First: This is not a question about electronics, it's about large (three-phase) power systems where the components have ratings in kV and MVA. I'm mainly interested in symmetrical components, and how the zero-sequence network is modeled. Although the terminology and simulation tools might differ, the physics should be pretty much the same for small and large components.

I have a three-winding transformer connected YNynd, where the primary side is solidly grounded, the secondary is resistance grounded, and the tertiary winding is delta connected.

In the simulation tool I'm using (PSS/E), three-winding transformers can be modeled, but not transformer with the parameters described aboce. I'm wondering if the following will be a physically correct way to model the system.

Disregard this paragraph (I'll let it be, as this was the original proposal:

If I'm modelling this as a YNdd transformer with solid ground on the primary side, and delta windings with angle 30 deg on the secondary and tertiary sides. Between the secondary winding and the bus on the primary side, I insert a zero-impedance 1:1 two-winding transformer with vector group Dyn where the secondary side is resistance grounded, and the angle reverses the 30 degrees from the three-winding transformer.

New proposal (after Lewis' comments):

Model the three-winding transformer as YNynd1, transformer, with both primary and secondary solidly grounded. Then, I add a 1:1 transformer with the primary side solidly grounded, and the secondary resistance grounded, as shown in the last figure. I'm not sure if the intermediate windings (secondary of three-winding and primary on two-winding should be grounded or not). I can't wrap my head around the zero-sequence equivalent of this (a few years since my university days).

I'm mainly interested in simulating line-ground faults on the secondary side of the transformer, but I obviously want the model to be correct for all other cases. I'm not interested in internal faults in the transformer.

Real system:

enter image description here

Suggested model:

enter image description here

  • \$\begingroup\$ PS modelling a YNyn0d1 transformer as a YNd1d1 transformer is never going to go well. YNyn has some peculiar characteristics which are important. \$\endgroup\$ Commented Aug 27, 2015 at 8:55
  • \$\begingroup\$ A hand-drawn sketch of the SLD of the physical system, on the back of a napkin, would be better than nothing. (I've drawn someting, then taken a photo of the drawing on my phone, before: electronics.stackexchange.com/questions/156519/… ). \$\endgroup\$ Commented Aug 27, 2015 at 9:41
  • \$\begingroup\$ I appreciate the comments @Li-aungYip! I've edited the question, and proposed a new solution. Any views on this? \$\endgroup\$ Commented Aug 27, 2015 at 10:11
  • \$\begingroup\$ What did you end up doing here? \$\endgroup\$ Commented Oct 14, 2015 at 15:54

1 Answer 1


I have done this before.

enter image description here

Make the 132kV winding of the 3W transformer be un-earthed wye.

Insert a "node bus" at 132kV.

At the node bus, model a shunt impedance (if PSS/E will let you do such a thing - maybe it's called a 'constant admittance'.)

Make the shunt impedance have infinite positive sequence impedance, and 30 ohm zero sequence resistance, such that the phase to ground fault current is 2,540 amps. Note 2,540 A = 132kV / sqrt(3) / 30 ohms. (If you get a funny number, try using Z0 = 1/3rd of 30 ohms, or 3 times 30 ohms - funny things happen with 3 × I0.)

enter image description here

I have to run now, please comment if anything requires further explanation.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.