Replacing a 1500kVA 3 phase Delta 2

[ffont]

An industrial customer is currently fed via a 2000kVA 13860V delta-600 GrdY 347V. The medium term plan is to replace the 13860V system by a 24940GrdY system, and install a new 2500kVA transformer connected GrdY-GrdY

What would be the drawbacks to remove the present transformer (delta primary) and feed him until then, ie a few years, with a 2000kVA 13860V Y - 600 GrdY 347V. (Note the transformer H0 would be insulated and left floating; only X0 would be connected to the ground grid).

Thanks for your help!

 

[mgtrp]

The delta winding meant that the transformer is an effective source of zero sequence current for the LV system. With this removed and the H0 bushing isolated, you will no longer have this.

 

[DaveMeharg]

I would think that the original configuration delta to wye blocks the zero sequence current. If you go with the wye wye configuration I would think that a ground fault on the LV side could be fed from the HV side and vise versa. In addition you may want to consider the affect of a voltage sag on one phase on the HV side might be more severe on the LV side then it would normally have been with a delta primary.

 

[DTR2011]

Don't forget the surge arrestor ratings!

 

[stevenal]

The original configuration blocked zero sequence on the 13.8 system, not the 600. The proposed configuration would perform as mgtrp stated.

 

[ffont]

Stevenal and mgtrp, will there really be no zero sequence currents flowing in the LV grounded neutral? The Primary will be an ungrounded wye, but the secondary will be solidly grounded. With no zero sequence current, I am increasingly worried that this may lead to ferroresonnance if one of the primary phases looses voltage.

 

[mgtrp]

The main problem that you would have is that in the event of a ground fault on one phase on the 600V side, the L-N voltage on that phase will collapse to zero, meaning that no fault current will flow, while on the other two phases, the L-N voltage will rise to 600V (assuming the transformer isn't saturating). The reason for this is that the ungrounded wye point on the high-voltage side is not tied to anything, and so it will "shift" so that its potential is equal to that of the potential of the high-voltage side of the faulted phase.

Presumably, your existing circuit protection on the 600V is based around detecting overcurrent to trip in the event of an earth fault. This won't function any more.

If you want to muddle through a more technical explanation, google "The Whys of Wyes", which is an excellent paper produced by GE on the subject of transformer wye connections, and freely available.

 

[prc]

In India we had seen use of delta/star transformers for supplying power to industrial loads. But today it is no more adopted and replaced by star/star transformers with earthed neutral on primary and secondary neutral earthed through a resistor.Why do you want to keep Ho floating?

 

[Marmite]

Zero sequence current will flow if the transformer is a three legged core type due to circulation of the flux via the tank. The classical representation of an open circuit zero sequence network is not correct for a three legged core type transformer.

Regards
Marmite

 

[ffont]

For mgtrp.... thanks! I can see the voltage collapse in the LV ground faulted phase and the corresponding shift of the neutral in the primary. I taught there would be some zero sequence flowing in the ground, but with no voltage there can be no current. It makes sense. But then, can you help clarifying what is the impact of grounding X0 if H0 is not grounded?

I also read that leaving the H0 ungrounded, will increase the chances of ferroresonance,(even on a 15kV system), specially because there will be some length of underground cable, and these new Wye padmounted transfomers will be low loss.

For prc..... H0 floating was an early tought. The early thinking was to this would help us to keep the upstream protection as it is currently (ie protecting the existing 13.8kV Delta transformer) until the primary voltage conversion to 25kV in a couple of years. But a floating H0 does not appears to be a good idea in this case!

 

[DaveMeharg]

For mgtrp: I would think if the secondary side is solidly grounded (600/347 side) then during a ground on one phase on the secondary you will have a large groud fault current flowing on this phase to ground. The voltage on the other two phases with respect to the neutral point would stay fairly stable on a grounded system. Do you agree?

 

[mgtrp]

Not if you have no earth on the HV neutral (and as Marmite points out, presuming no phantom deltas). Star windings do not act as a source of earth current, they can only pass it through. If the HV neutral bushing is unearthed, then there is no way for it to be passed through and so the neutral shifts.

See attached sketches to (hopefully) explain what happens in each of the two cases - X0 earthed, and X0 unearthed.

 

[stevenal]

You really don't need sequence components or voltages to see how this doesn't work. Connect a 1 pu current single phase load from Phase a to X0 at 347 V on the secondary side with no other loads. You'll have 1 pu current on a and returning to X0. b=c=0. To satisfy amperes law you will now need 1 pu on the primary A and N while B=C=0. But H0 is floating, so ampere's law cannot be satisfied. No amount can possibly flow on the secondary neutral for single phase loads or for LG faults. If you float the X0 as well, I see no difference. But do you want an ungrounded system?

 

[DaveMeharg]

mgtrp and stevenal: I reviewed the zero sequence connections for a wye-wye transformer and I agree that without the primary side grounded there will be no zero sequence current on the secondary. Thanks for clarifying that for me. I have not worked much with wye-wye configurations. On the industrial side most of our transformers are delta-wye grounded (either resistance grounded or solidly grounded). It seems like on the utility side there is more of a tendency to use the wye-wye grounded. Why is this and why don't they want to take advantage of the delta wye grounded configuration which would isolate the high voltage side from low voltage ground faults or third harmonics? Is it strickly due to minimizing the feroresonance concerns?