Delta wye transfomer during a grounded phase 5

[Mbrooke]

What happens when one or two phases fault to the MGN (multi grounded neutral) of a distribution system with the effected phase fuse blowing (but still grounded) while feeding a delta wye transformer? Will any inductive tank heating take place? Will a 3 limb vs 5 limb core make a difference?


Basically I am entertaining the idea of using a delta wye padmount instead of a conventional 5 limb wye grounded wye grounded, however I am unsure if certain fault scenarios create special protection concerns. Both the over head feeder and underground riser are protected via fuses, so single phasing and shorted phasing is a concern.

 

[magoo2]

I don't see where you would get any tank heating.

Assume you lose A phase on the primary due to the open fuse. The delta will actually collapse. A-B and A-C will add up to B-C voltage. I don't think the ground fault matters at this point other than causing the fuse to blow. Result on secondary is two phase to neutral voltages at half value and the other one at rated voltage.

 

[HamburgerHelper]

He is talking about tank heating during core saturation due to zero sequence voltage. Zero sequence voltage can't saturate the delta side due to it not impacting the delta windings. The delta windings are acted on with line to line quantities and those are only composed of positive and negative sequence.

You'll have some trade offs going to delta wye ground wye ground. Your phase voltages will be impacted by unbalanced load. Delta wye ground I think is preferable in industrial systems due to the loads being mostly 3 phase (motors) and balanced and it segregates the grounding and providing a ground source.

 

[Mbrooke]

Correct, tank heating from zero sequence voltages. I know this is possible with a 3 limb wye-wye, hence why a 5 limb or tri-plex is specified.


Also my worry is not just an open fuse, but that phase back feeding (shorted) into the MGN before line crews show up. As long as the transformer itself does not become damaged I am content.


And- for those who know about ferrorsaonance. As long as the transformer remains under about 15-20% load, the resonant condition will be dampened out?

 

[magoo2]

For ferroresonance, you don't need that much load. Less than 5% load is adequate. What primary voltage are you dealing with?

 

[Mbrooke]

22.5kv primary voltage.

 

[HamburgerHelper]

Here are a couple of good papers we reference here when look into distribution transformer configuration and resonance.

http://s000.tinyupload.com/index.php?file_id=30823192245048788904

This link works. Just don't click on the bait. The zip file downloads when you click on the file name.

 

[Mbrooke]

Thank you, and that this is a life saver, would be an serious understatement! I owe you 2x now

 

[jghrist]

You will have zero-sequence voltage with one phase connected to the MGN. See the attached calculations that assume ØB is connected to the MGN (VB=0V). I've assumed 100A secondary current in each phase, in phase with the secondary voltage, to calculate some currents.

There will be no zero-sequence current, even though there is current from ØB to the MGN (it is all pos- and neg-sequence).

 

[Mbrooke]

Perfect and spot on assumptions. This is gold

 

[HamburgerHelper]

Jghrist,

Line to line quantities on the delta would never be affected by zero sequence because zero sequence doesn't produce a voltage differential across each winding.

The thing when I looked over your calculations that I had never thought of before is that the type of load I think could dictate what happens on the secondary of the transformer. The load that you used in your calcs was 100 A load in phase with the voltage on each of the phases, a constant current load, probably single phase. Zero sequence current on the secondary is generated by the change in angle of the secondary voltages and zero sequence current is produced due to the currents not adding up to zero. If you had just a passive load, I don't think you would have seen any zero sequence currents on the secondary because the secondary voltages with just positive and negative voltages would have only produced positive and negative sequence currents. I suppose you might see something similar with constant power loads. I suppose this isn't profound or anything but I guess the type of load you have could determine if you should care what is happening on the high side in respect to tank heating.

 

[Mbrooke]

@HambergerHelper: great point. I am assuming single phase resistors in that simulation, but what if you had other loads?

 

[prc]

Hamburgerhelper- Thank you for that old Westinghouse paper. There is a similar GE paper of 1967

http://www.geindustrial.com/publibrary/checkout/White Papers|GET-3388B|generic–The

Why of the Wyes: The behavior of transformer Y connections. All these incorporated in the IEEE Standard C57.105-1978(R2008) Transformer Connections in 3 phase distribution systems.

 

[HamburgerHelper]

Mbrooke,

I am looking at this wrong. The zero sequence voltage at the transformer secondary due to the load would be zero because it is grounded. It can't be anything but zero. The load is generating the zero sequence voltage due to how it interacting with the voltage imbalance. I find it curious that the secondary voltage consisting of just positive and negative components could cause zero sequence currents to flow but I don't think that has anything to do with tank heating. I don't think the loads can impact your problem.

 

[Mbrooke]

So in a nut shell tank heating is not a concern- even with varying loads and various line faults. Thus a 3 limb core will be of no issue- unless of course someone sees a 5 limb being superior.

 

[HamburgerHelper]

I don't think it is for a delta-wye-grounded transformer since it is a ground source. Zero sequence voltage can't affected the high side because they are connected line to line. The zero sequence voltage on the low side will be the zero sequence current times the transformer's impedance to ground and you have current circulating in the high side delta which will reduce zero sequence flux in the transformer core. So, zero sequence can't do anything on the delta side and on the secondary I think you would have to pull a lot of zero sequence current to get stray heating flux.

 

[Mbrooke]

Makes sense. Or we can view it like this: if a wye-wye had a delta tertiary, would we still get stray flux?

 

[HamburgerHelper]

I have seen wind farms with wye-gnd-delta-wye-gnd transformers. I don't know why exactly they select those transformers but the impedance of the tertiary delta makes it act like a wye-gnd-wye-gnd if it has really high impedance or a back to back wye-gnd-delta to a delta-wye-gnd for really low impedance. Anything, though that circulates in a delta provides counter flux to your transformer core and reduces its flux density.

 

[prc]

With a Dyn connection,or YN(d)yn and a Ynyn 5 limbed core, you will not have straying zero sequence flux from core and hence tank heating. But with a Ynyn connection with 3 limbed core, zero sequence flux in 3 limbs(say under unbalanced secondary loads)have no return path through core and it goes to tank to close the path ie tank acts as a virtual tertiary delta( one turn winding,carrying current and heating)
In case of wind farm transformers, please check the core construction. If it is 4 or 5 limbed, stabilizing delta will be required to bring down the zero sequence impedance. With a 3 limbed core it may not be necessary.

 

[waross]

What is the effect of the back EMF of any motors until the thermal over current trips or thermal differential trips clear? Won't you be back feeding the fault with a current limited by the motor impedances and the transformer impedance?

Bill
--------------------
"Why not the best?"
Jimmy Carter