wye -> delta transformer

[electrofelon]

Was playing with a 45 KVA 240 delta -> 208 wye transformer today that was hooked up backwards, before I took it out (this is the one from my thread from half a week ago where we had that meltdown due to a utility phase loss). So just curious about a couple things. With no load connected, and the XO (now primary) floating, I metered 3.5 amps on X1 and X3, and 5.5 amps on X2. IS this typical? Why is the B phase drawing more? When I jumpered the X0 to ground (metal raceway back to primary neutral), I got 40 amps flowing through the bonding jumper, and X1-X3 went up to 16 amps! Wow, I didnt think it would be that much. It was neat to play with this firsthand, now I really know why you must float the X0, although I am not clear on the electrical theory behind it. Does anyone want to take a stab at what exactly happens with the impedance of the windings when the X0 is bonded?

Oh and one addition question: I was curious how much energy this thing as been wasting. Any rough guesses assuming the figures I gave? Having pretty much balance on the phases but 40 amps on the neutral is really throwing me.......

 

[templdl]

I'll see if I can jump start your post this morning. Other than 45kva you didn't advise the temp rise, efficiency, non ventilated, ventilated, if it has a 'K' factor, etc. But a std eff 45kva ventilated 150degC rise transformer's expected no load losses would be about 400watts to about 700watts for 80degC and K13. As such this should account for the amps W/O the X0 connected of course.
When you think about it when you connect the X0 now you are applying a L-N voltage across each winding instead of the L-L voltage across two windings in series when the X0 in left floating which would result in the very high currents you have measured. With the X0 left floating the currents the you have measured should reflect the no load loss.
Does this make sense?

 

[mivey]

Were the source voltages balanced? What were they?

 

[Bugman1400]

Sounds like zero-seq current flowing to me.

 

[electrofelon]

I'll see if I can jump start your post this morning. Other than 45kva you didn't advise the temp rise, efficiency, non ventilated, ventilated, if it has a 'K' factor, etc. But a std eff 45kva ventilated 150degC rise transformer's expected no load losses would be about 400watts to about 700watts for 80degC and K13. As such this should account for the amps W/O the X0 connected of course.
When you think about it when you connect the X0 now you are applying a L-N voltage across each winding instead of the L-L voltage across two windings in series when the X0 in left floating which would result in the very high currents you have measured. With the X0 left floating the currents the you have measured should reflect the no load loss.
Does this make sense?

Yes, but how come this works with a wye wye transformer?

Were the source voltages balanced? What were they?

About 208 plus or minus a few volts

 

[mivey]

About 208 plus or minus a few volts

Then you have a grounding bank with X0 bonded and high circulating currents in the delta.

 

[mivey]

Does anyone want to take a stab at what exactly happens with the impedance of the windings when the X0 is bonded?

Nothing happens with the impedance. You are just creating a bunch of voltage drop across the windings as the transformer is trying to equalize the source voltages.

 

[mivey]

With no load connected, and the XO (now primary) floating, I metered 3.5 amps on X1 and X3, and 5.5 amps on X2. IS this typical?

I would expect them to be more equal than that with no load.

 

[mivey]

Oh and one addition question: I was curious how much energy this thing as been wasting. Any rough guesses assuming the figures I gave?

Without X0 bonded I'm going to guess about 1kW. For X0 bonded I'm guessing 2-3 times that.

 

[GoldDigger]

Without X0 bonded I'm going to guess about 1kW. For X0 bonded I'm guessing 2-3 times that.

I think it is a lot worse than that.
Unconnected and unloaded the current is mainly magnetizing current at low power factor. The circulating currents, on the other hand, are mostly resistive. As a result I would guess that the losses could be 10-40 times as high with the wye point connected and the degree of voltage imbalance the OP reported.
Again, the most immediate danger is that the neutral conductor or bonding jumper connected to the wye point can be subjected to more than the full load current of the transformer and at least twice the OCPD current of the phase lines.

 

[kwired]

If voltages are a few volts difference you will start to throw the current balance off.

On top of that if this is the transformer from the other thread as mentioned - any overheating that may have happened could result in turn to turn shorts and will change impedance of those sections of the windings further changing current balancing.

As mentioned with no load majority of current drawn is low power factor reactive current, only the much smaller resistive components are drawing any true power.

 

[electrofelon]

Thanks all for the discussion. What I dont understand is with the XO bonded, how could I be reading about 16 amps on each phase (pretty close to equal on each) and 40 amps on the neutral/XO bonding jumper? That literally doesnt add up with phases 120 degrees apart. Would just like to understand what explains how I can have such a high current on the neutral with nearly balanced phase currents.

 

[electrofelon]

Sounds like zero-seq current flowing to me.

want to take a stab at explaining that? My theory is a little weak in this area. Thanks!

 

[GoldDigger]

Thanks all for the discussion. What I dont understand is with the XO bonded, how could I be reading about 16 amps on each phase (pretty close to equal on each) and 40 amps on the neutral/XO bonding jumper? That literally doesnt add up with phases 120 degrees apart. Would just like to understand what explains how I can have such a high current on the neutral with nearly balanced phase currents.

The answer is basically that the circulating currents are driven directly by the voltage imbalance rather than the overall voltage. And the imbalance voltages, and thrrefore currents, are close to being in phase rather than at 120 degree angles.

 

[GoldDigger]

I have posted more detailed explanations in earlier threads and and not in a position to retype them right now.

 

[Bugman1400]

want to take a stab at explaining that? My theory is a little weak in this area. Thanks!

It seems you have single phase loads on the same voltage level as the grounded-wye xfmr. The imbalance of those loads will cause zero-seq current to flow. Since you grounded the X0, your xfmr became a zero-seq source for the current. There is also zero-seq current coming from the mains as well but, your grounded X0 will share the duty.

 

[SparkyHC]

Does anyone want to take a stab at what exactly happens with the impedance of the windings when the X0 is bonded?

Theoretically speaking Inductive reactance "ZL" = 2*pi*f*L where f is frequency. Notice in the first 1/60th of a second "ZL" is equal to zero. (ignoring equipment and cable impedance) When you turn on a WYE winding
with the center-tap connected to ground you get what starts out as a short circuit but quickly reduces in a second or by the 60th cycle. The amount of current generated in the first few cycles can melt a ground wire, I've seen it happen.

 

[Bugman1400]

Theoretically speaking Inductive reactance "ZL" = 2*pi*f*L where f is frequency. Notice in the first 1/60th of a second "ZL" is equal to zero. (ignoring equipment and cable impedance) When you turn on a WYE winding
with the center-tap connected to ground you get what starts out as a short circuit but quickly reduces in a second or by the 60th cycle. The amount of current generated in the first few cycles can melt a ground wire, I've seen it happen.

Sparky, are you describing transformer inrush here?

 

[GoldDigger]

It seems you have single phase loads on the same voltage level as the grounded-wye xfmr. The imbalance of those loads will cause zero-seq current to flow. Since you grounded the X0, your xfmr became a zero-seq source for the current. There is also zero-seq current coming from the mains as well but, your grounded X0 will share the duty.

The way that I look at it is best illustrated by considering a single phase 120/240 service. If the two legs are not equal in line to neutral voltage for any reason, consider what happens when you connect a single center tapped winding to L1 and L2 only. The center tap will be halfway between L1 and L2 in voltage, while the neutral is not.
When you connect the center tap to the neutral current will flow until the IR drop in the two hot leads and the neutral equalizes the voltages. You do not need to consider it as supplying current to any unbalanced loads in the grid, although that may well be the origin of the voltage imbalance. The same current will flow if the origin of the voltage imbalance is a mismatched secondary tap on a transformer somewhere. If you do not connect the center tap an the 240V primary you can get a balanced 120/240 on a center tapped secondary.
It is slightly more complicated in the case of three phase.
By itself a wye primary connected to the unbalanced service will not draw unusual current, whether the wye point is connected to neutral or not. Nor will adding a wye secondary change things.
But if you add a delta secondary to the transformer suddenly the vectors of the transformed phase voltages must form a closed triangle, a vector sum of zero.
To make this happen, again current will flow in a loop around the delta and corresponding currents (according to turn ratios) will flow in the primary coils.)
Those currents will include a substantial component in the neutral.
If you do not bond the wye point on the primary it will float to a voltage such that the line to center voltages are all balanced. There will be no circulating currents in the delta secondary because there is no voltage imbalance. And there will be no anomalous currents in the primary neutral because it is not connected to anything. Just as in the single phase example the center point and the service neutral will end up at different voltages to ground or other references

 

[Bugman1400]

The way that I look at it is best illustrated by considering a single phase 120/240 service. If the two legs are not equal in line to neutral voltage for any reason, consider what happens when you connect a single center tapped winding to L1 and L2 only. The center tap will be halfway between L1 and L2 in voltage, while the neutral is not.
When you connect the center tap to the neutral current will flow until the IR drop in the two hot leads and the neutral equalizes the voltages. You do not need to consider it as supplying current to any unbalanced loads in the grid, although that may well be the origin of the voltage imbalance. The same current will flow if the origin of the voltage imbalance is a mismatched secondary tap on a transformer somewhere. If you do not connect the center tap an the 240V primary you can get a balanced 120/240 on a center tapped secondary.
It is slightly more complicated in the case of three phase.
By itself a wye primary connected to the unbalanced service will not draw unusual current, whether the wye point is connected to neutral or not. Nor will adding a wye secondary change things.
But if you add a delta secondary to the transformer suddenly the vectors of the transformed phase voltages must form a closed triangle, a vector sum of zero.
To make this happen, again current will flow in a loop around the delta and corresponding currents (according to turn ratios) will flow in the primary coils.)
Those currents will include a substantial component in the neutral.
If you do not bond the wye point on the primary it will float to a voltage such that the line to center voltages are all balanced. There will be no circulating currents in the delta secondary because there is no voltage imbalance. And there will be no anomalous currents in the primary neutral because it is not connected to anything. Just as in the single phase example the center point and the service neutral will end up at different voltages to ground or other references

What would happen for the 3 single phase loads if the voltage was exactly the same on all three phases but, the loads were all different? Would you expect to have zero-seq current or not?