Withstand ratings for Switchboard and Switchgear

[elec_eng]

The withstand/short circuit rating for the switchboard is only 3 cycles based on UL 891.

If I have a time delay in the main circuit breaker more than 3 cycles to coordinate with the downstream circuit breakers, do I have to go with the switchgear, which will have 30 cycles withstnad rating per ANSI standard?

 

[Mike01]

UL

UL

If you are installing a circuit breaker in a UL891 switchboard the breaker I assume is a UL489 device in that the instantenous cannot be removed or turned off as these devices per UL must have an instantenous, however If you modify the breaker or allow it to carry fault current for more than three cycles than you risk damaging your equipment so switchgear would be more appropriate however this requires more space typically front and rear access. There is currently no UL test for switchboards for 30 cycles although some manufacturers will ?Self-certify ? their switchboards for a 30cycle rating, but then you have to have a 30Cycle breaker upstream. The magic question is if the AIC rating of the equipment is say 65Kaic and your fault current is less how long will the bus be able to handle the fault prior to the breaker clearing? If you have a 3cycle board 65Kaic and make your breaker settings above 3 cycles and you get a fault at 35K will the board be able to handle this fault current and for how long? I imagine its all about heat and magnetic forces.

 

[jim dungar]

When UL gives an SCCR (bus bracing rating) to a switchboard/panelboard it is based on being protected by a UL489 device regardless of the settings you choose.

 

[elec_eng]

If you are installing a circuit breaker in a UL891 switchboard the breaker I assume is a UL489 device in that the instantenous cannot be removed or turned off as these devices per UL must have an instantenous, however If you modify the breaker or allow it to carry fault current for more than three cycles than you risk damaging your equipment

Don't they allow delays up to 60 cycles for Ground fault protection?

The magic question is if the AIC rating of the equipment is say 65Kaic and your fault current is less how long will the bus be able to handle the fault prior to the breaker clearing?

That is a good point..

If you have a 3cycle board 65Kaic and make your breaker settings above 3 cycles and you get a fault at 35K will the board be able to handle this fault current and for how long? I imagine its all about heat and magnetic forces.

If the board is 65K aic, I would expect to see close to 65K aic fault current..otherwise, it would be a waste of money?

 

[elec_eng]

When UL gives an SCCR (bus bracing rating) to a switchboard/panelboard it is based on being protected by a UL489 device regardless of the settings you choose.

That was what I was wondering...if UL gives a SCCR to a swichboard being protected by a UL 489 device regardless of the settings, why they limit the test to 3 cycles?

 

[Cold Fusion]

Don't they allow delays up to 60 cycles for Ground fault protection? ---

First, don't confuse NEC minimums with prudent design specs.

Second, 230.95 says 1 second at 3000A. That's a pretty small fireball compared with an arcing phase to phase 65kA fireball on the main bus.

---If the board is 65K aic, I would expect to see close to 65K aic fault current..otherwise, it would be a waste of money?

About 20 years ago I'm fighting an SCC calc trying to get it down to under 42kA, so we don't have to spec the 65kA gear. The Chief Engineer comes by to see how I am doing and says, "Any job that needs a 2mVA xfm, can afford the 65kA gear". I think the message was, "Don't be marginal. It is a waste of money if the facility gets a fault and the equipment melts down instead of just getting black."

That was what I was wondering...if UL gives a SCCR to a swichboard being protected by a UL 489 device regardless of the settings, why they limit the test to 3 cycles?

You already had that answer in your first post - Money. UL1558 switchgear costs maybe 50% more than UL891 switchboards. But if the xfm is that big and the coordination has to distinguish between a 40kA feeder fault and a 65kA bus fault then the extra money won't be a problem - if that's what the customer wants, they will happily pay.

Disclaimer: I don't design substations or distribution. I just get them working and keep them working. So everything I said could easily be all wet:-?

cf

 

[elec_eng]

First, don't confuse NEC minimums with prudent design specs.

Understood..but the safety shouldn't be comprimised..that is what I am concerned.

Second, 230.95 says 1 second at 3000A. That's a pretty small fireball compared with an arcing phase to phase 65kA fireball on the main bus.

No..actually, it says 1 second for ground fault current equal to or greater than 3000 A. but, the ground fault will be roughly 1/3 of the bolted fault, it might withstand more than 3-cycles..

About 20 years ago I'm fighting an SCC calc trying to get it down to under 42kA, so we don't have to spec the 65kA gear. The Chief Engineer comes by to see how I am doing and says, "Any job that needs a 2mVA xfm, can afford the 65kA gear". I think the message was, "Don't be marginal. It is a waste of money if the facility gets a fault and the equipment melts down instead of just getting black."

I don't want to be marginal, either...but you wouldn't spec 65Kaic, if you only expecting 35kaic... that is my point.

You already had that answer in your first post - Money. UL1558 switchgear costs maybe 50% more than UL891 switchboards. But if the xfm is that big and the coordination has to distinguish between a 40kA feeder fault and a 65kA bus fault then the extra money won't be a problem - if that's what the customer wants, they will happily pay.

My point is that if the switchboard is only rated for 3-cycles, should they prohibit the time delay more than 3-cycles for a bloted fualt?

 

[Cold Fusion]

---My point is that if the switchboard is only rated for 3-cycles, should they prohibit the time delay more than 3-cycles for a bloted fualt?

Who is the "they"?
cf

 

[Cold Fusion]

---My point is that if the switchboard is only rated for 3-cycles, should they prohibit the time delay more than 3-cycles for a bloted fualt?

First, be sure you read my previous disclaimer:-? I'm a bit out out on my area - I do my danmedest to stay out of the design business. I'm a fixer. With that in mind:

I'm thinking this stuff shows up on the coordination study. For this size project, the study is pretty through. The first sheet shows the cable damage curves, transformer damage points, inrush currents, bolted fault currents, ground fault currents ... And the over current protection device curves have to be inside of all of the damage curves. i don't do the studies, but i look at them and do a lot of verification they look to be truth.

Now I'm thinking that if the switchboard damage toa fault is questionable, then that damage point has to be on the coordination sheet and the OCP set inside of that damage point.

I don't think anybody is going to give you a receipe for this. You're the one that is doing the coordination/fault current study and is making sure the OCP is coordinated with the equipment damage curves. I don't see a switchboard damage point as any different than a transformer damage point or a cable damage curve.

What do you think?

cf

 

[jim dungar]

UL891 and UL489 devices are coordinated. You do not need to worry about a 3 cycle rating with these paired devices.

A 65KAIC rating means it can interrupt a fault on a system with 65KA available, not that it can actually interrupt 65K. In some case the internal breaker impedance + the 'contact separation' arcing keeps the current from actually reaching 65KA, in other cases the fault is cleared so fast that the current does not have time to rise to 65KA.