolated except for a huge opening onto the bus bars. However within the bus bar cavity outside of the MCC buckets it's all open inside.

The question is whether or not an arc can "hop" a breaker from load to line side. It can and has been shown to do this but how often it can do it in real world conditions vs. a lab is anyone's guess.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: MCC has proper segregation?
PostPosted: Tue Jul 18, 2017 11:37 am 
Arc Level

Joined: Mon Jan 18, 2010 11:35 am
Posts: 514
Location: Wisconsin
PaulEngr wrote:
When it comes to MCC's, it's not so clear. Testing has shown that it can go either way since within a bucket it is FAIRLY isolated except for a huge opening onto the bus bars. However within the bus bar cavity outside of the MCC buckets it's all open inside.

The question is whether or not an arc can "hop" a breaker from load to line side. It can and has been shown to do this but how often it can do it in real world conditions vs. a lab is anyone's guess.


I think an MCC has three issues for consideration. For sure, if the main device is not operating in its Instantaneous region, all bets are off.

1) the isolation between an individual buck and the vertical busing in a section. As you mentioned this is minimal to non existent. I am always amazed when I see Arc flash labels on individual buckets.
2) the isolation between vertical sections. Will an arc, or plasma cloud, move up a vertical bus bar then cross over and down into an adjacent section
3) the isolation between the main device lugs and the bussing in vertical sections. Say the main device is fed from the bottom, then an arcing event would need to travel up a vertical section, move horizontal between sections, and then move down the vertical bus and jump over the main device to the line lugs


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: MCC has proper segregation?
PostPosted: Tue Jul 18, 2017 1:29 pm 
Plasma Level
User avatar

Joined: Tue Oct 26, 2010 9:08 am
Posts: 2174
Location: North Carolina
JBD wrote:
PaulEngr wrote:
When it comes to MCC's, it's not so clear. Testing has shown that it can go either way since within a bucket it is FAIRLY isolated except for a huge opening onto the bus bars. However within the bus bar cavity outside of the MCC buckets it's all open inside.

The question is whether or not an arc can "hop" a breaker from load to line side. It can and has been shown to do this but how often it can do it in real world conditions vs. a lab is anyone's guess.


I think an MCC has three issues for consideration. For sure, if the main device is not operating in its Instantaneous region, all bets are off.


Not necessarily. Work by Mike Lang and others at Mersen and confirmed by other tests shows that arc propagation phase-to-phase happens in about 1 cycle. Given an instantaneous trip up in the 400+ A range typical of an MCC main breaker of about 3 cycles that still gives you 2 cycles of full 3 phase arcing fault and some portion of the first cycle of a single phase arcing fault...so let's just count it as 3 cycles. It's not until you get down to the small (<250 A? <100 A?) MCCB's that you start to see 1 cycle tripping.

Quote:
1) the isolation between an individual buck and the vertical busing in a section. As you mentioned this is minimal to non existent. I am always amazed when I see Arc flash labels on individual buckets.


More to the point one of the common ways to cause one (not that the practice should be done) is plugging a bucket into a live bus. Interestingly in the couple cases I've seen the arc is magnetically propelled to the bottom of the bus bar where it then blows out the door at the BOTTOM of the enclosure, not at the bucket where the arc is initiated. In fact even with the actual MCC testing I've seen done in open literature, the same thing happens.

Quote:
2) the isolation between vertical sections. Will an arc, or plasma cloud, move up a vertical bus bar then cross over and down into an adjacent section


The plasma moves away from the power source like a rail gun so no it is not going to move "up". The only way that this would/could happen and I forgot the location of the paper but testing on this in I think one case finally got enough hot gases and the like to cause arcing to move "up"...the air above the arc heated to the point where it started further up the busbar. This isn't arc tracking like normal...it's something entirely different. They were only able to get it to happen by removing a lot of buckets and other hardware to make an "empty" chamber. Otherwise the obstructions block radiant heat and conductive (smoke) heat transfer isn't really very effective since by the time you'd get it that hot, the doors would blow off and it ventilates...at least that's my speculation, without any scientific evidence whatsoever. Also arcs don't "anchor" due to insulated bus bars like what is claimed in the UNTESTED literature...they move across it as if it's not even there. Insulated bus may contribute to blocking initiation of an arcing fault but won't impede one in progress.

Quote:
3) the isolation between the main device lugs and the bussing in vertical sections. Say the main device is fed from the bottom, then an arcing event would need to travel up a vertical section, move horizontal between sections, and then move down the vertical bus and jump over the main device to the line lugs


Moving up, across, and down, would seem pretty likely since we have magnetic forces for propulsion that well exceed gravity. The effect is called "arc tracking" and detecting it is one way to find the point of origin, as documented in the NFPA standard on fire investigations. It makes very characteristic looking "teeth" marks on the bus bars. It's also what is happening in the relatively harmless "Jacob's Ladder" devices although obviously with a wide arc, you can see some lifting due to rising hot air around the arc. But every case of "jumping" that I've ever seen reported in the literature somehow contains enough smoke and hot gas to lower the conductivity of air enough to go around the obstruction (overcurrent device). I never paid any attention to whether it is "up" or "down" although now that you mention it, it would seem like vertical up is probably the most probable (heat rises). Vertical down just doesn't seem likely at least past an obstruction.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: MCC has proper segregation?
PostPosted: Tue Jul 18, 2017 9:55 pm 
Sparks Level

Joined: Sat Apr 04, 2009 9:50 pm
Posts: 104
Location: San Antonio, TX
Thanks for all your answers.

A couple of questions regarding your explanations.

1.
Am I interpreting your comments that an arc flash event on the line side (incoming section) of a standard MCC with metal barrier between the incoming section and the adjacent load side section can propagate to the load side adjacent sections making the AFIE at the load side adjacent section the same as the AFIE at the line side of the main CB?

2.
The question is to Jim Phillips or any other member that is in close touch with the NFPA 70E committee members.
Do you know what is the interpretation of the majority of the NFPA 70E committee members about this subject? Do they have the opinion that in standard MCC's, the metal barrier has enough segregation so to use two labels for the MCC's one for the line side and the other for the load side of the main CB?


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: MCC has proper segregation?
PostPosted: Wed Jul 19, 2017 7:19 am 
Arc Level

Joined: Mon Jan 18, 2010 11:35 am
Posts: 514
Location: Wisconsin
[quote="PaulEngr"]
Not necessarily. Work by Mike Lang and others at Mersen and confirmed by other tests shows that arc propagation phase-to-phase happens in about 1 cycle. Given an instantaneous trip up in the 400+ A range typical of an MCC main breaker of about 3 cycles that still gives you 2 cycles of full 3 phase arcing fault and some portion of the first cycle of a single phase arcing fault...so let's just count it as 3 cycles. It's not until you get down to the small (<250 A? <100 A?) MCCB's that you start to see 1 cycle tripping.

[quote]

Dont' go off topic too far. Propagation from L-N to L-L-L is germane.
My discussion has been concerning propagation through the MCC and across its incoming protective device. Based on the OP description, I have been concentrating on LV MCCs.

From your last comment, it appears you fall into the group that might accept the risk analysis of a bottom fed main breaker as likely to clear, if the fault starts in a different vertical section and construction like 'fish tape barriers' has been employed. Faults in the vertical section with the main, and top fed devices, are likely to engulf the main device.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: MCC has proper segregation?
PostPosted: Wed Jul 19, 2017 8:38 am 
Arc Level

Joined: Tue Jan 13, 2009 5:00 pm
Posts: 524
RECS wrote:



1.
Am I interpreting your comments that an arc flash event on the line side (incoming section) of a standard MCC with metal barrier between the incoming section and the adjacent load side section can propagate to the load side adjacent sections making the AFIE at the load side adjacent section the same as the AFIE at the line side of the main CB?


I believe the real question is the reverse of this one. Work is taking place on the load side of the device in question, and the worker inadvertently initiates an arc. Does the device in question operate quickly and clear the fault with low resultant IE, or does the arc break over to the source side where the clearing times are longer and the IE is higher?


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: MCC has proper segregation?
PostPosted: Thu Jul 20, 2017 8:28 am 
Plasma Level
User avatar

Joined: Tue Oct 26, 2010 9:08 am
Posts: 2174
Location: North Carolina
JBD wrote:
Dont' go off topic too far. Propagation from L-N to L-L-L is germane.
My discussion has been concerning propagation through the MCC and across its incoming protective device. Based on the OP description, I have been concentrating on LV MCCs.

From your last comment, it appears you fall into the group that might accept the risk analysis of a bottom fed main breaker as likely to clear, if the fault starts in a different vertical section and construction like 'fish tape barriers' has been employed. Faults in the vertical section with the main, and top fed devices, are likely to engulf the main device.


http://www.neiengineering.com/wp-conten ... esting.pdf

That should answer most of your questions, at least somewhat.

Also, don't bother with "Part 1". In that publication, the authors claimed that insulated bus bars would prevent arc propagation altogether and a bunch of similar claims. Part 2 which is the actual test work refutes most of the claims in Part 1. The available fault currents in Part 2 also set something of an upper limit on what sort of propagation would seem to be "possible"...higher fault currents might produce different results but at least this gives us an idea even if it's essentially a single data point.

I don't think there is anyone expressing an opinion about propagation between sections of switchgear with separate enclosures for each section. I've worked a lot with mining gear and the medium voltage stuff generally falls under metal enclosed rules...some of it is much more open from one section to another than others. Every failure I've seen at 4160 and 2300 stays in it's own enclosure on that stuff. For example I've seen a switchgear where arcing propagated across about 3 sections but it was in the back where everything is open and has horizontal bus bars and splice plates tying everything together essentially n one long compartment. Again...not really LV MCC but the principal is similar.

I've seen plenty of "real world" MCC's as I said where the bottom panel blew out in the same vertical MCC section. I have not personally seen any cases but have heard reports from folks doing testing of where it "hops" a circuit breaker. So I think I am pretty comfortable with the idea that from an arc flash point of view we need to at least treat individual sections as if they are all one enclosure.

When it comes to going from section to section I have no evidence of this happening. It is plausible and so I can see someone making a possible case for arcs travelling across the horizontal bus from one section to another. This is irrespective of how much plating is between sections. As you can see from the report above any barrier to arc tracking would have to be probably something like mill board and very substantial. But I have zero evidence for cases of arcs travelling from section to section. except in the above "single compartment" example.

The "substantial barrier" is the part I have problems with. I don't think that a 14 gauge plate even with some thin rubber or glastic "glands" between sections is going to stop arc propagation based on the above MCC testing. The substantial size of the gland is going to be the determining factor.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: MCC has proper segregation?
PostPosted: Thu Jul 20, 2017 1:26 pm 
Sparks Level

Joined: Sat Apr 04, 2009 9:50 pm
Posts: 104
Location: San Antonio, TX
Thanks for all these interesting answers.

If you are doing an arc flash hazard analysis, what would you do for the labels of a LV MCC with a main over current protective device?

1. Label the incoming section where the main OCPD is located with the AFIE calculate at the line side of the main OCPD and also label the sections on the load side of the main OCPD with the AFIE calculated at the line side of the main OCPD?

2. Label the entire MCC with the worst case AFIE which is normally the one calculated at the line side of the main OCPD?

Thanks.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: MCC has proper segregation?
PostPosted: Fri Jul 21, 2017 5:03 am 

Joined: Fri Nov 06, 2015 1:05 am
Posts: 25
Location: Evansville IN
I am currently working in a very large facility. We have come to a consensus at this time.
We will only label with the incoming source above the main breaker. Until such time as more definitive testing is completed, we will not take credit for this breaker.
The one action we did take is : discussion on breaker settings, coordination and re evaluating trip settings of our switchgear upstream.
We have already evaluated some bus ways also. The good news is we have implemented some fairly significant reductions in incident energy calculation.
The first finding and adjustment was the toughest. Nobody wanted to start changing settings.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: MCC has proper segregation?
PostPosted: Sat Jul 22, 2017 4:31 pm 
Plasma Level
User avatar

Joined: Tue Oct 26, 2010 9:08 am
Posts: 2174
Location: North Carolina
I treat each section like a panelboard...the section with an incoming cable (main lug or main breaker) is all treated the same. The other sections are treated as downstream of the main breaker. It's sort of anyone's guess right now but with some evidence of "jumping" breakers that's the best you can do that I know of.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: MCC has proper segregation?
PostPosted: Mon Jul 24, 2017 9:21 am 
Sparks Level

Joined: Fri Apr 15, 2011 7:43 am
Posts: 177
Location: Colorado
We don't always know the construction of the MCC and how it will react to arcs and plasma and.... We always treat the MCC with the possibility of propagation. We also do not know if the MCC may get reconfigured (buckets moved/replaced) and an incorrect sticker gets applied. We always treat any MCC or panel as if propagation may occur unless they are fed from an separate enclosure.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: MCC has proper segregation?
PostPosted: Tue Aug 01, 2017 8:13 am 
Sparks Level

Joined: Sat Nov 21, 2009 10:23 pm
Posts: 105
Location: Ohio
The line between practice and reality is blurred on this topic. It is highly doubtful that any MCC manufacturer will agree that there is anything but one IE number for the entire MCC. I have tried this multiple times and in each case the manufacturer falls back to the one number for IE.

The question then becomes, if you use a second value below the main OCPD, who is liable in the case of an accident if the manufacturer has gone on record that the entire MCC took on one value.


Top
 Profile Send private message  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 14 posts ] 

All times are UTC - 7 hours


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Jump to:  
© 2019 Arcflash Forum / Brainfiller, Inc. | P.O. Box 12024 | Scottsdale, AZ 85267 USA | 800-874-8883