It is currently Fri Aug 14, 2020 4:15 am



Post new topic Reply to topic
Author Message
ekstra   ara
 Post subject: Generator Modeling
PostPosted: Tue Jan 22, 2008 10:31 am 

Joined: Thu Jul 12, 2007 4:27 pm
Posts: 5
This one is puzzling me. Do people model (care about) smaller emergency generators and arc flash. I see this as a complicated issue. Based on the generator size and subtransient reactance, fault currents tend to be low however due to the quick decay of the fault current, now there is the question about the device actually tripping i.e. creating a sustained fault. Then again, if it did sustain, it is a collapsing fault current so it does not produce much energy. For smaller units, do we even need to model this and if so, how detailed do people get. Do we need to get machine constants for small units (which often is difficult to obtain?) - Thanks for any input.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Tue Jan 22, 2008 9:21 pm 
Arc Level

Joined: Thu Jan 10, 2008 8:49 pm
Posts: 498
Location: New England
Wouldn't it be common to have a Main Breaker on a generator. IF so, then use that breaker curve for clearing time, and use the x' of the generator to calculate fault current on the main.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Fri Jan 25, 2008 8:06 am 

Joined: Thu Jul 12, 2007 4:27 pm
Posts: 5
Thanks Haze,
I guess my question was more along the lines of what size of generator no longer matters for arc flash and when can you ignore decrement detail on larger ones.

When performing short circuit studies it is common to ignore the study below very small 208V transformers since you reach a point where the fault current is below 10 KA - lowest device rating.

Arc flash studies sometimes stop at transformers less than 125 kVA i.e. 112.5 kVA below 240V for similar reasons.

That got me thinking about generators. If you have a small generator, has anyone come up with a lower limit where you don't model it? But then the question becomes, if it is a small unit, will there be enough fault current to trip the breaker? Where that question came from is if it has a subtransient reactance of let's say 15%, the Generator Short Circuit Current = Full Load Amps / 0.15 This means the maximum short circuit current is approximately 7.5 X FLA which can be below the instantaneous trip of some breakers.

So now, is the low fault current considered dangerous becuase the device in theory may not clear, or is it recognized that the field will collapse (transient and synchronous reactance), or perhaps the arc just can't sustain itself, or do you assume it sustains the subtransient portion of the fault and cut it off at 2 seconds. I am now wondering how the software vendors handle this as well.

I think the problem here is I began thinking about this too hard!

I would be interested to know how others have been dealing with this.

Thanks!


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Fri Jan 25, 2008 8:57 pm 
Arc Level

Joined: Thu Jan 10, 2008 8:49 pm
Posts: 498
Location: New England
No, I think you are bring some good thoughts to the forum. With accountants and lawyers running manufacturing today, we all have grown to point of reaching beyond what is practical. I, along with every electrician I know, and I know over 30 of them, think the whole 70E has gone to far.

Let me ask you this. On a generator this small, can you look up the clearing time on the breaker for 7x fault current, then take that clearing time and the fault current and put it in one of the online free Incident Energy Calcs. If the answer comes out less than 1.2 cal/cm2 - then you don't need any PPE. Its hard to say if there would be a magic threshold value where you don't have to calculate below a generator of x size, as the clearing times may change and they will have the bigger influence in this situation.

But if you care to do the first step and report back what you find it would be a learning lesson for us all.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Tue Jan 29, 2008 9:26 am 

Joined: Thu Jul 12, 2007 4:27 pm
Posts: 5
I have a 300 kW generator at 480V with a FLA of 451 Amps. There is an in line breaker rated 500A. When the trip is set on the high side for coordination with downstream breakers, there is not enough bolted short circuit current to trip it. Using a 450A breaker has the same problem. This is with bolted fault current and the lower arcing current would be much worse. Using one of the computer programs it does not solve the problem but rather shows a default cut off of 2 seconds (from IEEE 1584) and lists it as a 2* (from NFPA 70E Table). This seems to be one of the cases that falls through the cracks. We can go with 2* based on the tables, except the NPFA footnotes give a maxium time duration which we would greatly exceed. The practical solution is only work on it when it is not running which is typically the case or crank the breaker setting down if we had a rare case of needing to work live but I would still question if it would trip for the low arcing current at the low setting. We began digging into this particular scenario but perhaps we are digging a little to deep. Thanks for any comments.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Tue Jan 29, 2008 11:16 am 
Sparks Level

Joined: Thu Jul 19, 2007 7:54 am
Posts: 201
Location: St. Louis, MO
I've read this thread, but quite frankly, I'm not sure what the answer would be. I think this would be similar to motor contribution, as talked about here: http://www.arcflashforum.com/showthread.php?t=37

Basically, to approximate the decrement curve, you could include the full generator contribution for the first 6 cycles, then regress to full load current after that for the remainder of the time, or 2 seconds. This would be a very rough step calculation approximation of the generator contribution, regardless of size. SKM currently offers this method.

Kyle K wrote:
We began digging into this particular scenario but perhaps we are digging a little to deep.

Nah, that's how we advance this fledgling field, and in my opinion, what makes it interesting, and is also what this forum is for.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Wed Jan 30, 2008 7:29 pm 
Arc Level

Joined: Thu Jan 10, 2008 8:49 pm
Posts: 498
Location: New England
The prime mover on the generator, diesel, gas engine, gas turbine, etc, is sized very specific to the generator. At full generator load you pretty much are at full engine load, with very little engine HP left. So when the fault occurs, you are really working off inertia. How long can the generator produce bolted fault current, I'm not sure, but probably only a couple seconds, then the engine is going to bog down to whatever it is capable to produce in hp and that it.

Does anyone have a feel for how many cycles or seconds you would get fault current before the engine stalls.

After that point, when the engine stalls, and you go back to FLA, my gut feeling is that the arc would extingish.

There is a good dicussion and I would like to explore it further.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Thu Jan 31, 2008 12:48 pm 
Sparks Level

Joined: Thu Oct 18, 2007 7:10 pm
Posts: 261
Location: NW USA
Although much of this can be modelled, and we've done so on larger machines with published transient response; for a standby or emergency generator we have adopted the position that this is not a normal configuration and we don't calculate exposure for other than normal conditions.

It would seem unlikely that switchgear internal maintenance needs to be performed just the one day of the year that the emergency generator is running.

By avoiding all the different possible abnormal conditions, much confusion has been saved. I have yet to find maintenance staff (the actual guys doing the work) that would prefer multiple calculated exposure based on differing configuration.

We've adopted the same philosophy on alternate feeders, looped systems etc. First put the system back to normal before working on it.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Thu Jan 31, 2008 4:18 pm 
Arc Level

Joined: Thu Jan 10, 2008 8:49 pm
Posts: 498
Location: New England
I'd like to see if the PPE/Task table would still apply when the generator is on. It could be necessary to perform voltage tests, or even change out a MCC bucket.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Fri Feb 01, 2008 2:20 pm 
Sparks Level

Joined: Thu Oct 18, 2007 7:10 pm
Posts: 261
Location: NW USA
haze10 wrote:
I'd like to see if the PPE/Task table would still apply when the generator is on. It could be necessary to perform voltage tests, or even change out a MCC bucket.


It should not be more difficult than any othe arc flash calculation, to enter the generator parameters and come up with a number. On the systems I've worked the generator supplied lower fault current but still had enough to trip instantaneous (when reduced by arc impedance) so the resulting arc flash exposure was lower, but that should be checked prior to assumption.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Tue Feb 05, 2008 9:27 am 
Plasma Level
User avatar

Joined: Mon Sep 17, 2007 5:00 pm
Posts: 1503
Location: Scottsdale, Arizona
Generator Time Current Curve

Hi Everyone,

This is a great discussion - lots of good comments and thoughts. I attached a time current curve that hopefully helps the cause. It looks like the generator MCCB with a high setting would probably not trip according to the generator decrement curve. If the fault current is arcing, the liklihood of the breaker tripping goes down although at lower fault currents, the difference between bolted and estimated arcing currents shrinks. So now the question is:

1) Could the arcing current sustain itself?
2) If it does sustain itself, would it last as long as it takes for the breaker to trip according to its curve (I doubt it)
3) Do you cut the time off at 2 seconds (makes more sense but could give high HR category using 120 cycles)
4) Do you use the 70E Tables which have limits on the arc duration in the footnotes that would be exceeded

I'm not sure there is a right answer yet (I'm not sure if that means there aren't any wrong answers :cool: ) This seems to be another one of those cases where we are trying to get precise with a piece of the puzzle while the main IEEE 1584 equations aren't that precise (yet)

And..... to make this more exciting (??) the NEC requires that the emergency and legally required systems must coordinate with the supply overcurrent devices i.e. generator breaker would ususally be set high for coordination - :eek: Augh!

I am interested in anyone's comments and thoughts - this is a good one!


You do not have the required permissions to view the files attached to this post.

_________________
Jim Phillips, P.E.
Brainfiller.com


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Tue Feb 05, 2008 1:10 pm 
Sparks Level

Joined: Thu Jul 19, 2007 7:54 am
Posts: 201
Location: St. Louis, MO
brainfiller wrote:
1) Could the arcing current sustain itself?
2) If it does sustain itself, would it last as long as it takes for the breaker to trip according to its curve (I doubt it)
3) Do you cut the time off at 2 seconds (makes more sense but could give high HR category using 120 cycles)
4) Do you use the 70E Tables which have limits on the arc duration in the footnotes that would be exceeded


1) I don't think the arcing current would sustain itself. I'm hopeful that more studies will substantiate this. There have already been some studies that suggest that most LV arcs will not sustain for more than 2 seconds.

2) See 1) above and 3) below.

3) A 2 second cut off would give overly conservative results. I try to make my calculations as accurate as possible. This increases electrician buy in and gives the program more credibility if something does go wrong and a flash occurs. This could also result in workers being overly dressed and unable to perform their jobs effectively. I still go with the 6 cycles at full arcing current, with a subsequent reduction to 100% rated current. In place of that, 10-30 cycles may give a close approximation without going too far.

4) The tables cannot be used in this instance. They simply fall outside the scope. Outside of that, the alternative is to have a mixed :( system of calculations and tables. (In this case, for the same voltage levels.) A bad combination sure to bring your program into question.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Tue Feb 05, 2008 7:56 pm 
Arc Level

Joined: Thu Jan 10, 2008 8:49 pm
Posts: 498
Location: New England
Small generators like this generally use molded case circuit breakers which are 80% rated. So a 500A breaker should not be loaded to more than 450A, which matches the generator output FLA. Installing a 450A breaker would shift the curve to the left and increase probability of tripping, but you would have to carefully monitor continuous load conditions and therefore not be able to use full capacity of the generator. This could be overcome by changing to a 100% rated breaker at 450A. You would stand a good chance of being in the instantaneous range and tripping in 2 cycles then.

I don't have the backgound of knowing arc sustainability, but my gut feeling is that sustaining an arc at 800A after 1 sec seems very unlikely, especially at typical gap distance.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Wed Feb 06, 2008 5:52 am 
Plasma Level
User avatar

Joined: Mon Sep 17, 2007 5:00 pm
Posts: 1503
Location: Scottsdale, Arizona
brainfiller wrote:
I had a copy and obtained approval from it's source at PG&E. It was presented at one of our IEEE 1584 meetings about a year ago. It confirms what Catcher stated. There are a few graphs of data points as well. It shows that we are still learning the physics of arc flash and have a lot of work ahead of us. You can download it at:

[url="http://www.brainfiller.com/documents/PGETestingbrainfillerposting.pdf"]PG&E Presentation[/url]

I added an intro page on it just so people know what it is based on and to use it at your own risk.


I agree with not being able to sustain the arc. The question I posted earlier was more rhetorical just to see what everyone else thought. Above is the info I posted a while ago about arc duration testing that was presented at one of the 1584 meetings. WDN, I think the shorter clearing times you are using make sense - maybe someday we'll have some solid "official" direction but for now, it's great to see everyone sharing their views to work through this. Haze, you are right, the 450A breaker would stand a better chance of tripping than the 500A under these conditions but as you correctly pointed out, it is ususally only 80% rated.

_________________
Jim Phillips, P.E.
Brainfiller.com


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Wed Feb 06, 2008 9:11 am 

Joined: Mon Jul 23, 2007 6:49 am
Posts: 19
It looks like with all the different info thoughts and ideas being posted, it was OK that I was confused. :eek: Thanks everyone!


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Wed Feb 06, 2008 8:34 pm 
Arc Level

Joined: Thu Jan 10, 2008 8:49 pm
Posts: 498
Location: New England
Jim,
Why does the Generator Decrement curve divide into two branches at the bottom? Are you trying to show min an max values?


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Thu Feb 07, 2008 1:57 pm 
Plasma Level
User avatar

Joined: Mon Sep 17, 2007 5:00 pm
Posts: 1503
Location: Scottsdale, Arizona
Haze, great question. I probably should have labeled it on the graph. The branch to the left is the AC current from the generator only. The branch to the right is the total RMS of the AC plus DC current. The DC component of the fault collapses pretty quickly - less than 1 second.

The total current could possibly trip the instantaneous of the breaker but I doubt it. The arcing value for this current (which I did not calculate) would be even less. The vertical band of the breaker's instantaneous represents the plus and minus breaker tripping tolerance. At best the bolted fault current just passes the minimum amount of current it takes to trip the breaker and the right side of the band means it could take a lot more current - not to mention the arcing current would be less. The curve depends on the generator parameters and I just used some typical numbers.

_________________
Jim Phillips, P.E.
Brainfiller.com


Top
 Profile Send private message  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 17 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