Random transformer issues

[Razzap]

I recently wired up a three-phase, 51 kVA isolation transformer to a fusable disconnect switch. When it was first energized, everything seemed to be fine. From then on, nearly every time the disconnect was turned off and the back on, a fuse (or fuses) would blow. Completely random, sometimes phase A, next time maybe phases B and C together, once in a while it would be OK, and so on. We swapped out transformers (three other identical units), cabling from the disconnect to the transformers, everything. We have gone through three boxes of fuses, still no fix. What is odd is that the secondary is not even hooked up! When only XO on the secondary was grounded to the enclosure, it blew fuses EVERY time. Here are the specs: Primary 480 V Delta with taps, Secondary is 480 V Wye with a XO connection. 51 kVA, 60 Hz, 3 Phase, 4.4% impedance. Fusable disconnect is 60 amp, with 45 amp LPJ time delay fuses (sometimes even 50 or 60 amp fuses!) Three #6 AWG wires with a #6 EGC in a cable feeds the transformer. This acts like a short circuit, too fast for an AMprobe reading. Fuses blow immediately upon energizing the transformer. Cleaning the factory primary tap connections seemed to help a little. I have run out of ideas, anyone out there have a similar expirience?

 

[don_resqcapt19]

The fuses are too small for the inrush. The full load current is 61.4 amps.

 

[Razzap]

There is no full load current because the secondary is not hooked up, only inrush and energizing current.

 

[mike7330]

Inrush is too high for your fuses. It does not matter if you have load on secondary

 

[don_resqcapt19]

Actually in some cases no load on the secondary increases the inrush. You need to have larger primay fuses. You will also have to change the disconnect, because 60 amp fuses won't be large enough to eliminate this problem.

 

[Cow]

We swapped out transformers (three other identical units), cabling from the disconnect to the transformers, everything. We have gone through three boxes of fuses, still no fix.

Who sized the fuses and disconnect? Why are you swapping parts without verifying the size of all components? A quick call to the transformer manufacturer would have confirmed the fuses were too small and saved you a lot of trouble and part swapping....

 

[Razzap]

Engineering specified the size of the transformer and disco. We have many other installations using this same size equipment (and with the secondary hooked up) with no blown fuse issues. That is interesting about a higher inrush with no load on the secondary. Can you explain why? Don't know if it makes a difference, but this is a machine tool company, the equipment is for an electrical control panel. The full load on the machine is around 30 to 35 amps.

 

[gar]

120130-2143 EST

Razzap:

The residual flux in the core at turn off is probably nearer saturation when the transformer is turned off without a load. With a large load on the transformer the primary current and voltage are more nearly in phase. Even with a mechanical switch circuit turn off will occur closer to a current zero crossing.

Thus, with a moderately resistive loaded transformer turn off will be closer to a volt-time integral of zero, and lower residual flux.

With an unloaded transformer the current is very lagging, and at a current zero crossing the volt-time integral will be closer to a maximum.

When you apply power to the transformer it may be such as to force the core into saturation and thus a high current.

.

 

[jim dungar]

Drive isolation transformers have higher inrush than they did just a few years ago (thank you energy efficiency requirements).

 

[Besoeker]

Drive isolation transformers have higher inrush than they did just a few years ago (thank you energy efficiency requirements).

Never understood the rationale for drive isolation transformers.
On the OP's problem, I'm with many others here. Inrush.
We routinely fit 400V to 110V transformers to provide control voltage in our equipment. Typically these are 500VA so less than 1A in the primary at rated output. The primary has 10A fuses. There is no expectation that this will protect the transformer. Just the wiring to it.

 

[LISHAJI]

Looks like resonance!! Non linear magnetizing current interacting with the winding capacitance. The core may be going to saturation as a result!

 

[hurk27]

We have many other installations using this same size equipment (and with the secondary hooked up) with no blown fuse issues. That is interesting about a higher inrush with no load on the secondary. Can you explain why?

Simple supply circuit impedance, do the other sites have the same amount or run distance?

Remember resistance is current limiting

 

[Joethemechanic]

Actually in some cases no load on the secondary increases the inrush.

I'm right there with you on this one. I've seen it happen

 

[Joethemechanic]

Why are you fusing too small for the full load capacity of the transformer anyway?

 

[gar]

120121-1435 EST

LISHAJ:

It is simply a core saturation problem. When a core is left in a residual flux level near saturation on turn-off, then when voltage is reapplied on the next turn-on and the volt-time integral continues to increase the flux in the direction of saturation a very large amount of current is required to drive that core. If the voltage polarity was opposite at turn on, then the transient would be small because the flux is being driven away from saturation.

See my photos P6 and P7 at http://beta-a2.com/EE-photos.html .

.

 

[Razzap]

All the machines have been nearly identical, with 60 amp fusable discos and 40 amp Time Delay LPJ fuses. The transformer sits on the floor next to the panel, with about 18 feet of primary and secondary #6 AWG 4-conductor cable.

It would seem that the Time Delay fuses would hold long enough for the inrush current to dissapate, unless the inrush was much higher than anyone expected.

I am just an electrician that wires up machine tools to the best of my ability and skill, using prints from Engineering. Any changes must have engineering approval. They make all the calculations, and determine wire size, and fuse size. Not me! When the machine is wired, I move on to the next one. A start-up/debug team takes over and makes sure it ready for the programming team. I do not have much input. Once in awhile I get to turn on the disconnect. Engineering calls all the shots.
From what I have learned from this post, it looks like the transformers were operating "on the edge" as far as inrush current and impedance and we just got lucky up till now. I will pass all this info on to engineering. Thanks to all for your input!

 

[Jraef]

All the machines have been nearly identical, with 60 amp fusable discos and 40 amp Time Delay LPJ fuses. The transformer sits on the floor next to the panel, with about 18 feet of primary and secondary #6 AWG 4-conductor cable.

It would seem that the Time Delay fuses would hold long enough for the inrush current to dissapate, unless the inrush was much higher than anyone expected.

I am just an electrician that wires up machine tools to the best of my ability and skill, using prints from Engineering. Any changes must have engineering approval. They make all the calculations, and determine wire size, and fuse size. Not me! When the machine is wired, I move on to the next one. A start-up/debug team takes over and makes sure it ready for the programming team. I do not have much input. Once in awhile I get to turn on the disconnect. Engineering calls all the shots.
From what I have learned from this post, it looks like the transformers were operating "on the edge" as far as inrush current and impedance and we just got lucky up till now. I will pass all this info on to engineering. Thanks to all for your input!

No, I don't think the transfpormer is operating "on the edge", I think your engineers are just not doing their job.

Gar is right, this appears to be just a common saturation / inrush problem. One issue that might explain why this one and not the others may be based on the transformer design. Improvements in energy efficiency have lead to an increase in this problem. If your other installs are older, that could easliy be it.