Vector Group 2

[Gestur3]

Hello,

Have a transformer with Vector group Yd5. Line come in with C-B-A instead of normal A-B-C phase order.
On terminal A arrives L3, on B arrives L2 and on C arrives L1.

How does this affect the Vector group? Does this transform to Yd7 or Yd1? Or will it remain Yd5?

Three engineers and three opinions, need some assistance to set transformer protection.
(Personally I think Yd7 or maybe Yd1, but not Yd5=)

BR Gestur

 

[dpc]

I live in ANSI land, but the phase shift will flip 60 degrees, so I'd assume Yd7. It will definitely not stay the same (I've learned that the hard way).

 

[Marmite]

You would need to check the LV connections as well. If A&C are crossed on the secondary the same as they are on the primary then your transformer will act as a Yd7. If not then you are just changing the phase rotation.
Regards
Marmite

 

[LiteYear]

Heh, this seemed like a straight forward question. Either it's Yd7, Yd1 or Yd5 right? Well, I found that as I went to explain it with voltage vectors and individual phase transformers, I quickly lost track of where my "zero" reference is. Of course, these "Yd#" abbreviations only work when you have a reference. So to make things completely clear, I turned to good old, reliable simulation. The results are quite as I expected, but much less ambiguous to explain:

See the attached image. The simulation consists of a 3-phase voltage source scaled to reach 100V peak and oscillate once every 12 seconds. The 12 seconds allows me to put units on the # in Yd# - in this simulation there are 12 "seconds" to a full rotation. The top transformer is a normal Yd5, and the load weakly references the secondary symmetrically about the 0V reference.

The top plot shows primary voltages A and B for the top transformer. Phase A's rising zero crossing is at 0 seconds and phase B follows it at 4 seconds.

The next plot shows secondary voltages a and b for the top transformer. Phase a crosses at 5 seconds and phase b follows it 4 seconds later. No surprises there, normal Yd5 behaviour.

The third plot shows the voltages at the primary terminals of the bottom transformer. In this case phases A and C have been swapped, while everything else remains the same. So phase A now crosses at 8 seconds and phase B crosses 4 seconds earlier since the phase rotation is backwards.

The final plot shows the secondary voltages of the bottom transformer. Phase a now crosses at 3 seconds, which is 7 seconds after phase A (or equivalently, 5 seconds before, but we'll stick to the positive convention for vector groups). Phase b crosses 4 seconds before phase a since the secondary reflects the phase rotation of the primary.

In summary, crossing A & C on the primary of a Yd5 will, provided the terminals marked A, B, C and a, b, c do not change, perform as if the transformer was a Yd7 and the phase rotation will be backwards. This makes sense once you study the voltage vectors across individual transformer windings, but I found it much more convincing in simulation.

 

[sibeen]

OK, I get something different again.

Yd9?

Diagram attached. Where did I stuff up?

 

[sibeen]

I should have mentioned on the diagram, in the transformer with the A & C phase swapped I get:

a-n lags A-n by 270 degrees
b-n lags B-n by 150 degrees
c-n lags C-n by 30 degrees

If a & c were also swapped at the secondary, the transformer would stay a Yd5, but the phase rotation would be c-b-a.

 

[GTstartup]

I don't get it, if you swap the primary you change the phase rotation on the secondary, the vector group will be the same. If you swap the primary and the secondary, you have effectively done nothing and the vector group will also be the same.

 

[sibeen]

GTstartup, not so.

If you do the same thing - swapping the same two phases on the primary and secondary - on a Dy1 transformer you end up with the equivalent of a Dy11 connection, and vice versa.

 

[dpc]

This topic has come up before and we even have a FAQ: faq238-1154

If you are not changing the internal connections, it's pretty straightforward if you just take the phasor diagram on the transformer nameplate and apply your actual system phase sequence to the appropriate transformer terminals.

 

[LiteYear]

Diagram attached. Where did I stuff up?

Same place I did before doing the simulation - you've double dipped on the swapping of the terminals. You've changed which phase is connected to which terminal, but then also changed which terminal you use as the reference. I don't think the result is wrong, it's just that by referencing to a different terminal the number comes out different. The overall effect is the same.

I actually like the convention suggested in the FAQ:

So was the DY11 converted to a DY1 by means of changing the external connections? If you look at DY? as referring to the system phases, yes. I prefer to see the DY? designation as an indication only of the internal connections. I can then more easily see that any DY transformer can be installed so the secondary is + or - 30 degrees, + or - 90 degrees, or + or - 150 degrees.

In our case, the Yd5 transformer is still a Yd5 transformer, but by changing the ordering of the phase connections the rotation is flipped and it behaves like a Yd7 (with flipped secondary rotation). If you go one step further, as sibeen did, and consider where the phases go, then it looks like Yd9! At the end of the day the behaviour is the same, it's just that the convention (and reference) has changed.

By the way, thanks for pointing out the FAQ dpc. It's a pity the referenced Basler document seems to have disappeared from the Internet.

 

[sibeen]

It's a pity the referenced Basler document seems to have disappeared from the Internet.

LiteYear, fair dinkum, nothing is ever lost on the internet

Especially when dumb mongrels like myself made sure we downloaded the document years ago.

See attached - about a 1.3 meg file.

 

[stevenal]

Thanks. I've updated the FAQ to point to the engineering.com file.
If you like the FAQ, don't forget to rate it; so far I'm only scoring a 6 out of 10. And if you have suggestions, there is a comment form.

 

[LiteYear]

You beauty. Thanks mate.

 

[odlanor]

Marmite:
"You would need to check the LV connections as well. If A&C are crossed on the secondary the same as they are on the primary then your transformer will act as a Yd7. If not then you are just changing the phase rotation."


LiteYear:
"crossing A & C on the primary of a Yd5 will, provided the terminals marked A, B, C and a, b, c do not change, perform as if the transformer was a Yd7 and the phase rotation will be backwards."

crossing A & C on the primary of a Yd5 and secondary do not change
Marmite=> transformer was a Yd5 and the phase rotation will be backwards
LiteYear=> transformer was a Yd7 and the phase rotation will be backwards

Please clarify my conclusion!

 

[LiteYear]

Well confusingly enough, Marmite and I are saying the same thing, just from a different perspective. Note that 5 thirty degree rotations in the forward direction is the same as 7 thirty degree rotations on the other direction. My notation is from the perspective of the primary's (original, pre-swapped) rotation, Marmite's is from the secondary's rotation (or equivalently, the "new" primary's rotation due to the swapping of phases).

 

[odlanor]

Marmite,
below my interpretation of your post for setting relay


Siemens Transformer Relay Parametrization (extract):
270 Rated Frequency 60 Hz
271 Phase Sequence L3 L2 L1
313 Starpoint of Side 1 is Earthed
314 Transf. Winding Connection Side 1 Y (Wye)
323 Starpoint of Side 2 is Isolated
324 Transf. Winding Connection Side 2 D (Delta)
325 Vector Group Numeral of Side 2 5

 

[odlanor]

attached assistance to set transformer protection