Vault Transformer Cabling

[JoeStillman]

In this scenario, we have three single-phase oil-filled transformers installed in an above-grade vault by the utility company. The primary cables are installed and terminated by the POCO, secondaries by the owner. The transformers have the usual four spades on the secondary, X1, X2, X3, X4. The transformers feed a 4000A switchboard with 9 runs of (4) 750 kCM. Because the transformers are arranged in 208Y/120V secondary, the X1-X3 and X2-X4 lugs are paralleled on each transformer.

On each transformer, how many cables do we have to install between X1-X3? or X2-X4? The outgoing cables can be arranged with 4 on X2 and 5 on X4. Is a single jumper enough to ensure equal current in the coils? I've seen it done with a single jumper and also with 4 jumpers on each phase. Is there some standard I can refer to for advice?

 

[Elect117]

Single phase current of the transformer.

XFMR VA / 120

That is the max current through that wire (other than a fault).

It works out to be the same as the 3PH current of the 3 transformers. So typically the same size wire as the secondary feeder.

It could theoretically be reduced to that of the same size as the neutral wire but I don't know if people actually reduce it.

 

[electrofelon]

Single phase current of the transformer.

XFMR VA / 120

That is the max current through that wire (other than a fault).

It works out to be the same as the 3PH current of the 3 transformers. So typically the same size wire as the secondary feeder.

It could theoretically be reduced to that of the same size as the neutral wire but I don't know if people actually reduce it.

Wouldn't it be half the FLC? Here is what SCL does: you can't see because of the insulation, but it's a braided copper jumper. Not sure if these will provided with the transformers or added in the field.

 

[Elect117]

Wouldn't it be half the FLC? Here is what SCL does: you can't see because of the insulation, but it's a braided copper jumper. Not sure if these will provided with the transformers or added in the field.

You could be right. The only way to know for certain is a photo of the nameplate. One of us is probably right.

Also, as you said, the connection between X1-X3 might also be able to be a reduced ampacity for other engineering considerations. Like that connection being the grounded / neutral conductor.

As I read the question again, the connection between X2 and X4 at the transformer, was to ensure that the current was equal in both windings (4 wires vs 5 wires) or for bonding the windings of X2 and X4?

If X2-X4 were the same potential and all the wires were moved to the X4 paddle, than I would say the wires between X2 and X4 should be equal to the number of runs landed on the single paddle. That is why I worded my response like that.

But to be honest, I don't know anymore. I might be overthinking it but more information about the set up might help.

 

[electrofelon]

You could be right. The only way to know for certain is a photo of the nameplate. One of us is probably right.

Also, as you said, the connection between X1-X3 might also be able to be a reduced ampacity for other engineering considerations. Like that connection being the grounded / neutral conductor.

As I read the question again, the connection between X2 and X4 at the transformer, was to ensure that the current was equal in both windings (4 wires vs 5 wires) or for bonding the windings of X2 and X4?

If X2-X4 were the same potential and all the wires were moved to the X4 paddle, than I would say the wires between X2 and X4 should be equal to the number of runs landed on the single paddle. That is why I worded my response like that.

But to be honest, I don't know anymore. I might be overthinking it but more information about the set up might help.

If you are just paralleling two windings, then each does half the KVA right?

 

[Elect117]

If you are just paralleling two windings, then each does half the KVA right?

Ya. That is the theory. Each winding is 1/2 the VA's FLA and when combined, they are the FLA. Where series type increases the voltage and therefore reduces the current.

But like you said, it might not even need to be connected with 1/2 of FLA if it is just to work as a bonding at the transformer between the phases and the actual current is on the 4 + 5 runs. At that point you want to reduce voltage gradient in the two runs and any possible impedance differences at the far end. A single conductor would probably suffice to make that work. Sizing of it, would probably need to be more of a impedance matching than a current or ampacity thing.

My original comment was not not a good answer. It was oversimplistic to the question. I appreciate you calling that out.

 

[electrofelon]

Ya. That is the theory. Each winding is 1/2 the VA's FLA and when combined, they are the FLA. Where series type increases the voltage and therefore reduces the current.

But like you said, it might not even need to be connected with 1/2 of FLA if it is just to work as a bonding at the transformer between the phases and the actual current is on the 4 + 5 runs. At that point you want to reduce voltage gradient in the two runs and any possible impedance differences at the far end. A single conductor would probably suffice to make that work. Sizing of it, would probably need to be more of a impedance matching than a current or ampacity thing.

My original comment was not not a good answer. It was oversimplistic to the question. I appreciate you calling that out.

Yeah in an absolute perfect world you connect half of the total load to each spade and don't even run a jumper between them. I was just saying that, I think, worst case for your jumper would be half of nameplate A

 

[JoeStillman]

If I have an odd number of cables running from the transformer, there will more of them on one spade than the other. I think that could have a significant impact on how the winding-currents divide.

IMHO, the ampacity of whatever you use for a jumper should come from the "free air" table. But what is my target ampacity? It's only going to carry a few amps.