Grounding of stabilizing Tertiary winding. 1

[prc]

Normally one corner of delta connected stabilizing tertiary winding is grounded in service or neutral grounding achieved through grounding transformers or star connected electromagnetic PTs connected to tertiary terminals. Latest IEC 60076-3 Insulation levels of Transformers proposes as below:
B.3 Power-frequency transferred overvoltage
If a low-voltage winding which is physically adjacent to the high-voltage winding is left without connection to earth or with only a high-impedance connection to earth while the high-voltage winding is energised, there is a risk of power frequency overvoltage by capacitance division.

The risk is obvious for a single-phase winding, but it can also exist for a three-phase winding if the primary winding voltage becomes asymmetric, as occurs during earth faults. In particular circumstances, resonance conditions may arise.

Tertiary windings and stabilizing windings in large transformers are also subjected to the same
risk. It is the responsibility of the purchaser to prevent a tertiary winding from being accidentally left with too high an impedance to earth. A stabilizing winding should normally be arranged for permanent connection to earth (tank) either externally or internally.

The overvoltage is determined by capacitances between windings and between windings and
earth. These can be measured at low frequency from the terminal of the transformer in
different combinations, and they can also be calculated with sufficient accuracy. Unquote

But grounding of one corner of delta has a problem esp in large EHV auto banks where the bus work on bus support insulators will be substantial for forming the closed delta. In case of one more grounding(through bus insulators etc) it will form a LL fault on tertiary. My query is- with enhanced tertiary insulation (say 52 kv in 33 KV tertiary) can we avoid the corner grounding of delta tertiary? How is the practice in other countries?

 

[Power0020]

Without grounding one corner the voltage can go floating and exceed winding BIL causing uncontrolled flashover.

The practice with tertiary winding is usually to ground one corner and the other two corners are left open with a shunt surge arrestor to ground.

In case of large single phase units transformers, it is common to have a "delta" bus hanging over the transformer units, where the tertiary is being connected, a bit complicated but works well.

I have seen a substation that the delta tube bus was supported on transformer units firewalls, with a 33 kV insulators, it is much easier than a strung bus with extra supporting structures.

You may ground two separate winding points to form the grounded corner using the ground grid. the other two will need to be connected with the high profile delta bus with shunt surge arrestors.

 

[prc]

Actually I had checked the surge transfer and it is well below the 250 kV,the BIL for 52 kV. In such a case, is it permissible to leave tertiary floating?

 

[Power0020]

No, again the transferred voltage isn't the cause of floating voltage rise, the floating voltage rise may take place due to any stray induction, this is very similar for the need to ground the current transformer secondaries start point just to avoid stary induction coming from anywhere is an HV substation.

The tertiary transferred voltage is useful in selecting the tertiary winding BIL and Surge Arrestors accordingly, given that one corner is grounded to "create" a ground reference for the potential.

 

[FPelec]

We do not use tertiary winding in our standard EHV/HV autotransformer since the 70's.
The oldest autotransformers, which are 3-phase units made in the 60's and still in service, have a delta tertiary winding (Um=24 kV) with a delta corner permanently connected to the ground.
However, for some recent oversea projects, the local TSO explicitly requests for a delta connected tertiary winding (Um=36 or 72 kV) for single-phase EHV/HV autotransformers, with single phase units.
In the latter case the delta is closed through external connection on auxiliary busbars (in order to allow for the fast switching of the spare unit) and it is not possible to solidly connect a delta corner to the ground.
We assessed the possible induced power frequency voltage on tertiary winding and we found that, with a 72 kV insulation level, there are no risks.
It is thus possible to keep the tertiary floating.
However, ferroresonance shall be carefully checked, due to presence of inductive voltage transformers.
A possible solution could be to use an auxiliary transformer for substation load (e.g. 36 kV / 0.4 kV with Zy connection) , with a proper winding connection to allow for a high impedance neutral grounding on the tertiary winding side; this would allow for an overvoltage mitigation and would avoid high current faults.


Si duri puer ingeni videtur,
preconem facias vel architectum.

 

[prc]

Thank you Fpelec for detailed info. You are referring to which area? US? When you say no tertiary with autos means 3 phase 3 limbed autos or with single phase banks also?
In case of banks, you are referring to 33 kV Tertiary with 72 kV class insulation?"We assessed the possible induced power frequency voltage on tertiary winding and we found that" -Can this method of calculation be explained a bit more. Believe this power frequency over voltages cannot be mitigated by lightning arresters. Any surge absorbers or LAs provided for tertiary protection? VTs are provided for ground fault detection through broken delta secondary? Instead of auxiliary transformer,if VT primaries are connected in star and grounded, will not serve grounding purpose?

 

[prc]

I got answer to part of my query in a draft IEEE Standard PC 57.158 D6A -2016 Guide for the Application of Tertiary and Stabilizing windings in Power Transformers.

 

[JensenDrive]

Grounding of one corner means the second ground on another phase would produce a major LL fault, as you seem to mention. Seems a bad idea. I would recommend always using the VT approach to provide ground reference.
I recall a case where a single line to ground VT was used for the ground reference, which I suppose would do the trick if $ were an object.

 

[FPelec]

prc, I refer to Western Europe(and particularly Italy ).
All our recent autotransformer (i.e. made after the '70s) are made without a delta connected tertiary; they are generally 3 phase - 3 limbs units, but we also use, especially for oversea projects, 3-phase, 5 limbs and single phase units. Of course, in all these cases the EHV system neutral is solidly grounding.
The induced overvoltage on the tertiary is because of both capacitive and inductive coupling, the first being often the most significant.
The induced overvoltages have been assessed for oversea projects, with single phase units, both by calculation (FEM) and field tests. We never found troublesome overvoltage, if the tertiary winding rated voltage is 72 kV or higher.
In some particular winding arrangements (when the regulation is not the innermost winding) and with a 36 kV rated tertiary we found some risk of overvoltages. This overvoltage risk can me mitigated by Surge Arresters.



Si duri puer ingeni videtur,
preconem facias vel architectum.