What assumptions are engineers making for utility arc flash studies for OH three-phase transformer bank secondaries?

(1) Do you automatically assume that the flash event will escallate into a three-phase event given the possible spacings for the physical XFMR installations?

(2) How do you treat unbalanced banks?

(3) Does the secondary connection (wye or delta) affect the results?

(4) Do you count on "local" XFMR protection fuses to clear? What if the fuses are sized differently for each XFMR?

I have to tackle these calcs shortly as part of our study. I'm curious to know what others are doing.

Thanks

What assumptions are engineers making for utility arc flash studies for OH three-phase transformer bank secondaries?

(1) Do you automatically assume that the flash event will escallate into a three-phase event given the possible spacings for the physical XFMR installations?

It might be conservative to do so, and also able to be calculated with existing software.

(2) How do you treat unbalanced banks?

Assume three of the larger, so it could be calculated with existing software. It is assumed this is fuse protected, and the percentage impedance similar.

(3) Does the secondary connection (wye or delta) affect the results?

Software will ask for this and frankly I don't remember if there is a diffferent result.

(4) Do you count on "local" XFMR protection fuses to clear? What if the fuses are sized differently for each XFMR?

Software will typically determine what is the clearing device if modelled accurately. Typical software will not allow different fuse types so I would try the largest fuse and the largest transformer.

I have to tackle these calcs shortly as part of our study. I'm curious to know what others are doing.

We have worked with ETAP software which I am comfortable is as good as any and better than some. But there are still shortcomings: recently we needed to model plug in busway that is not within the library of conductors so we fudged a bit by using equivalent cables. It is good to be familiar with the disclaimers within NFPA70E that indicate even the most comprehensive study will not avoid all injury, but should reduce injury. Also it is good to be aware that most engineering studies start with an assumption and even though 20 steps later it might appear to be black and white calculations, much is determined by the initial assumption. It is good to verify what the consequences of the initial assumption are.

"Thanks" Have fun

I have done so, mainly because the existing analysis methods covered in the industry standards use 3Ø faults. At the transformer, this is not practical because of the phase spacing, but the secondaries are made up into a service drop at some point where they are close together.

If haven't run into this, but I think assuming a bank of three of the largest transformers would be acceptable.

Not if you consider only 3Ø faults. This does bring up a potential problem with only looking at 3Ø faults. The primary current in a delta-wye bank for a single phase fault on the secondary will be only 58% of the referred secondary current. This may make clearing times much longer for 1Ø faults. 1Ø fault currents will also be higher than 3Ø for a delta-wye bank.

If you assume all transformers the same size, then all fuses would be the same size. You might want to do detailed hand calcs of one instance to see if assuming all one size is conservative or unconservative.