Autotransformer neutral impedance earthing

[smacd]

In order to limit 110kV single phase fault levels it has been suggested to employ impedance in the neutral of 250MVA 220kV/110kV autotransformers (wye-connected with a delta tertiary). To 'spec' the neutral impedance do i have to give a fault current level (say full-load of the auto) and a fault time/duration, say 20secs ??
Do i also have to give an ohmic value of impedance ?
I'm assuming reactive impedance at present. Would resistance be suitable also ?
Has anyone got similar experience of the above possible method ?

 

[electricuwe]

I know this type of grounding for 110 kV grids, the impedance used for this purpose is called Petersen-Spule in Germany.

But you can not apply this technique if the 110 kV grid is supplied by an autotransformer, because usually the 220kV grid is solidly grounded to limit voltage stress on insulation.

So unless you replace the autotransformer with an isolation transformer the impedance will be shorted.

 

[jwerthman]

I agree with electricuwe that impedance grounding the neutral of the autotransformer is probably not feasible nor would it solve the 110 kV ground fault duty problem.

Unless the transformer was specified with sufficient neutral-end insulation (BIL), any impedance inserted into the neutral would jeopardize the winding insulation during transients. Most grounded wye HV transformer windings in the U.S. are designed with graded insulation to save money, size, and weight. For example, the BIL rating for the 230 kV line side terminals might be 825 kV, but the neutral end BIL could be 150 kV. This is only applicable for a solidly grounded neutral however. In Asia and I believe in Europe also, it is not uncommon to find high-resistance grounded neutrals on large wye-delta transformers. But the neutral insulation level is increased substantially for this type of application.

Adding impedance to the neutral would reduce the zero sequence fault contribution from the tertiary branch of the transformer, but would do nothing to reduce the zero sequence current that flows straight through the transformer from the 230 kV system to the 110 kV system.

 

[smacd]

Thanks for the replies !
I take your point about the insulation levels for autotransformers, and it is an issue we're looking into with the manufacturer.
I also take the points made in jwerthman's last paragraph. However, our studies show that the reduction of fault current from the 'T' branch [should this actually be called the tertiary branch, as you mention ?? Zero sequence impedances would be measured between the 'H' (220kV), the 'L' (110kV) and the 'N' neutral terminals]is significant enough to aid us in reducing the 1 L-G fault level by 12%. An impedance of 48 ohms was employed. I'm assuming this value of impedance would have to be "spec'd".