We have an available fault current at the service entrance of roughly 135 kA (nice) and it's causing the following Note (*N) on the main switchboard:

(*N_ka) - Out of IEEE 1584 kA Range. Lee Equation Used. Consider NESC/EPRI method to compare

There IS an option for "EPRI Report 2011" under the Arc Flash Study Options (see below), but I am hesitant to use this method as I don't know what the results mean, and something seems REALLY off.


For Example, it changes the Incident Energy from roughly 317 cal/cm^2 to almost 35,500,000 cal/cm^2



I know I'm missing something, so hit me with it!

Not sure I am going to be able to help as I have not run a study with that high of available fault current. Is this part of a spot network with multiple parallel transformers? In any case, it looks like your bus voltage is 480V and the NESC and EPRI formulas are for over 15kV per the SKM setup screenshot you inserted so I don't think that is an option. Not sure why SKM would recommend those options. I don't think you can use 1584 or Doughty because your fault current is too high. According to NFPA 70E Table D.1, the Lee equations are for open air. However, I am not sure about that. There are some old threads on this forum about the Lee equation like this one: viewtopic.php?f=8&t=3590

Here is the limitations table from 70E:



Can you post a screenshot of your service up to and including switchboard? A little more context may help. Sorry I don't have any answer for you.

Ah, I just noticed in your EPRI method screenshot you have selected AIR so maybe the Lee method is the best you can do. It would still help to know a little more about the distribution system and what exactly you plan to label. Is a label even required at the COMBINED UTILITY BUS?

You nailed it. 4 paralleled 3-Meg transformers. I'm not sure I'm allowed to share much more than that.

And thanks for the table, didn't know that was there. I was piecemealing it together.
And thanks for always responding with a thoughtful answer, it is much appreciated.

You're welcome.

There was a very large and costly arc blast event that you may be aware of that happened about 3 or 4 years ago in Muncie (not BSU). I was asked to investigate it to help determine the cause. I wish I could share the surveillance video and other information but the lawsuits prevent that. I highly doubt that one would have survived if they had been in the electrical room at the time.

In any case, that building was fed from two paralleled substation transformers in a vault. Spot network configuration. The collector bus in the vault had available high fault currents, as did the downstream main switchboard in the building, not nearly as high as yours as the transformers were much smaller and there were only two. The power company electricians referred to the collector bus as the crab or spider as I recall. I had not heard those terms for that before then, or since.

If yours is a new design, or if you need to make recommendations on an existing installation, I would highly suggest installing cable limiters/cable protectors on each end of each cable between the transformers and the collector bus if they don't already have them. Spot networks are basically what they are for as an individual cable fault will clear quickly and not take down the network so the building or buildings stay online. Those may get your incident energy down as well. They won't help your available fault current though.

https://www.eaton.com/content/dam/eaton ... miters.pdf

https://us.mersen.com/en/products/cable ... r-limitors

https://www.littelfuse.com/assetdocs/lf ... 8e2ba0b13f

I know the incident you are referring to, and I have seen the video. IIRC it blew the doors off the electrical room.



It's not new, but we did recommend using cable limiters. Which, while we're on the subject what method do you use to model cable limiters in SKM?

Model as a fuse...cable limiters are in the library.

Mike

What Mike said although I seem to recall that there is a check box on some screen somewhere that says Cable Protector but I don't recall now where what screen or what checking the box does. I don't run into those often.

Found it. The cable protector check box appears when you have parallel sets of cable for a feeder. Checking the box will divide the fault current between the number of parallel sets, so the fault current "seen" by the cable protectors will be a half or a third, or a fourth or whatever the number of parallel sets you have and do the math for clearing based on that.

Thanks Mike, so then would you typically model the paralleled cables individually as well? For example, we have 5 parallel #600 kcmil sets to make a 2000A feeder. Each wire gets a 400A cable limiter; however, it seems wrong to use a 400A fuse in SKM unless they are taking that into account.

Thanks Brent,
I do wonder how that tripping mechanism would play out in reality. Like I can see a scenario where the load end of the only half of the wires gets shorted; it doesn't seem like all of the cable limiters would trip in this scenario. If that is the case, there would only be INCREASED resistance (not the infinite resistance we see in an open circuit). If that's true, then dividing the available fault current at a given point does not follow the Kirchoff Current Law. The Current doesn't go from V/(5R) to V/0 (infinite, open circuit), it goes from V/(5R) to V/(4R), then V/(3R) if that cable limiter blows at all.

Guess I need to do more research on these...

On my phone so short answer. The limiters are not sized by amps. They are by cable size. They are pretty much just short circuit protection for a single cable so if you have a cable fault it takes that cable out very fast and your system stays online with a little less ampacity. If you look at the TCC for a #500 limiter it would have like a 2000 or 2500 amp long time rating if it had such a rating. So no overload protection just short circuit protection.

I just went back and looked at an old job where I had compared #500 cable limiters to a Eaton Magnum DS limiter installed in the gear in series with the breaker and a fuse. It looks like my memory failed me. If the #500 cable limiter had such a thing as a long time rating, it would be around 1000 amps, not 2000-2500a that I mentioned yesterday.