Star/Star TX - Zero Seq circuit 3

[CDG16]

Could someone please explain the equivalent zero sequence circuit for a 2-winding Star/Star TX (HV ungrounded and LV grounded). The confusion is that PRAG shows that no zero sequence current will flow for ph-E faults. But, the J&P TX Book shows a link to the zero bus on the LV side. (I found several other books showing these circuits both ways).

The only explanation for this is that J&P might assume the tank acting as a delta and therefore the closed link to the zero bus. The problem is, if you assume the PRAG is correct, then for a ph-E fault on the LV, no current will flow. But the winding has basically a short across it - it does not make sense...

Could someone please explain this better.

Thanks in advance.

 

[tinfoil]

I am confused.

jghrist said 'The voltage across both primary and secondary windings will collapse'.

Our typical model has an infinite bus connected through an impedance (call it Zgrid) to the area under study. If the voltage on H1 has collapsed, then the infinite, unchangeable bus now has to drop the full pre-fault system voltage across Zgrid. Wouldn't this have to compute out to an SLG on the primary side?


I never tried to imply that 'zero seq. current' was flowing through the xmfr; only that a lot of 'real' current will be passed through.

 

[tinfoil]

BTW, when you say 'collapsed', do you mean 'has reached the point of the vector sum of the H2 and H3 legs'?

I do not have access to the quoted IEEE document, but I thought that this table was interesting:

http://www.ieee-kc.org/library/pq/singlelinegroundfault.htm

 

[jghrist]

The primary winding neutral point (which isn't connected to the source neutral) shifts to the voltage of the faulted phase. If ØA is faulted, primary A-n voltage becomes zero, B-n and C-n voltages equal the system phase-to-phase voltage.

Your referenced table is interesting. Note that Van is non-zero in some grounded wye secondary connections, but zero in all grounded wye primary cases (except delta secondary). This must mean that the single-line-to-ground fault that is referenced is a single-line-to-ground fault on the transformer primary. A single-line-to-ground fault on the transformer secondary would make Van zero for all grounded wye secondary connections.

IEEE Explore doesn't find any paper by those authors and title, but it does find Voltage sag analysis case studies by Lamoree, J.; Mueller, D.; Vinett, P.; Jones, W.; Samotyj, M.; Industry Applications, IEEE Transactions on, Volume 30, Issue 4, July-Aug. 1994 Page(s):1083 - 1089. This paper doesn't have the table but does discuss sags of customer systems caused by faults on the utility distribution and transmission systems. A quick read doesn't indicate any mention of the effect of transformer connection.

 

[stevenal]

The "real" fault current you speak of, for the faulted phase of an SLG, is I1+I2+I0 where I1=I2=I0. The other phases see no additional current from the fault.

Your infinite bus model maintains the line to line voltages and possibly the line to ground voltages. The unconnected transformer neutral is free to move about based on load or fault conditions. As LionelHutz said, not a very useful connection.

 

[waross]

There are a couple of points here that should be clarified.
Do not confuse grounding with connection to the neutral.
A ground fault is often a line to neutral fault with the grounding system forming part of the circuit, particularly at utilization voltages.

If the primary wye point of the transformers is connected to the neutral (but not grounded) then fault level currents will flow.

If the transformer wye point is not connected to the system neutral then in the event of a line to neutral/ground fault on the secondary, the transformer windings will act as a three phase voltage divider.
An easy example to follow and one with which most of us are familiar is a 120/240 volt secondary with an open neutral. The voltages in each 120 volt circuit are inversly proportional to the impedance when the neutral goes open. If there is a high wattage load such as a toaster on one line and a low wattage load such as a stereo set on the other line the resulting voltage across the toaster will be very low and the voltage across the stereo will be around 200 volts or more. The toaster will "Toast" the stereo.

The voltages, both secondary and primary will behave similarly. Current may only flow in the primary of the faulted phase if it also flows through one or both of the other phase windings.
The current resulting from the line to neutral fault will not be at fault levels. The currents on the unfaulted phases will increase as a result of the increased voltages.
If the increased voltage across the primaries of the unfaulted phases drives the transformers into the knee of the saturation curve, there will be increased current. This increased current will result in a higher voltage across the faulted phase and the system may be expected to find a voltage balance point with the unfaulted transformers in the knee of the saturation curve. The currents will still be at load or overload levels.

In the real world:
A wye/wye transformer with a floating primary wye point is only suitable for a balanced secondary load.
A wye/wye transformer with a floating primary wye point has no primary return path for harmonics. Overheating may be expected. In a large bank, the harmonics generated by the magnetizing current may be an issue.
A wye/wye transformer designed to operate with a floating primary wye point will have a delta connected tertiary winding that will hold the voltages balanced and fault current will flow in the primary, the secondary and the tertiary windings.
respectfully

 

[dpc]

http://www.geindustrial.com/publibrary/checkout/38652.30055.7762.44019/generic/GET-3388B.pdf

This old GE white paper (The Why of the Wyes) has some good background on wye windings.

If that link doesn't work, it is GE publication GET-3388B in their "White Paper" section.

 

[bigamp]

The performance of a three-phase STAR - STAR transformer (YNyn) with primary star un-grounded and secondary star grounded during an earth fault (or phase to neutral load) on the secondary side will depend upon the transformer type construction; shell or core.

For shell type, the zero sequence impedance will be very high, equal to the open circuit reactance (i.e. magnetising reactance) i.e. near enough open circuit so you will get very little earth fault current (i.e. effectively zero).

For core type the zero sequence flux passes through to the tank wall of the transformer and this acts as a tertiary. The effect is that the reactance will be somewhere in the range between 50% to 200% and you will most definately get earth fault current for an earth fault on the secondary side. There will be no zero sequence current on the primary side.

The effects are described in the J&P transformer book and also on pages 73 - 75 in a brilliant text "Electrical Power Systems Engineering Problems and Solutions" by Alvin Knable (McGraw Hill 1967). The GE paper referenced by dpc is a good read, well worth its suggested price of 50 cents (thanks for the link dpc), and it explains things as per above really well plus all sorts of other goodies.

The problem these days is that analysis software packages seem to model star - star transformers with primary star unearthed as having infinite zero sequence impedance hence no earth fault current. This is OK for a shell type but very wrong for a core type (how did we get by before there was software....). To get around this for a core type transformer you need to cheat a little.....either model it as a star - star with a fictitious tertiary winding or as a delta - star with the secondary neutral resistively earthed.

Core type star - star transformers without a tertiary do exist and are not all that uncommon. They owe their existance to the "best" of reasons...apparently they cost a bit less to manufacture than a delta - star of the same capacity (don't know exactly why...maybe you can use graded insulation on the primary?). Anyone ever contemplating buying such transformers had best get some additional testing done by the manufacturer to establish the zero sequence reactance so that the transformer may be adequately modelled.

 

[isquaredr]

Waross
I have just been to site to check a transformer rating plate on a 14.5MVA 33/11Kv transformer, Yyn0. It has a resistance grounded neutral with a value choosen (6.35-ohms) to allow a maximum of 1000amps earth fault current. There is no mention on the rating plate of a tertiary delta winding. I checked the value of earth fault current which flowed in to the last fault, 797A. According to the above a tertiary winding must be present for this current to flow.
Is it normal practice for the manufacturer to ommit imformation regarding tertiary windings on rating plates.

 

[jghrist]

isquaredr,

Is the primary neutral grounded? Or connected internally to the secondary neutral? If so, you don't need a tertiary to get ground fault current on the secondary, so long as the primary system has a ground source.

 

[isquaredr]

jghrist
No according to the rating plate and schematic diagram of both HV & lv windings there is no connection to the HV neutral, and therefore not grounded.

 

[stevenal]

isquaredr,

You need to look beyond the nameplate and look at the installation. You have four primary air bushings with H0 left unconnected? Or is it connected with three elbow style bushings with insulated cable?

 

[isquaredr]

The transformer in question is cabled both HV & lv, 3-connections HV, 4-connections lv (1-being the neutral which is grounded via a resistor)
I have also gone to many other sites within this group and all except one are identical, the odd one out being Dzn0, no mention at any site about a tertiary windings.

 

[jghrist]

isquaredr,

Perhaps the transformer is core-type, in which case the phantom tertiary as mentioned by dpc will give a zero-sequence impedance about 5 times the pos-sequence. This would still be small compared to the grounding resistor and could result in your measured fault current.

 

[stevenal]

If by "cabled" you mean insulated cable with elbows, each connector provides a connection to the concentric neutral.

 

[isquaredr]

jghrist
I accept that 5 times the pos-sequence value supplied via the phantom tertiary would be more than enough to source zero sequence current (ground fault) as our neutral resistor restricts ground faults to 19MVA max anyway.

Stevenal
All our connections at this voltage level are cabled, and I accept that the HV cable neutrals are bonded together but there is no connection what so ever to the star point of this winding, it is just not available.