Banking Single Phase Transformers

[WoodyM]

Can you bank three single phase transformers in a delta to grounded wye with one transformer having a different impedance?

 

[rbulsara]

I would think so, especially if they are not fully loaded. How much different impedance? I am not sure what would exactly happen though. Haven't ran into one yet. As long as voltage and current remains within the rating, it should be OK.


Rafiq Bulsara

http://www.srengineersct.com

 

[mgtrp]

It is possible, but you will get circulating currents. From my slightly sketchy memory, I believe that we would limit the difference between transformers to no more than 10% of the impedance value. You wouldn't want to try this on a heavily loaded transformer bank that is already running hot.

 

[waross]

A four wire wye:delta will often have serious circulating currents.
A delta:wye is basically three single phase circuits. No circulating currents.
For unity power factor loads compare the regulation numbers rather than impedance. As the load changes you may have slight phase angle errors. Any motor loads will sacrifice some of their thermal capacity in an attempt to balance any unbalanced voltages and to correct phase angle errors.
It depends, a small impedance difference may not cause issues but an impedance difference large enough to cause a few percent voltage differences will cause hot running motors as the motors try to correct the errors.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[stevenal]

"A delta:wye is basically three single phase circuits. No circulating currents".

Did you mean to say wye:wye? The OP's delta:wye will certainly have any secondary zero sequence current transformed to the primary where at will be confined to circulate around the delta. Not a problem, as long as current is within the individual transformer ratings. The lower impedance unit(s) will hog the load, so this bank will not be able to run to a capacity based on the sum of the individual units.

 

[magoo2]

You certainly can bank 3 single phase transformers in a delta grounded-wye connection. We did so with a GSU some years back when we had a failure of one unit and the nearest available unit had a different impedance.

You can model this situation using symmetrical components. Check out open conductor faults with impedance.

You'll probably want to model this so you can see how much negative sequence current you get and to make sure the settings of the negative sequence relays permit this.

 

[waross]

A single phase load or fault will appear on the primary as a line to line load or fault.
I may have misspoke as part of the harmonic currents may circulate in the delta but that is the nature of harmonics. It doesn't matter if the harmonics are present on the line, or generated by your load. If harmonics are present, the mismatched impedance will not change that or in itself cause harmonics. The impedance mismatch may have a slight effect on the amplitude of the harmonic current.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[stevenal]

Waross,

Harmonics are not needed to have circulating zero sequence primary delta current, just unbalanced secondary fundamental loading will work just fine. Yes they do not reach the primary conductors outside the delta where it looks like line to line current. Like Charlie they doomed to ride forever 'neath the streets of Boston within the delta.

'later

 

[waross]

I am fairly familiar with the circulating currents that occur in four wire wye:delta banks. I am aware of the phantom delta effect found with three legged cores, but this is three single phase transformers.
If I hook a transformer line to line on a delta system I can draw single phase current from the secondary. I am at a loss as to how doing this three times with only one common point on the secondary windings will cause circulating current.
I hope that I am not about to learn something new. (Yes it happens and it's painful).


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[davidbeach]

Assuming three transformers connected delta-wye, with the wye grounded.

If there is a zero-sequence component to the current on the wye side, there will be a corresponding zero-sequence current circulating in the delta of the delta side. It can't leave the delta, so it never appears at the terminals, but it can certainly be measured by CTs in the delta. This is the same for three singles as it is for a single 3-phase transformer.

Put the three singles in a black box, put a comparably rated 3-phase in another black box. Three wires into the delta side, four wires out of the wye side in both cases, with the wye-point wire grounded in both cases. You'd be hard pressed to determine which black box contains which type of transformer. If the 3-phase unit were of shell construction it could be impossible. If the 3-phase unit were of core construction there will be a difference between Z1 and Z0 for the 3-phase and not of the three singles.

 

[prc]

It is really interesting to think over. What is the effect on no-load and under loaded condition with delta/grounded whye with different impedance in phases.In india we had an experience with a 600 MVA auto bank with stabilising tertiary.When one of the single phase units was replaced with a unit with different tertiary impedance,but with the same HV-LV impedance,heavy circulating current was noticed in tertiary.(HV-try in one unit 60% and two units 45 %)

 

[waross]

Hi fellows, help me here, please. No sarcasm intended, i respect both Steven and David.
I am visualizing a delta distribution system. The system carries the secondary neutral throughout the system.
On Maple street, there is a transformer connected from A phase to B phase. The secondary has the neutral point grounded. There is a line to neutral fault on the secondary. There is a corresponding current in the primary A and B lines.
If this is fed from a delta winding back at the sub, the load current will divide about 50/50 between A-B and A-C-B. This is not a circulating current, it may be measured on the primaries of the three windings.
But no circulating current in the single phase primary.
Now if we connect a similar transformer over on Bay Street from line B to line C and on Green Avenue we connect another similar transformer from C to A.
Now, if we have a line to neutral fault on Maple Street will we get a circulating current between Maple Street and Green Avenue by way of Bay Street?
This is the same circuit in that three transformers with a common connection on their secondaries are connected to the three phases of a delta supply system.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[davidbeach]

Probably. Would also depend on how those separated transformers are connected. If the A-B is always H1 on A and H0 on B, while the B-C is H1 on B and H0 on C and likewise H1 on C and H0 on A you will get that circulating current.

The source, delta or wye, doesn't matter, wants to supply the fault current equally from all three phases. The Maple St transformer has current flowing from A-B. Some of the ground return will arrive at the X0 of Bay St, and current will flow from B to C. Some of the ground return will also arrive at Green Ave and primary current will flow C to A.

Some circulating current. One limiting case is that there is no impedance between the three X0 terminals. The other limiting case would be infinite impedance between the X0 terminals.

With no impedance we have the case described in my previous post. In the infinite impedance case there would be no circulating current as only one X0 would be participating in the fault. Any real case will be somewhere in between and the amount of circulating current will vary.

 

[stevenal]

Waross,

The situation you describe is somewhat different from the OP's, since your single phase secondaries are likely grounded mid-winding. Offhand I don't know the answer. Never analyzed a six phase system. To make it simple, let's ground each winding end so that the three secondaries form a three phase wye. Yes, circulating current will flow in the primary wires that connect the three streets.

For a graphical approach, draw your delta wye, show 1pu current leaving each wye leg in phase to represent the zero sequence currents. On the delta side you will have 1pu in each winding, all in the same direction. Summing the currents to zero at each delta node shows that no contribution from the primary lines are needed. If the nodes are remote from any of the windings, there is little difference.

 

[jghrist]

Let's simplify things a bit. Three transformers, one connected from A-B, one from B-C, and one from C-A. Single phase secondaries with one side connected to a grounded neutral, common to all three transformers. If these transformers were all in one place, they would be a delta-grd Y bank. If they were in different places with negligible impedance lines, the situation would be essentially the same.

Say there is only load on one of the transformers. There is clearly only current in one of the primary windings, right? What is with this circulating current??? It is a mathematical construct. In two of the transformers, the zero-sequence current is completely balanced by positive and negative sequence current. The sum of the three sequence currents equals zero.

 

[davidbeach]

I think stevenal may have hit the nail on the head. If Bill's three transformers are all center-tapped on the low side with the center tap grounded, then most of the discussion wouldn't apply.

 

[waross]

David, ignore one single phase winding on each secondary. Then we have a three phase wye. Now consider a line to ground short on one transformer.
The primary of this transformer is a single phase load on the delta supply. On the one hand it will work just fine. Other single phase loads will work fine on the other phases. If we add transformers on the other phases, we expect to see any secondary current reflected accurately in the primary. I have never heard of or seen primary currents that were not in reasonable proportion to the secondary current that were attributed to circulating currents. That would be the case if there were any transformers anywhere in the system on the other phases.

For a graphical approach, draw your delta wye, show 1pu current leaving each wye leg in phase to represent the zero sequence currents. On the delta side you will have 1pu in each winding, all in the same direction. Summing the currents to zero at each delta node shows that no contribution from the primary lines are needed. If the nodes are remote from any of the windings, there is little difference.

Steven, I like and respect you. But, respectfully I have to throw the flag on this one.
A transformer bank with full load on the secondaries and no contribution from the primary lines?? Is this patentable?
I was starting to doubt myself, but now, considering the high caliber of the other parties to this discussion, I suspect a massive misunderstanding of some sort.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[davidbeach]

Bill, I think its a matter of Apples, Oranges, and Kumquats. All of my discussion dealt with two winding transformers where the there is only a connection at each end of each winding. I've really no idea what happens when you ground a center tap and then fault from one end of a winding to the middle of the same winding. Without doing a whole lot more work than I feel like at the moment, I have no idea even where to start to analyze the situation you presented 21 Jan 11 0:19 if you were in fact talking about transformers with center-tapped low sides.

 

[jghrist]

Steven, I like and respect you. But, respectfully I have to throw the flag on this one.A transformer bank with full load on the secondaries and no contribution from the primary lines?? Is this patentable?I was starting to doubt myself, but now, considering the high caliber of the other parties to this discussion, I suspect a massive misunderstanding of some sort.

You're being too harsh on Steven. He was discussing only the zero-sequence currents. The primary lines do not contribute zero-sequence currents. As I noted in my previous post, you can have zero-sequence currents in a winding while the total current in the winding is zero. Similarly, you can have current in the line without zero-sequence current in the line.

The problem with analyzing three separate transformers connected phase-to-phase in separate places with symmetrical components is that you are using a mathematical device applicable to three-phase systems, not three separate single-phase systems.

 

[waross]

Well either ignore one winding on each transformer or connect it in parallel with the other winding as a typical 120:208 Volt bank of single transformers.

Let's look at it another way.
Connect a transformer A phase to B phase.
Connect a transformer B phase to C phase.
Connect a transformer C phase to A phase.
Leave the secondaries disconnected and ungrounded. What would cause a circulating current in the primaries?
Although the magnetizing currents may form a delta vector diagram and appear to sum to zero, the magnetizing currents are still fed from the primary lines.
Any load on any transformer will have a corresponding primary current. I don't see any reason for any circulating current at the fundamental frequency. I can't see how a load on one transformer secondary will have any effect on the other transformer primaries.
Now, any connection of one end of any winding will not cause any current to flow in that connection. The connection may be to another winding, to ground, or to both, it will not by itself cause any change in winding currents or voltages as it is open circuited at the other end of each winding.

I can see the phantom delta in a three legged core of a three phase transformer causing circulating currents and I have many times seen a four wire wye:delta bank develop circulating currents to the point of self destruction.
In a delta wye with three independant transformers there is no need for the voltage delta to close on itself. The delta may have phase angle errors and dissimilar phase voltages and it will have no effect on a delta primary winding unless there is also a delta secondary winding.
The OP asked about Delta:Wye, not Wyeelta which does have serious circulating current issues.
As a comment regarding four wire Wyeelta transformer banks that were common in the earlier part of the last century when utilities were upgrading 2400V systems to 4160V systems;
It was common practice to reconnect the 2400V transformers from line to line to line to ground/neutral. Most industrial plants were using delta services.
This caused system wide circulating currents to the point that a hard ground fault on the secondary of a large transformer bank would cause widespread fuse blowing throughout the system. Wyeelta, not Delta:Wye.

Bill
--------------------
"Why not the best?"
Jimmy Carter