Do I need a tertiary? 4

[stevenal]

This is related to my other question. I was incorrect on the voltage, neighbor has 24.9 kV, ours is 12.47. Both are grounded wye systems feeding unbalanced loads. The 60 degree phase shift can be accommodated with a Yy6 connection by choosing the proper bushings. The transformer size is 3000 kVA. Due to the size, I understand I need a delta tertiary. Do you agree? I understand the tertiary helps reduce tank heating and balances the loading somewhat, but have never seen guidelines based on capacity. Thanks.

 

[9wearSuave]

Depends. How much unbalance are we talking about?

 

[stevenal]

I expect this will be fuse protected. I would assume the worst: Operating near three phase capacity when one phase loses all load when a fuse opens.

 

[dpc]

I don't know the answer, but there are a lot of transformers in this size range on 480 V services that don't have tertiary windings.

 

[stevenal]

dpc,

Thanks. I take it those are gnded wye secondaries. I looked though our stock list, and it looks like we only go as high as 1500.

 

[RRaghunath]

If the transformer is expected to supply unbalanced loads, it does require a closed tertiary Delta winding to reduce unbalanced currents to circulate.
Tank heating happens because the unbalanced flux cannot circulate in a three limbed core and thus leaks on to the tank walls.
A 5 limbed core will allow the unbalanced flux to circulate.
Some times, the tank is shielded internally with blocks of metal that shunts the flux and thus prevents it from linking with the tank walls.

R Raghunath

 

[prc]

If YN and yn are solidly grounded, we can go without a stabilising tertiary winding.

 

[stevenal]

Both neutrals would be solidly grounded.

 

[davidbeach]

At that size, simply ordering a Y-Y transformer without a highly detailed spec could result in a transformer with a buried delta tertiary anyway. In many cases the manufacturer may may decide that the buried delta is less expensive than a 5-limb core.

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

 

[waross]

My experience is from the distribution perspective.
I understand that while the same principles apply, there are significant differences between transmission and distribution due to protection schemes and because loss of one phase is much less likely on a transmission circuit.
I spent about 15 years in an area where grounded wye/delta transformer banks were common, let me tell you about the possible issues.
How does a delta winding stabilize voltages?
First lets look at the difference between EMF and terminal voltage.
The terminal voltage is the EMF minus the internal voltage drop.
Second, lets look at the delta as represented by an equilateral triangle.
The with a balanced supply, the equilateral triangle may represent the delta EMFs, the delta or the delta terminal voltages.
The angles of the sides of the triangle are locked by the phase angles pf the grounded wye primary circuit.

What happens when the voltage is low on one phase?
If you draw the vector diagram, the triangle will not close on itself.
Let's state a 10% difference.
But how can we close the delta (closed physically by the delta connection) if the angles are locked at 120 degrees but the voltages are different?
The answer is by circulating currents.
The circulating current is limited by three times the transformer impedance.
Thus a 10% voltage unbalance on a 3% imp transformer will cause a current of 10%/9% = 111% of full load current to flow in the delta.
This current is in a direction to cause a voltage rise on the low voltage phase and a voltage drop on the full voltage phases.
So while the EMFs may differ from phase to phase, the terminal voltages will be equal due to the combination of internal voltage drops and rises.
I am skipping over the effects of voltage drops in the supply conductors and possible phase angle errors due to a voltage drop in the primary neutral conductor.
An unloaded grounded wye/delta transformer bank may be overloaded by a voltage %unbalance over 3 times the %imp of the transformer.

Let's hope that the tertiary impedance is higher than the working impedance, although the higher the tertiary impedance the less effective it is in balancing the voltages.

Without a tertiary, the plant motors try to act as the balance mechanism.
A effect related to the impedance of the motor interacting with the impedance of the supply conductors and the impedance of the supply transformer determine how effective the motor is at balancing the umbalace, but as more motors go online, they share the correction duties and the voltage unbalance at each motor is less.

A tertiary delta will reduce the unbalance and extra heating at the motors and replace it with extra heating of the transformer.

What is the downside of a grounded wye/delta connection?
Worst case example, your plant is on a shared distribution circuit.
One supply phase is lost to your plant and other customers.
Now the two healthy phases act as an open delta and back-feed the winding with the missing primary.
The transformer back-feeds to the primary circuit and now the failed phase of the distribution circuit is energized at both sides of the open circuit.
That back-feed will supply the missing phase to the other shared services on the primary circuit.
Your wye/delta transformer may be badly overloaded.
A grounded primary conductor may have the effect of 1/3 of a short circuit on the wye/delta transformer.

When considering a tertiary winding you may want to consider the following questions:
How likely is a loss of phase?
What is the impedance of the tertiary winding?
Will we have phase loss protection on the system?
Are other other loads on the distribution circuit that may be fed by the wye/delta in the event of a phase loss?
On the other hand, a lot of large motors may develop a similar back-feed until differential protection or overload protection takes them offline.

Oh, and by the way, with the loss of two primary phases, the two lost phases will be back-fed by approximately 50% of rated voltage.


--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[stevenal]

We use phantom tertiary's in larger auto transformers, as it make the protection of such easer. But buried tertiaries, is not that bad either. But either one will add a zero sequence source. The question might be if you need another ground source.

No ground source is desired, as it would de-sensitize the ground relay at the substation. Isn't a phantom tertiary just the tank, and causes heating?

 

[waross]

I thought that the phantom tertiary was the flux produced by the two healthy phases returning trough the core of the low or missing phase.
Not all returns through the core.
Some may return through the tank.
It is the flux returning through the core of the compromised phase that causes the phantom delta effect.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[cranky108]

Not all transformers are three leg cores. It is possible to have four and five leg cores.
The problem with a return path through the case is case heating, paint blistering, and additional losses. So this is only acceptable for short term events, like faults.
I am not an expert on transformers, so I don't know much about the internal design.

 

[dpc]

My understanding has always been that wye-wye transformers are built with five-legged cores as standard. Delta-wye may have three-legged cores. At least in pad-mounted xfmrs. So there should be a path for zero-sequence flux in the iron for wye-wye units.

 

[prc]

dpc, My understanding is the requirement of stabilising transformers may be there only for distribution transformers (having YNyn) due to unbalanced loads in phases. In India we are not providing stabilising tertiary in all transmission and sub-transmission transformers with star-star connection. We provide it when the core is five-limbed to reduce transformer zero sequence impedance! Of course, all distribution transformers have a delta-star (Dyn) connection. So, in India, we will not provide a stabilising tertiary in a sub-transmission transformer mentioned by Stevenal.

 

[wroggent]

Will there be any load served either between where the 12.47 source is and this transformer, or between where the 24.9 source is and this transformer. I.E., is this a tie somewhere out on two distribution circuits?

If yes, I think I'd go with the five limb Yg-yg.

If no, I think it would depend on if you wanted it to be a ground source, which you already so no, so probably still the above.

 

[stevenal]

Yes, load between source and tie point on both sides.

 

[stevenal]

My understanding has always been that wye-wye transformers are built with five-legged cores as standard. Delta-wye may have three-legged cores. At least in pad-mounted xfmrs. So there should be a path for zero-sequence flux in the iron for wye-wye units.

Based on a quick look on the Eaton site, I believe this is correct. Thanks.