Arc Flash Forum
https://brainfiller.com/arcflashforum/

Multiple Circuits in One Junction Box
https://brainfiller.com/arcflashforum/viewtopic.php?f=23&t=2519		 Page 1 of 1
Author:  engrick [ Mon Nov 05, 2012 3:35 pm ]
Post subject:  Multiple Circuits in One Junction Box
Anybody have an idea on how to model multiple circuits in one junction box? What we have is a remote junction box with three circuits of terminal blocks, each is a three phase circuit fed from different breakers. We are concerned about propagation and not worst-case.

I do not know of any studies done on aggregate incident energy - if anyone has seen any I would love to see something on it. It seems there is the possibility of a combined incident but no simple solution.
Author:  PaulEngr [ Sat Nov 10, 2012 5:39 am ]
Post subject: 
Propagation does indeed happen. Actual tests on arc flash prove this. However, there are several factors that may make your job harder or easier.

First, if they are all fed from the same source then you could just model the impedance of the cabling itself as a parallel feeder situation which will reduce the impedance. Unfortunately where this will get tricky is that most modelling software will also model the breakers as tripping early due to the higher incident energy that it is assumed the breakers will see. So you may have to manually calculate the resulting cable impedance and feed that info into the software if you are using software. Second issue is that you'd have to consider what happens depending on the order that breakers trip if they have different settings.

If they have different sources then you have to consider that this means effectively that you have parallel feeders coming from whatever bus is the actual common source and model the transformers and wiring in between as impedances leading back to the common source.

Third issue is propagation itself. This is not instantaneous. If you search around through some of the publicly available arc flash studies you'll find that it takes a while before you get to a 3 phase fault. So depending on how fast you are tripping and the available fault current you might never get to a propagation issue. IEEE 1584 current assumes that a full 3 phase fault occurs and does not give any credit for time delays in developing a 3 phase fault. This is even more important in outdoor open wire conditions where 3 phase faults are almost nonexistant in many cases.

So as you can see, you're asking the right question and you can model an "instantaneous" case where your entire junction box turns into one giant arcing slag pile but if you even start to think about propagation, you open a can of worms that has yet to be solved.
Author:  engrick [ Sat Nov 10, 2012 2:54 pm ]
Post subject: 
PaulEngr, thanks for the post. You echoed my thoughts and concerns, never thought about the time delay that way

We are using SKM for our study and have though about paralleling the feeders into a single bus but were concerned about the tripping time as you indicated. We only have one utility connection but two different paths for each of the circuits with two upstream breakers in one path and three in the other.

I think this the part of arc flash that has been over looked and will continue for some time. The days of combining multiple circuits in one can is probably over. Any more I keep circuits separate just to avoid issue with arc flash like this one. Unfortunately there is more plants pre-arc flash than post.
Author:  jvrielink [ Sun Nov 11, 2012 3:37 pm ]
Post subject: 
Why not put a very small impedance between the two buses and include a scenario where they are connected? This should properly model the case of an arc flash with both sources contributing in terms of protective devices (although not taking into account the delay). That way your model still makes sense, and you simply take the worst case for labeling the equipment.
Page 1 of 1 All times are UTC - 7 hours
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/