Problem 1:
We have a three-phase Powerware UPS with a built-in input protection (thermo-magnetic) of 40Amp (C-curve) and an output protection of 40Amp (C-curve). In the downstream distribution panel is a supplementary protection of 40Amp (C-curve) with a distribution to 5 lines of 16Amp (C-curve).
When a short circuit is made on one of these lines the 16Amp protection and all of the 40Amp protections trip because the fault-current (Isc) of the power supply is bigger than the max. magnetic value of the 40 Amp protection. (8.5 x 40Amp = 340Amps)
What can we do to get selectivity in this UPS system?

Problem 2:
If the UPS switches to battery-supply the max fault current will be very low. Maybe below the magnetic value of the 16Amp protection. (8.5 x 16 Amp = 136Amp)
With a fault-current (Isc) through a person to the earth, only the thermo part of the protection will be activated which means that it takes to much time to trip. This way a person is not protected.
How much is the typical fault current of a battery-supply and what can we do to make it safe?

We have to solve both problems at the same time. Where can I find info?

Many thanks,
Leon

 

I would suggest that you contact application engineering at Powerware. The technical support line number is 1-800-843-9433 select hardware support option. They should be able to assist you, make certain you have model # and size available, they will need the information.

Dan Princinsky

 

The typical ways to get selectivity are through time delay between levels of protection or current settings. The input to the UPS requires a higher amperage setting due to inefficiency of the unit (typically 5-10 percent energy lost in heat). So, the input breaker typically can be set 10 percent higher than the output and if the breakers have time pickup settings, they can be set to provide separation. The smaller breakers do not normally have much flexibility in setting delays or managing selectivity.

I expect this will not help much as most low cost systems have breakers without the ability to adjust (only higher end breakers can be adjusted).

Anyway, hope this helps some.

Gene

 

Thanks Gene.
You're right about the time delay etc. but what about the high short cicuit current? A 40A and a 16A circuit breaker are time-selective but only up to a short cicuit current of 340A (8.5 x 40A with a C-curve circuit breaker). Our 3 phase 400V power supply for the UPS has a short cicuit current of at least 1kA. Even if we use a 40A and a 2A circuit breaker both will trip at that short circuit current. The time-selectivity doesn't matter in this case because the current is to high.
In normal control panels we use an isolating transformer which brings the short circuit current down to a level where the time-selectivity works. But in a UPS system we want as less components as possible to make it failsafe.
I wonder how it is done in other factories.

Leon

 

Thanks Dan.
We've contacted the Belgian Powerware distributor, but they have never thought about these problems. I'm gonna ask them to contact the Powerware factory. I'l let you know the results.

Leon

 

Responding to Leon Vangorp's query:

It amazes me that this problem still persists after some 40+ years of this technology. I suggest the following:

The three solutions I have found to be most effective are: a) install current-limiting breakers or fuses; b) introduce small inductors
downstream of the of the branch circuit protective devices; and c) use an appropriate single inductor between the source LV protective device and the branch circuit protective devices.

Regards,
Phil Corso, PE
(Boca Raton, FL)