Voltage Sag Due To Transformer Starting

[3winding]

Does anyone know how to estimate the voltage sag on system voltage upon energizing a transformer, using transformer and system impedance? I am trying to determine a ball park number before going to some type of transient analysis evaluation.

 

[dpc]

Transformer inrush is unpredictable. But if you assume about 10-12 times full load current for about 0.1 sec, you can get a rough idea. So take your system model and add a load equivalent to that amount of current (at power factor of about 0.1) and you might get some idea of the voltage dip. Of course, this high inrush is typically seen in only one phase. or at any rate, one phase will see higher inrush than the other two.

Transformer inrush currents are rich in harmonics, so that complicates the analysis. I'm sure there are a lot of papers written on this topic - you might search IEEE website.

I'd be interested to know what you come up with.

 

[3winding]

dpc, I am aware of the 12 times for 0.1 sec inrush factor. I'm not sure how to apply that to voltage drop. I thought maybe there was some rule of thumb such as - if the utility had a 500MVA SC capacity and I was energizing a 50MVA xfmr, I would see a worst case voltage drop of some percent. I do have access to motor starting software. I wonder if I could model a transformer as a motor and plug that into the software to get an instaneous voltage drop?

 

[dpc]

Yes, my thinking was to apply it as a load to the transformer primary bus. That would simulate the voltage dip created when the transformer was energized.

 

[jghrist]

See Reducing the magnetizing inrush current by means of controlled energization and de-energization of large power transformers,
László Prikler, György Bánfai, Gábor Bán, & Péter Beckerhttp at

http://www.ipst.org/TechPapers/2003/IPST03Paper9d-4.pdf

This includes an approximate equation for transformer inrush.

 

[ArtAllm]

>I am aware of the 12 times for 0.1 sec inrush factor.
>I'm not sure how to apply that to voltage drop.


Well, 3winding, if you know the inrush-current and the impedance of your grid you have to apply the formula:

U = I*Z