Protection coordination problem 1

[Eclair]

Hi

I have simulated my electrical system on ETAP on it gives me a short circuit current too high if a 3 phase short circuit happens. It makes the distribution panel breaker useless: the upstream breaker trip and I lose all my loads (servers...).

What are my solutions then?

I thought to introduce a fused panelboard instead of breaker. It would probably solve my problem. Expensive though cause the panel have to be hand-made.

I also thought to limit the short circuit current but how and what? By introducing an additional a serial impedance between my MCC (600V) and my panelboard, it would probably reduce the short circuit current. Extra cable? An inductance?

What do you think about that?

 

[davidbeach]

Welcome to the joys of coordinating molded case circuit breakers. The things you mention are all options, probably none of them good. Any additional series impedance that includes resistance will increase your voltage drop while it limits fault current. One way of inserting blocks of impedance without too much impact on voltage is transformers. The delta-wye transformer has the advantage of preventing secondary ground faults from being seen as ground faults on the primary and ground faults are far more common than three phase faults. Good Luck.

 

[Eclair]

My circuit is:

MCC(600) -> cable ->600-120/208 transformer-> cable -> 120/208 panelboard

There is already a transformer. Why is inserting a serial inductance between my cable and my panelboard would not work?

 

[dpc]

As David said, molded case circuit breakers in series will generally not coordinate.

For the main 480 V switchboard, use of power circuit breakers instead of molded case breakers would allow elimination of the instantaneous trip and allow at least one level of coordination.

Use of fuses is another option.

The problem with adding inductance (a reactor) in series is that it is always there, even when you don't need it. The additional reactance will have its own voltage drop, heat loss, and will generally be a big pain in the neck.

In general, you just have to live with it, or spend the money to put in breakers that will coordinate.

 

[Eclair]

Ok I'll go into the details.

The circuit explained in my previous post is the bypass of an 40KVA UPS. There is a serial external by-pass cabinet containing another breaker as well.

The problem that we observed on a similar system, is that when a fault occurs, the UPS reacts a lot faster than the downstream breaker and goes on its bypass. Then it is the breaker of the external bypass cabinet that trips 50% of the time even if the coordination is okay on paper. Result: all loads are down.

Now, on a similar system not yet installed, I am not coordinating on paper. My thought is that a serial reactor put downstream after the UPS would limit the fault current and give more time to the distribution panel breaker to react before that the UPS sees the fault.

What do you think about that?

 

[davidbeach]

A much better idea: Use two UPS systems and all dual corded equipment. Size both sides to supply 100% of the load. One side trips and the other side continues to supply all of the load. The reactor won't accomplish what you want, the UPS doesn't have any inverse time characteristics, either it sees the fault instantaneously or it never sees it, you can't delay the UPS response to the fault unless you can delay it forever.

 

[Eclair]

I think I will go with the fused panelboard. Can you confirm that it would work?

And is this a recurrent problem that two serial breakers could not coordinate? What are the usual solutions? Cause it is my first real engineering project I don't see any solution than to put this extreme fused panelboard.

(the choice of the upstream breaker can't be replaced...)

 

[dpc]

Fuses in series may coordinate if properly sized. Coordination of fuse in series with a molded case breaker is more problematic. Might work and it might not. It will depend on the magnitude of fault current and response of the breakers.

 

[alehman]

Yes, fuses may or may not coordinate. There is no guarantee. As dpc said, closely coordinating fuses and breakers is at best a guesstimate.

The system can be designed to coordinate with careful selection of breakers or fuses, but this must be considered during the design and may lead to OCPD sized larger than required for the load.

Squared D offers a line of high instantaneous trip breakers designed specifically to help with this problem in some situations.

 

[Rodmcm]

Sorry DPC but I have to take you to task here. In general MCCBs do coordinate but you need usually two breaker sizes between them to get the inbuilt hisets to coordinate

If you start with a sensible lower rating for the cicuit to be protected then usually the normal adjustable current pickups on the larger breaker allows normal overload discimination. Often this resuts in a an 'unusual' sized upstream breaker, but a note on the drawing can explain the rational.

I have seen fuses in series with circuit breakers to limit fault levels and allow smaller cables to be used for small loads. This is an economic solution, but I agree usually no easy discimination between fuse and breaker.

 

[DanDel]

Is this a coordination issue, or are the devices not rated for the available fault current? Or both?

 

[dpc]

Rodmcm,

The only way two molded case breakers in series will coordinate is if the fault current level is below the instantaneous trip level of the upstream breaker. In practice, this does occur, especially for small systems will lower levels of available fault current.

But I would never expect these to coordinate unless I had done the necessary engineering review.

Two molded case breakers in series that both see fault current above their instantaneous trip levels will NOT reliably coordinate.