Choosing switchgear for an arc flash test

[Chenopod]

Hi, I've been asked to spec out switchgear enclosures to do arc flash testing per IEEE C37.20.7. We know the 'arc rating' parameters we plan to use: 5kV, 50kA (or more), 30 cycles @ 60Hz. The arc will be induced by a copper shorting wire between phases on the main bus, at a specialized test lab. No breakers will be installed. We will use pressure venting to release most of the blast energy and we hope to re-use the switchgear several times for our "practice runs" and then use a fresh enclosure for the actual test. It is OK if the bus bar fails, we will buy and bring extra copper blocks. I need help spec'ing the operating voltage and current of the switchgear based on the arc rating I mentioned above. I am less concerned about the mechanical construction of the cabinet, more interested in advice on the electrical specs. For instance, can we get away with a 480V, 1600A system to do this testing? Or would we need a medium-voltage system such as 5kV, 1200A? Or does it really not matter because they're all going to blow to smitherines? Thank you!

 

[Zogzog]

I would imagine you would want to use 5kV rated gear for a 5kV arc test. You want to get as close to real world conditions as possible. Different gaps and designs.

What are you doing this for? I did some similar testing like this a few years ago researching retrofits of older gear to meet the ANSi arc rated criteria, not easy to accomplish.

 

[Overvoltage]

I think EPRI has done quite a bit of this testing in recent years. You might look into obtaining/purchasing their results instead as it may be cheaper than the cost to test this yourself and "do it right".

Unfortunately arc flash testing and calculations are very dependent on many variables and assumptions, and the results you test yourself may end up being valid, or may lead you down the wrong path. Just some things to consider...

 

[Chenopod]

Thank you Zogzog, Overvoltage, this is very helpful advice.

I am designing a component that gets installed into a panel of the enclosure. In a nutshell, we want to build a system with arc pressure vents, create an arc flash inside, and see if our component causes failure per ANSI/IEEE C37.20.7. So this is more of a one-off R&D test than anything real-world.

Being a mechanical engineer, the switchgear ratings are what confuse me, for instance... MVA rating vs. AIC rating - What is the difference in their meaning (besides the unit)? How are they related? And what actually happens when you exceed them?

 

[dpc]

Forget MVA ratings - those are deprecated for ANSI switchgear for about 40 years. They relate to the short circuit rating for the circuit breakers.

Switchgear bus has a continuous current rating, based on thermal constraints, and a short circuit rating based on mechanical bracing and strength of the bus to resist bending, etc due to the magnetic forces during extremely high currents.

If you have the AIC (interrupting capacity) you have no need to know the MVA rating - that is also a short circuit rating but is now based on the interrupting current rating.

The main variables for arc-flash seem to be the type of bus material, bus spacing, bus mounting to the steel of the enclosure, size of enclosure, and relative position of the bus in the enclosure. There are big variations for horizontal and vertical bus based on past testing, but assume you'll be dealing with horizontal bus.

I'd suggest trying to review data from prior testing - there has been quite a lot.

David Castor

http://www.cvoes.com

 

[Zogzog]

Sounds fun, any reason I can find to go to a test lab like KEMA I go, had some fun days there.

I can get you some old switchgear for this if you need it, I have a little laying around.

 

[Chenopod]

dpc, thanks for the explanation - I did some more digging and found a couple papers explaining AIC vs. short circuit current, which is more relevant to lab-induced arc faults.

Zog, our testing will be at KEMA, I'm pretty excited about it. And I am indeed interested in your older used gear. The problem is in specifying exactly what I need, which is why I'm on this here forum =). How do we get in touch?

 

[Zogzog]

Not sure, forum rules are pretty strict here about posting info like that.

 

[davidbeach]

Links to company web pages seem to be left alone. Email addresses seem to disappear quickly.

 

[Chenopod]

Thanks. I still could use some help with the specs...

Regardless of voltage, we would like to pump 50kA and/or 63kA for 30 cycles at 60Hz because that is what we want the test report to say.

If we go with a 480V system, can someone suggest a nominal current rating that wouldn't result in a vaporized bus bar at this short-time current? It's ok if it gets destroyed mechanically from the pressure wave. And would your recommendation change for higher voltage (but same short circuit current)?

I'm trying to figure out what to ask for quotes on.

Is it considered personal info if one were to write something like david(dot)pelletier(at)extech(dot)com?

... just wondering =)

 

[Chenopod]

I have one other question to add -

IEEE C37.20.7 calls for switchgear that is "fully equipped". Since they leave it up to the manufacturer to interpret this, I'm wondering what components would comprise the simplest possible system. Does this just mean the enclosure(s), bussing, and one breaker? Or is there some other internal components that are essential for basic operation?

 

[Zogzog]

I think you could call it a "fully equiped spare compartment", but that may depend on what you are really testing.Worst case you add a breaker.

As far as your test current, if you stay within the ratings of the switchgear bus bracing you should be OK. (This is getting interesting, I might have to come watch this)

 

[LionelHutz]

This sounds interesting but I'm somewhat lost about what you are tring to achieve. If you create an arc fault in the switchgear than what are you testing? If you purposely cause an arc fault then I don't see how you could be testing if your component will cause an arc fault.

 

[Chenopod]

We want to see if the explosion creates a hole in the enclosure at the location where our product is installed. We're not testing whether the product itself causes the arc.