Phase loss behavior of YG-YG autotransformers

[mb3928]

For a standard YG-YG three phase transformer with a 3-leg core, I understand that a phase loss on the primary side will be regenerated due to the magnetic flux path through the 3rd leg of the core corresponding to the lost phase. Less or no regeneration will occur for a 4 or 5 leg core or shell type.

Does the same hold true for a three-phase YG-YG three leg core autotransformer? I would think so, because the magnetic flux paths are the same, but looking for confirmation.

 

[waross]

It's called the phantom delta effect.
The flux through the two "Hot" cores will mostly return through the third core.
There may be a lot of leakage leading to tank and hardware heating.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[mb3928]

It's called the phantom delta effect.
The flux through the two "Hot" cores will mostly return through the third core.
There may be a lot of leakage leading to tank and hardware heating.

Thanks! Just to confirm, you're saying that the same effect happens, regardless of whether it is a standard isolation transformer or an autotransformer?

 

[waross]

I don't think that the core cares that much.
The more load on the open phase, the more leakage flux there may be.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[mb3928]

I don't think that the core cares that much.

Probably not, they're just dumb iron!

Do you think the below scheme would work for phase loss detection? It might also provide some ground fault protection but that is not the point. The inverter shown is a voltage-controlled current source with high Z1 such that I_LLL is on the order of 1.5pu. The inverter neutral is for sensing only and Z0 is effectively infinite.