Y-Yg with buried Delta tertiary - Secondary Ground Fault 2

[ZeroSeq]

Hello,

For a Y-Yg transformer (primary floating, secondary solidly grounded), a secondary ground fault will have no current flow as there is no path on the primary windings for the necessary congruent current to flow. However, will this still be the case for a transformer with a buried delta wdg (i.e. same Y-Yg transformer but with buried delta, vector group I believe would be Yynd0)?

My intuition says no, the buried delta will not provide any zero sequence path for secondary ground fault current.

I have tried to model this within ATPdraw, and the case for a simple Y-Yg transformer is clear (no current flow for secondary ground fault), however it is very difficult to model the buried delta and therefore I cannot conclusively say it will have no effect.

Thank you.

 

[RRaghunath]

YNyn0 transformer is straight forward - Rhase earth fault current in secondary is reflected as Rhase to neutral current in primary.
In case of Yyn0 transformer, as you rightly mentioned, wan earth fault on secondary side will not cause any current to flow as there cannot be corresponding current on primary side due to star point being floating.
In case of Yyn0d11 (or Yyn0d1), the buried delta will set up circulating current within Delta when there is current in one phase on secondary side. This will show up as current in two phases on primary side (for a single phase to ground fault on secondary side). The behaviour is similar to if it were Dyn instead of Yyn with buried Delta tertiary.
I worked with YNd connected transformers at 132kV. These transformers used to act like zero sequence current source for earth faults in 132kV OHLs.

 

[ZeroSeq]

Thanks for the reply. So the buried delta will facilitate current flow for a secondary SLG fault? The question I would then ask is how would you calculate the zero sequence impedance for the SLG fault? If the buried delta (obviously rated much lower than the primary and secondary) facilitates the current flow on the yn wdg, would it not considerably limit the magnitude?

 

[prc]

Zero sequence impedance of Ynynd transformers is 0.8-0.9 times positive sequence impedance.

 

[ZeroSeq]

Thanks prc - is that 0.8-0.9 per-unit Z1 the same for a Yynd (floating primary)?

 

[prc]

Z0 value as measured from HV and LV will be different. In the case of Yynd, from HV side value is infinity and LV side 0.8-0.9 Z1. You can get Zo of all combinations from Table 1 of IEC 60076-8 Transformer Application Guide.

 

[dfdt]

YNYn0 transformer will always transform line to ground fault on LV side even though the neutral HV winding is isolated. Depending on Trf construction (like 3 limb) core ; the zero seq balancing flux flows via the tank of the transformer. It called phantom tertiary. Zero seq impedance of such transformer is relatively high i.e. 0.5 pu to 1 pu depending upon the design.Adding buried delta winding will reduce zero seq impedance. Also L- G fault on LV will be seen on two phases on HV side.

 

[RRaghunath]

Vector Group Yynd suggests that the primary star winding neutral is isolated (from earth).
Hence, for a single line to ground fault on secondary side of transformer, there cannot be any zero sequence currents circulating on primary side.
But due to the availability of tertiary delta, the zero sequence currents can circulate on secondary side of transformer.
With Zo of 0.8-0.9 of Z1, the magnitude of earth fault currents can be as high as three phase fault currents (and even a little more in case of a bolted fault).

 

[dfdt]

Transformation of phase to ground fault current in YNYno vector group transformer is one of the least understood concept. As mentioned in my previous post, the balancing zero sequence flux flows through the tank of the transformer in case of three limb core type transformer.Zero seq impedance of such transformer is in the range of 0.5-1.0 per unit (not of positive seq impedance as mentioned in previous post). FYI all of our distribution transformers (110/22 kV) are of similar vector group and we do see balancing fault current on hv side as phase to phase for phase to ground fault on LV side.

 

[mLp85]

I agree with "dfdt" and my experience with ~50 Ynyn0 transformers in our fleet I can assure you that there certainly is fault current for secondary side earth faults.
The zero sequence pathway is created via the 'Tank Effect', giving rise to a large zero sequence impedance( but not infinity) so your earth fault level can be quite low but still significant. We commonly see ~ 100%Z.

One difficulty is predicting this value , because it relates mostly to the construction of the transformer tank and earthed components. It's also non-linear throughout the load range. Best to measure it at several points to get a rough idea and then set downstream EF protection conservatively with a decent sensitivity safety margins.

Adding a buried delta improves the situation, by allowing the behavior to become linear and predictable aswell as reducing the zero sequence impedance by giving it a controlled pathway between windings ( but no way near what a DY transformer would give). You can find values in the order of 300% x pos sequence impedance, and/or fault levels 150% of what a YY would have given you.

 

[davidbeach]

What's being called "Tank Effect" is often labeled as "phantom delta". It happens in 3-phase core form transformers. To a less extent, if at all, in 3-phase shell form transformers. It does not occur in banks of three single phase transformers.

I’ll see your silver lining and raise you two black clouds. - Protection Operations

 

[ZeroSeq]

Hi mLp85 - should the vector group you're referring to be "Yyn0" and not "YNyn0"?

Wouldn't the Z0 of a YNyn0 be exceptionally low, and the tank effect being a consideration only where one of the Y-windings is open circuited (e.g. Yyn0)?


Just bringing to back to my initial inquiry, here's a few questions:

1. Starting at a Yyn0 transformer (and ignoring any tank permeance), is the ground fault magnitude on the secondary windings zero given that the zero sequence network is completely open circuited?

2. Now, adding a buried delta (Yynd0), and again ignoring tank permeance, does the embedded delta provide a path for ground fault current on the secondary winding to flow (I believe the above responses point to yes)? The magnitude is up for debate.

3. Now, adding non-zero tank permeance to a Yyn0 transformer, a zero sequence path may be closed through the relatively high reluctance of the transformer tank, allowing ground fault current to flow on the secondary winding? According to mLp85, this may result in a Z0 ~ 3.0x Z1.

4. And finally, a YNynd0 transformer zero sequence impedance is around 0.8-0.9 x Z1 (per prc above). Does the buried delta change the Z0 magnitude given that both Y star points are grounded (lets assume solidly for arguments sake)?

 

[mLp85]

Hi ZeroSeq,

I see where some confusion has been added, but to clarify - having a "YNyn0" vector group doesn't predicate that the HV side is earthed. It's not uncommon for the neutral to be taken out on both sides, but left floating on the HV."YNyn0" and "Yyn0" would behave the same in this instance, so long as the HV side in unearthed. The comments made referred to the original statement that your transformer had HV side floating and LV side earthed. If you earthed both sides of the transformer, then the behavior is totally different and all those numbers quoted would not be true.

With star-star transformer, HV side unearthed, LV side earthed...
If you theoretically chose to ignore the "Tank Effect", then there is no zero sequence pathway and no fault current - but we know in reality there is the "Tank Effect" present.
Adding the buried delta, provides a zero sequence pathway, irrespective of the "Tank Effect".

 

[dfdt]

I agree 100% with mLp85.There seems to be a confusion regarding correct notation of Trf vector group where HV winding is star unearthed and LV side is star earthed. In my view we can use YNyn0 symbol to represent this vector group. Earthing a trf neutral or not earthing is not governed by the vector group. All of our distribution transformers are star unearthed on HV side and star grounded on LV side. In fact we notate our Trf as YN0yno (0 in HV side indicating that neutral bushing is brought but not earthed)
Now coming back to reflected fault current on HV side for LV phase to ground fault. Yes you surely will get amp turn balance due to tank tertiary effect regardless if it has buried delta winding or not. Zero Seq imp of our Trf ( with no delta tertiary)is typically in the range of 0.5 to 1.0 pu there by yielding 1 to 2 times of full load as fault current

 

[prc]

dfdt-
1)YNyn means both Hv and LV neutrals are grounded. It can be YNyno or YNyn6 etc. The last number is the clock number for the yn position with respect to the YN vector. So it is not correct to put YNoyno.

2) When you have a Yyn0 winding connection, zero sequence impedance is available only when excited from the yn side. It will be pretty high -say 60 % or more. With 5 limbed core or with single phase banks, Zo will increase to values near to magnetizing impedance, say 10,000%.

3) If you have a YNyn0 connection, zero sequence impedance when excited from the YN side will be 80 % of transformer impedance and 90 % of the impedance if excited from yn side. If the transformer is with a 5-limbed core or a bank of single-phase units, Z0 will be the same as transformer impedance. When a stabilizing delta is added, Zo will come down to 80 % of impedance or lower, irrespective of core or single or three phases.

4) Yn means neutral is grounded. So if you have brought out neutral and not grounded, then call it a Y connection and not YN.