Transformer primary circuit breaker tripping on ATS switch

[PaultheBigFinn]

The site is provided with 100% back-up generator power. After working fine for 6 months last Friday there was a power failure. The open transition ATS energized the generator and closed to it without a problem. On utility power return, the circuit breaker supplying a 150kVA transformer tripped as did the panelboard’s main circuit breaker as well as the switchboard’s electronic circuit breaker supplying the panelboard.

The system is a temporary arrangement where one small out building is operating while the main building is under construction. As such the main circuit breakers are rated 250A, and the circuit breaker supplying the transformer is 200A. The site is also in Canada so the primary voltage is 600V.

The outbuilding will house racks of electronic equipment, which are currently being set-up, so the electrical load very light. The problem appears to be transformer in rush (although, I'm open to other suggestions).

Does anyone have any suggestions on the best way to reduce in rush? We have considered putting a false load on the transformer, but that seem inelegant and hard to justify to the owner. The problem seems to be similar to motor in rush where capacitance banks have been used to dampen the in rush. Does anyone have any experience with this or other suggestions?

 

[dpc]

I'd suggest looking at the breaker settings. Tripping of the feeder to the transformer could be simply too low of an instantaneous setting. Tripping of the main breaker is more unusual.

It's very common to undersize these primary breakers serving small dry-type transformers. Inrush can be 12X for 0.1 seconds. Since the inrush is a function of the voltage phase angle at closing, it may trip only one out of 10 times.

Limiting the inrush will be very difficult. Having small amount of load on the secondary can help speed up the inrush decay.

Are you trying to pickup a bunch of motor loads on the re-transfer? This can make your problems worse.

Best solution is to increase the instantaneous trip on the primary breakers, if adjustable. If not, you may need to look at a larger feeder breaker, but first you need to review the settings of the feeder breaker and main breaker.

"The more the universe seems comprehensible, the more it also seems pointless." -- Steven Weinberg

 

[PaultheBigFinn]

Thanks for the reply.

I am also puzzled by the tripping of the switchboard circuit breaker which has adjustable settings that are reported to be maxed. We will verify them tomorrow. The breakers on the panelboard are thermal magnetic without adjustments and should have tripped in advance of the switchboard breaker.

The way the system is wired, we could replace the panelboard circuit breakers without much cost and that maybe the simpliest solution. Just how big the replacement breakers should be is another queastion.

We have found that the generator was delivering 590V to the ATS and when the utility returned the voltage was ~630V. Our site is fairly remote and the utility is supplying our transformer from a high voltage transmission line. Do you think this might have added to the problem?

It has also been suggested the transformer ground may have been compromised. Since it was a temporary installation, a single ground rod was used and I’m not sure how deep it was driven or if the winter heaving of the soil loosened the ground connections. Another thing to investigate Tuesday.

Do you have any other suggestions?

 

[davidbeach]

How fast was the transition? How was the phase angle at transition managed? Did the load include any thing like a UPS?

 

[PaultheBigFinn]

The ATS switching performed with charged springs, breaking and making in 2 or 3 cycles.

The ATS is open transifition to break the connenction to the generator before making the connection to the utility. Nothing is done to attempt to match the phase conenction.

There are two small UPS supplied from the panelboard, but not supplied from the 150kVA transformer that tripped it's circuit breaker.

 

[dpc]

I don't think 630 V on a 600 V system should be a problem for the transformer.

Ground connection should not be a factor for transformer inrush.

Can you provide sizes and types for the breakers that tripped?



"The more the universe seems comprehensible, the more it also seems pointless." -- Steven Weinberg

 

[PaultheBigFinn]

The circuit breaker supplying the transformer is 200A, molded case without adjustment. The panelboard main circuit breaker is 250A molded case also without adjustment. The switchboard circuit breaker is a 250A electronic power breaker with long and short trip adjustments. Because this is the preliminary phase of a larger project, the switchboard is supplied from a 1mVA transformer with the 250 panelboard the only load.

I agree that 150kVA transformer should be able to handle 630V. I wonder if the switch from 590V supplied by the generator to 630V by the utility in a half dozen cycles may have increased the transformer in rush.

 

[dpc]

I think the 200 A breaker is too small to reliably handle the inrush, especially if non-adjustable.

For a 150 kVA transformer with a 480 V primary, we use a 300 A breaker on the primary. So at 600V, I'd probably use a 250 A.

Of course, at this point, you will be limited by the size of the primary conductors.

Adjustable trip instantaneous units might solve the problem.

If you have something on the secondary of the transformer that is holding up the voltage (rotary UPS, etc), then you may have an out-of-sync condition - but sounds like that is not the case.

"The more the universe seems comprehensible, the more it also seems pointless." -- Steven Weinberg

 

[sibeen]

I'd be adjusting the ATS so there is a decent break between the transfer. With only 2 or 3 cycles switching time between the sources, the transformer will still be partialy magnetised. Depending upon the voltage and phase angle of the two sources the inrush current can be up to twice what was normally expected, up to 20 - 25 times In.

 

[PaultheBigFinn]

Our technicians have been at the site and have found that I was wrong about several things. First both the switchboard and MDP main circuit breakers are Square D type JL with over current adjustment only. They were both set to max at 10X250A=2,500A.

The ATS is a single blade type with a phase monitoring window. It looks at the generator and utility power and when they are within 5 degrees of each other for 1.5 seconds the ATS switches with a time between make and break of ~5 cycles.

The system is extremely under loaded the measured steady state load is ~10kVA on the 150 kVA transformer. The i9mpedance of the 150kVA transformer is 2.4%, making our dead fault current ~6 kAIC.

It has been suggested that we should scrap the 150 kVA transformer and install a 75kVA transformer in its place. This seems like a good idea, but even with that the system will be only 12% loaded (although the load will increase as the systems come on line). The 75 kVA transformer has an impedance of 4.85% with a dead fault current ~1.8 kAIC.

I appreciate all of teh good input I have received. Is there a more accurate method of determining the instantaneous current we might experience and it duration?

 

[dpc]

I don't think the system loading is a big factor in the transformer inrush or breaker tripping. The breakers trip on instantaneous so changing the inrush time constant is probably not going to make much difference. And the system should be able to energize a small transformer without worrying about the transformer load.

Inrush is very difficult to predict, but it's safe to say it will be a lot less with a 75 kVA compared with a 150 kVA unit.



"The more the universe seems comprehensible, the more it also seems pointless." -- Steven Weinberg

 

[Ceast]

Phase angle could the the problem. Is the generator driven into sync with the Normal source on return to Normal? Permissive?

 

[Ceast]

Whoops that's a very small gen. Not likely equipped with an auto syncronizer. Double check the freq on the gen. I've seen similar problems on units with permissive sync-check when the gen freq is too fast or slow.

 

[oldfieldguy]

Some ATS have an ability to close on a synch condition between the emergency and the normal source on return to normal. Check the documentation for your switch and see if it has this capability. If so, make sure it' is properly set up.

I had a similar problem at one of my stations. found the synch check feature wasn't turned on.

old field guy