It is currently Sun Aug 20, 2017 2:42 am



Post new topic Reply to topic
Author Message
ekstra   ara
 Post subject: Weird Analysis Calculations
PostPosted: Tue Mar 07, 2017 1:02 pm 
Offline

Joined: Tue Mar 07, 2017 12:49 pm
Posts: 5
Hi all,

I am looking for some input on an odd incident energy value I received while running an analysis recently.

The system has one main switchboard is entirely powered by generators.

There are 3 different operation modes, 2 of which return generally low incident energy levels. The 3rd mode involves the main switchboard being fed by 3 parallel generators.

When I run my analysis during this operational mode the arc fault current is so low (1.7kA) that the fault clearing time is near 8 minutes which ends up producing an IE of around 4000 cal/cm2. There are coordination factors that don't allow for the feeder breakers to be adjusted to the point where they make a significant difference.

There are many factors which make me think this number is a bit off the deep end.

The Switchboard is 380V. energy levels throughout the system are mostly 0-6 cal/cm2. The IE at the board in other modes is at 6 cal/cm2. I am at a loss for how I end up with a number that high in the 3rd mode. (Besides the 8 minute fault clearing time)


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Weird Analysis Calculations
PostPosted: Tue Mar 07, 2017 1:49 pm 
Offline
Sparks Level

Joined: Fri Jul 08, 2016 10:01 am
Posts: 119
Is that even enough arcing current to sustain an arc at that voltage?


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Weird Analysis Calculations
PostPosted: Tue Mar 07, 2017 2:33 pm 
Offline
Arc Level
User avatar

Joined: Wed May 07, 2008 5:00 pm
Posts: 660
Location: Rutland, VT
What software are you using to analyze?

What are the other 2 modes of operation?

_________________
Barry Donovan, P.E.
www.workplacesafetysolutions.com


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Weird Analysis Calculations
PostPosted: Wed Mar 08, 2017 5:29 am 
Offline

Joined: Tue Mar 07, 2017 12:49 pm
Posts: 5
bbaumer wrote:
Is that even enough arcing current to sustain an arc at that voltage?


That was my first question when I saw the results. I am sure the math is correct, but would this actually happen in real life situation? I haven't been able to find much literature on how to tell if it is.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Weird Analysis Calculations
PostPosted: Wed Mar 08, 2017 5:38 am 
Offline

Joined: Tue Mar 07, 2017 12:49 pm
Posts: 5
wbd wrote:
What software are you using to analyze?

What are the other 2 modes of operation?


Thanks for the reply,

I am using ETAP V.16.0

The other two modes are as follows:

1) Fed from a single 1000kW Generator. In this case I get a higher fault current (approx. 4kA) and the IE ends up at about 6 cal/cm2.

2) Fed through a400 kVA transformer from a 12.47kV utility source, here I see a similar IE to mode 1 (5.7 cal/cm2)


My thought is there is no way possible I could ever see 4000 cal/cm2 in the 3rd mode for a lot of different reasons.
The arc might not be sustainable at that arc fault current level and if it was there is no way the switchboard would still be there after 8 minutes being two reasons that stick out to me.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Weird Analysis Calculations
PostPosted: Wed Mar 08, 2017 6:33 am 
Offline
Arc Level
User avatar

Joined: Wed May 07, 2008 5:00 pm
Posts: 660
Location: Rutland, VT
I don't know if ETAP has the capability of doing an Integrated analysis. There is the option in EasyPower to do this which for a case where there are multiple, in your case 3 gensets, the software looks at the incident energy for the initial fault, looks at which genset will trip first, then calculates the IE for the other 2 gensets, removes the next one to trip, looks at remaining IE and time and adds all together for one result.

The other thing I just noticed is you are running the fault out until it totally decays which is 8 minutes. What happens if you use a 2 min cutoff?

_________________
Barry Donovan, P.E.
www.workplacesafetysolutions.com


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Weird Analysis Calculations
PostPosted: Wed Mar 08, 2017 7:06 am 
Offline
Sparks Level

Joined: Fri Jul 08, 2016 10:01 am
Posts: 119
wbd wrote:
....What happens if you use a 2 min cutoff?


Do you mean 2 seconds?

I thought that was the standard max trip time override. The thinking being the worker will either be blown away or can get away from the event within 2 seconds.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Weird Analysis Calculations
PostPosted: Wed Mar 08, 2017 7:37 am 
Offline

Joined: Tue Mar 07, 2017 12:49 pm
Posts: 5
wbd wrote:
I don't know if ETAP has the capability of doing an Integrated analysis. There is the option in EasyPower to do this which for a case where there are multiple, in your case 3 gensets, the software looks at the incident energy for the initial fault, looks at which genset will trip first, then calculates the IE for the other 2 gensets, removes the next one to trip, looks at remaining IE and time and adds all together for one result.

The other thing I just noticed is you are running the fault out until it totally decays which is 8 minutes. What happens if you use a 2 min cutoff?


I'm not sure what ETAP's capabilities are for this situation either. I am going to take a look into what I can do with it. This is the first time I have run into a situation with this many generators feeding the same bus. The 3 LVCB's have different trip settings, and when I place a fault on the switchboard bus I can see the trip order of the breakers but I think the analysis is just taking the total clearing time into account and not the integrated analysis like you speak of.

I think I can try to the analysis with cutoff and see what I get.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Weird Analysis Calculations
PostPosted: Wed Mar 08, 2017 8:40 am 
Offline
Arc Level
User avatar

Joined: Wed May 07, 2008 5:00 pm
Posts: 660
Location: Rutland, VT
Ok, be sure to post your findings.

_________________
Barry Donovan, P.E.
www.workplacesafetysolutions.com


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Weird Analysis Calculations
PostPosted: Wed Mar 08, 2017 9:06 am 
Offline
Sparks Level

Joined: Fri Jul 08, 2016 10:01 am
Posts: 119
Not sure how to do this in ETAP anymore but again, if you can, change your max event duration or max trip time to 2 seconds. That is what SKM defaults to:

Attachment:
2 seconds.JPG


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


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Weird Analysis Calculations
PostPosted: Wed Mar 08, 2017 9:32 am 
Offline

Joined: Tue Mar 07, 2017 12:49 pm
Posts: 5
wbd wrote:
Ok, be sure to post your findings.


I was able to reduce the max fault clearing to 2 seconds.

I now have an IE of 50 cal/cm2 at the switchboard which is high but at least looks better than 4000.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Weird Analysis Calculations
PostPosted: Thu Mar 09, 2017 9:08 pm 
Offline
Plasma Level
User avatar

Joined: Tue Oct 26, 2010 9:08 am
Posts: 1736
Location: North Carolina
Something doesn't add up here. You are saying that you get a higher fault current with all 3 generators running compared to only 1? That doesn't make any sense.

Another thing that doesn't make sense here is that a relatively low arcing current of 1.7 kA is generating a 50 cal/cm2 incident energy at 2 seconds. This doesn't sound right because usually I see crazy high incident energy at 2 seconds with much higher fault currents. If you know arcing current, can you do the IEEE 1584 empirical equation by hand and post the results and/or hand-check the results? Sounds like something's not right here.

Typically with multiple sources (e.g. generators) the highest incident energy is seen with the highest impedance source because it has the lowest arcing fault current and thus the longest trip time. But this scenario is almost always one generator with the utility offline, not all 3 generators vs. a single generator. Something just doesn't add up here. Again, check your math very carefully because intuitively it just doesn't make a lot of sense.

From experience using power system analysis software whenever you seem to be getting results that don't make any sense, resort to spreadsheets and/or hand calculations and check the answers carefully. Errors do pop up frequently either because you are using some kind of "partial update" mode instead of doing a full recalculation and since there is some there are iterative calculations going on, it is pretty easy to corrupt them during normal development work. The second problem is when someone configures a parameter incorrectly or misses changing something from the default setting, causing all kinds of silly corruption. So if it looks wrong to you, it probably is.


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Weird Analysis Calculations
PostPosted: Mon Jun 05, 2017 8:23 pm 
Offline

Joined: Mon Jun 05, 2017 8:07 pm
Posts: 2
@ bbaumer ETAP does have capabilities to manually enter the Fault Clearing Time (FCT) to any value that is required by using the following options

1 . "Fixed FCT" from the bus editor -> Arc Flash page -> User-defined section
2. Open up the study case -> Clearing Time -> and select the "Limit Maximum FCT" option and enter value of 2 or above.

The first option is meant for individual buses or equipment and the second option is to limit the maximum Fault clearing time by 2 seconds. So For example, If your FCT is 480 seconds for example, with the second option set, the incident energy will be calculated based on the FCT limited to 2 seconds.


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


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Weird Analysis Calculations
PostPosted: Mon Jun 05, 2017 8:31 pm 
Offline

Joined: Mon Jun 05, 2017 8:07 pm
Posts: 2
@PaulEngr

The calculation of incident energy is not only based on the magnitude of the currents but also the clearing time of the protective devices. In case of Multiple sources feeding a fault location , simulation software like ETAP can provide calculation of incident energy based on stages of time . This can be accomplished using the following option

1. Go to Tools-> Options (Preferences) -> Arc Flash and set "Subtraction of Incident Energy for Multiple Source Systems " = "True"
2. If this entry is True, then the program removes incident energy contribution from each source branch at the time its protective device opens to de-energize the arc fault. The total incident energy by assuming that the incident energy contribution from each source branch is present up to the time that the last source protective device de-energizes the fault

Thank you


Top
 Profile Send private message  
Reply with quote  
 Post subject: Re: Weird Analysis Calculations
PostPosted: Wed Jun 07, 2017 6:33 am 
Offline
Plasma Level
User avatar

Joined: Tue Oct 26, 2010 9:08 am
Posts: 1736
Location: North Carolina
I understand that software is the way to go in most cases and not disagreeing with the idea that opening time matters and obviously it does. The issue though is that the underlying math is effectively a sparse matrix impedance calculation with two exceptions. First it is in this case driven by multiple sources so we don't have a simply Thevinin equivalent circuit. Indeed actually if the simulation actually does piecewise analysis (I forgot how E-Tap does it) it can get pretty complicated. The second issue is that arcing currents are a nonlinear function of available fault current so it takes a couple iterations (usually 3-4) before things settle out. No need to resort to complicated Newton, steepest gradient, etc...it settles out pretty quickly on it's own with good old "plug and chug".

The problem though is that as you add nodes, the size of the relatively sparse matrix quickly grows in size. Once you get to 1,000 nodes or more for instance it can take several minutes even on a high end machine to actually solve. Fortunately most of the time you only have to do a full run once, to a point. After that as you make changes the software only has to do incremental updates on a small part of the matrix, ASSUMInG that essentially that current flows only in one direction (from source to load) and that very little change occurs in the other direction (from inductive loads). This assumption holds most of the time, so maybe only a dozen nodes need to be updated and you can simply do incremental calculations on the submatrix. HOWEVER the errors remain and over time they can grow substantially.

So quite often what happens is that as you work with a given model and keep using the incremental/windowed update mode, everything seems fine. But then over time you keep getting more and mroe nonsensical answers. This is the point where a full recalculation is needed which fixes all those accumulated errors.

The second problem is that there are a lot of settings and parameters everywhere. It is really easy to put in something user defined or fat finger an entry somewhere and all of a sudden you start to notice things that don't look right. So whenever I'm doing arc flash modeling, I always run several reports where I'm looking at say highest incident energies or lowest incident energies and carefully look for what look like discrepancies because chances are something isn't right somewhere. A good example of this for instance is what arc flash calculation you are looking at (upstream, downstream, etc.). For instance you might be looking at say the bus on say a feeder breaker in some switchgear and see some crazy high result that happens because it is looking at the upstream (before the main) breaker instead of the correct value, or the tie and both mains are assumed closed but it is Kirk keyed or otherwise configured so that no more than two breakers are closed at a given time, or even that the plant practice is to always run with open ties (or the tie might be a physical device...just a splice box...so it can't be tied except very manually).

1700 A sounds more like a normal/overload situation especially on a 1000 kW generator where the normal maximum output is around 1200 A. Sounds like a math error has crept in somewhere.


Top
 Profile Send private message  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 15 posts ] 

All times are UTC - 7 hours


Who is online

Users browsing this forum: No registered users and 1 guest


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:  
cron
© 2017 Arcflash Forum / Brainfiller, Inc. | P.O. Box 12024 | Scottsdale, AZ 85267 USA | 800-874-8883