Transfer switch feeding transformer causes upstream circuit breaker trip

[TunaPower]

Fast, open transition transfer switch causes upstream circuit breaker trip on transfer. From transformer secondary currents with lower loads that survive without a breaker trip, I see an open time of about 30ms. Feeding a 45kVA dry, three-phase transformer 480V to 160V AC, 55FLA, for generator excitation power. Breakers trip going either direction (normal to emergency, emergency to normal). Upstream breakers are 70A Molded Case Circuit Breakers (MCCB) without adjustable magnetic trip (probably around 3X handle rating). Trips do not occur if we lower the excitation load on the secondary of the transformer. Excitation load is a six-SCR bridge to feed DC to a generator field coil.

I wonder if transformer inrush is exacerbated by having residual magnetic flux reverse the voltage on the primary terminals such as with DC relay coils, generator DC field coils, etc.

In my opinion, the breaker is simply undersized and should be 100A with adjustable magnetic trip set to 10-15X handle rating.

In a previous thread (Breaker tripping with ATS Transfer, thread238-103803), member rbulsara commented and referred to a paper about similar phenomenon, but I cannot find that--references to that paper would be appreciated.

I see lots of references an discussion related to transformer energization inrush but little about inrush after a short open time. Your help and comments are appreciated!

Regards
David

David

http://www.annainc.com

 

[itsmoked]

inrush after a short open time

Probably because in a transformer's case time doesn't matter. Short or long doesn't much alter the static magnetic polarity the last de-energization left it in. I believe that depending on that last state and the moment of re-energization there can be up to 20X inrush.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[RRaghunath]

Since the secondary remains connected, could it be that the primary terminal voltage of the transformer is being maintained by the connected loads on secondary. It can happen especially if there are capacitors in the excitation circuit.
Depending on the moment of switching, the voltage applied could be at 180 deg opposite to the voltage prevalent at the transformer terminals and this could give rise to huge inrush, making MCCB to see it as a fault.
Higher rated MCCB with higher Instt element tripping current should definitely help, provided the voltage dip during inrush is not going to cause problems to other feeders.

 

[beyond86]

Sure, residual magnetic flux can exacerbate inrush current. The high res.flux increases of inrush current and if we de energized transformer for a short period (some ms) res.flux will be higher and inrush current also.

 

[stevenal]

This sounds like hot load inrush. Cooper Eaton recommends 12X full load at .1s, and 25X at .01s.

 

[TunaPower]

Thanks for all the input


RRaghunath, the loads are a 6-SCR bridge that fires to feed an inductive field coil, but if terminated, the feedback from the field reverses and the internal crowbar would fire to duct energy from the field to keep that field voltage down. internal crowbar can fire in either direction since the 6-SCR bridge can create positive or negative energy. mostly fires on negative polarity from the field since the field current remains in the positive direction.

I like the concept you present, but I think there is little to no feedback from the field to the transformer. However on transformer re-energization, there is a lot of current (maybe twice what was there before loss of power) for a cycle or so. Know also that it takes a cycle or so to let the control circuit know where the zero crossings are so that the SCR firing angles can be calculated. FYI also, the current in the tests I refer to are perhaps only 30% of transformer FLA, so even twice that is not much on the primary side. Image I tried to attach shows secondary currents for two phases. other trips did not have current excursions. From this you can see a 30-35ms 0-current time.

From itsmoked (a questionable handle IMO, LOL), since its AC, the magnetic flux changes direction 120 times per second, I suppose that at each point in time, the mag flux can leave a static magnetic set at a certain level and polarity . . . so, a moment later (such as the 0.1 sec that Stevenal refers to and beyond86 (another lofty impending-doom handle) mentions as "some ms" the magnetic flux could be in opposite directions and lead to the perhaps 25X FLA at 0.1s Stevenal lists.

I wonder if there is a paper out there already that refers to this phenomenon? Maybe I will contact transfer switch manufacturers to see what they can tell me.

Thanks for all the input guys and gals

David
www.annainc.com

 

[itsmoked]

Tuna(hopefully mercury-free); There are electronic devices that keep track of where a transformer's magnetic circuit is and mostly eliminate the energization in-rush issues by accurately synchronizing closures to the polarity. Likely a provider of such controls would have usable documentation to assuage your interest.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[cuky2000]

See if the enclosed Link can help:

https://www.russelectric.com/wp-content/uploads/Tech-Brief-Dual-Operator-Trnsfr-Swtch.pdf

https://library.e.abb.com/public/f0...ormer Saturation WP 071018 WP-STS-MK-0074.pdf