Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations The Obturator on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Wye Primary Transformer Neutral 6

Status
Not open for further replies.

nightfox1925

Electrical
Apr 3, 2006
567
I have a 750KVA dry type transformer to be installed in a system. Primary winding is 480Volts Wye and Secondary winding is 208/120v Wye. I understand that the secondary neutral is required to be bonded to ground in order to prevent floating neutral in case there are 120v 1phase loads .

However my question is;

In case the power source is coming from 480 volts 3phase 3 wire delta, Is it required or necessary to bond the transformer primary neutral to ground?

In case the power source is coming from a 480 volts 3phase 4wire Wye, is it required to connect the neutral of the power source to the neutral of the drry type transformer primary?

GO PLACIDLY, AMIDST THE NOISE AND HASTE-Desiderata
 
Replies continue below

Recommended for you

There seems to be some improper use of the terms NEUTRAL and GROUND in this thread. The codes will require the 120/208 volt secondary to be grounded. Grounding the secondary neutral will have no effect on the floating primary neutral.
A zig-Zag transformer on the secondary might work but it is doing the job the hard way. The 208/120secondary already has a neutral. Please note, NEUTRAL NOT GROUND. The primary wye winding is the winding that needs a NEUTRAL. A primary NEUTRAL could be developed with a Zig-Zag transformer and the system would work.
You could also connect a wye delta transformer to the secondary with the wye points connected. This would balance your secondary voltages within a few percent. The transformer sizing would be based on the maximum unbalance possible rather than the full load current.
The overwhelmingly best solution is to sell or trade the transformer and acquire a delta wye transformer. This is the proper transformer for the job.
The second choice would be to generate a primary neutral with a zig-zag transformer. If this connection is grounded, (code requires a neutral connection to be grounded if the neutral is in use, but you may get an inspector to grant an exception here) the sizing would depend on the capacity of the 480 V delta system and for a large system may be larger than the 480:120/208 V transformer. In the event of a ground fault on the 480 V system the 480 v system would then behave as a grounded wye system. Ground faults would be much more destructive. If the inspector allows an exception to run the primary neutral ungrounded, you are looking at a much smaller zig-zag transformer.

Protection of the zig-zag transformer may be an issue. If the connections to the zig-zag are opened under load, you can develop over voltages on the 120/208 volt system regardless of the location of the zig-zag (primary or secondary) Under conditions of a fault to ground, loss of the zig-zag will result in 173% over voltages on the healthy phases and greatly reduced ground current. The reduced ground current may not trip the primary protection and leave the over voltage condition for an extended period of time.
These constraints will also apply to a secondary connected wye delta transformer.

I realize all the problems that arise when you are stuck with the wrong piece of equipment. Often the political problems are more difficult than the technical aspects.
But, the very first step is to investigate the possibility of exchanging the transformer for a more suitable transformer.
NOW, GROUNDING is a whole other issue and proper grounding will be dependent on the solution you choose for the transformer issue.
By the way, are you sure that the transformer has a wye primary winding? A delta primary winding is more common.

Bill
--------------------
"Why not the best?"
Jimmy Carter
 
Nope, I don't think it would work. With nothing fixing a voltage on the X0 terminal, the differing load currents cause differing voltage drops between X1-X0, X2-X0, and X3-X0. Nothing on the secondary, other than perfectly balanced load, can prevent this. Those differing voltage drops are then reflected to the secondary. With mismatched voltages applied to the zig-zag transformer, you may well have excessive circulating currents flowing in the zig-zag.
 
Nope, I don't think it would work. With nothing fixing a voltage on the X0 terminal, the differing load currents cause differing voltage drops between X1-X0, X2-X0, and X3-X0.
Do you mean nothing fixing the voltage on the H0 terminal? The X0 terminal would be fixed by a neutral grounded at the grounding transformer.


If X1-X0, X2-X0, and X3-X0 are fixed, then H1-H0, H2-H0, and H3-H0 would be fixed also by the transformer turns ratio.

If there is a load imbalance, then there will be different voltage drops in the secondary windings, but how is this different from secondary winding voltage drops in a delta-wye transformer?
 
Egads. Every place I used X, change to H.

davidbeach - revised said:
Nope, I don't think it would work. With nothing fixing a voltage on the H0 terminal, the differing load currents cause differing voltage drops between H1-H0, H2-H0, and H3-H0. Nothing on the secondary, other than perfectly balanced load, can prevent this. Those differing voltage drops are then reflected to the secondary. With mismatched voltages applied to the zig-zag transformer, you may well have excessive circulating currents flowing in the zig-zag.

Think about the condition with load on one phase of the secondary. You have current in one winding of the primary, say the H1-H0 winding, and no current (neglecting magnetizing current) in the other two windings. With no current in H2-H0 and H3-H0, H0 will attempt to become the voltage mid way between H2 and H3. This will result in H1-H0 being at about 150% of nominal voltage. Therefore X1-X0 will also be about 150% of nominal voltage with X2-X0 and X3-X0 both about 87% of nominal and nearly 180[°] apart in phase. The zig-zag won't much care for this situation, and no current can flow into (or out of) the X0 terminal since there is nowhere for the corresponding current to flow on the high side.

Much of that is assuming ideal or near ideal transformers and no zero-sequence path through the tank. Your mileage may vary, but the affects should be similar.
 
Call the zig-zag connections A, B, C and N to not confuse them with the step-down transformer.

The transformers are connected X1-A, X2-B, X3-C, X0-N.

When the voltage on A-N rises does the zig-zag not then try to also raise the voltages on the other 2 phases of the step-down transformer? Does it not do this by causing current flow X2-X0 and X3-X0 which will balance the step-down transformer?

Your example says that the voltages A, B and C to N on the zig-zag will go unbalanced but does the zig-zag not keep these voltages balanced by design? If the A, B and C to N voltages on the zig-zag are balanced then the X1, X2 and X3 to X0 secondary voltages of the step down transformer must be balanced. If the secondary voltages of the step-down transformer are balanced then the primary voltages must be balanced.

 
You can't have current flow in X2-X0 and X3-X0 without corresponding current flow in H0-H2 and H0-H3. With no connection to H0, the only possibility is that this current comes from H1-H0 and that would have a tendency to strengthen the voltage imbalance. Remember, you can never have current flow on one side of a transformer if you don't/can't have the corresponding current on the other side, and without current flow there is no voltage drop.
 
Hello LionelHutz;
Remember that the phase angles may be badly distorted and the voltages still sum to zero. I have not seen an assessment of the action of a zig-zag transformer with badly shifted phase angles. Although the zig-zag may work I don't think that we may make that assumtion.

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

There will be no zero-sequence current in either side of the transformer. The current going in one phase will come out of the other phases. See the attached sketch. Assume all windings are 1:1 for simplicity. I used a GrdY-Delta grounding transformer instead of a zig-zag because it is easier to understand.

Positive-sequence transformer current is I/3 @ -120°. Negative-sequence current is be I/3 @ 120°. Zero-sequence current in the grounding transformer is I/3.
 
 http://files.engineering.com/getfile.aspx?folder=dd96368d-83d7-4714-81bd-af19574348fc&file=012308_Drawing.pdf
Well. I don't have any problem with the current distribution, but the voltages bug me. I've set it up in ATP and find, much to my surprise, that H0 does wind up where it should be. Lots of circulating current and all of it in phase, pretty ugly, but it does work. For now that has to be a temporary answer though because I can't get ATP to produce a meaningful solution without the grounding transformer, I get 1/333 as much current through my load without the grounding transformer whether or not I ground the secondary neutral.

Certainly an interesting puzzle, but it is now much later than I should still be working on this.

 
Thanks for the confirmation, Jjghrist. From my experience in an area where two common issues are wye delta transformers with the primary wye connected to the system neutral, and burned up transformers I was sure that a wye delta connection would work. However I was developing a degree of uncertainty as to whether it would work on the secondary. You have now removed that uncertainty.
David, think about the action of a wye delta bank with the wye floating. The wye point is normally at the proper point, it only migrates when the primary phase voltages are unbalanced. If the primary wye is connected to the system neutral, the transformer bank tries desperately to pull the primary neutral back to the true neutral point. It is transferring power from the higher voltage phases to the lower voltage phases. That is the reason that the circulating current is real current rather than quadrature current.
I hope this explanation makes sense.
Respectfully

Bill
--------------------
"Why not the best?"
Jimmy Carter
 
That was the kind of thing I was expecting. Ugly but it would work.

Assume an ideal transformer.

If you connect the transformer by itself then a single phase load as the only connected load will have 0VAC. If this one transformer load is connected X1-X0 then the X1-X0 voltage will be pulled to 0VAC and the H1-H0 voltage will also be pulled to 0V. However, the X1-X2, X2-X3 and X1-X3 voltages will still measure 208VAC with the proper 120 degrees phase separation. The transformer becomes the equivalent of an open delta connection.

By design a zig-zag will keep its neutral terminal centered between the 3 phase voltages. As discussed above, the Y-Y connected transformer with the open Y connections will still be creating balanced 3-phase line-line voltages regardless of loading. So, the zig-zag is getting the proper L-L voltages and it will always be creating a centered neutral point. Connect that zig-zag transformer neutral to the supply transformer neutral and the zig-zag then forces whatever 3-phase currents are necessary to keep X0 centered. If X0 is centered then the floating H0 must also be centered.

Of course, all this is assuming an ideal transformer. In the real world, the transformer would saturate with sqrt(3) times more voltage applied to one of it's coils above the design point. But then, the zig-zag would never allow that to happen so it's really a moot point.

 
By the way, despite the fact that developing a neutral on the secondary side will work, I suggest that by far the preferred solution is to replace the transformer with a delta primary transformer for a delta supply system, or to connect the transformer wye point to the system neutral in the case of a wye supply system. A possible code issue will be grounding the neutral of the primary supply system. A 480 volt wye system may be impedance grounded if there are no neutral connected loads. When you connect the transformer wye point to the system neutral you must also solidly ground the system neutral. Developing the neutral on the secondary side avoids primary grounding issues.

Bill
--------------------
"Why not the best?"
Jimmy Carter
 
For now that has to be a temporary answer though because I can't get ATP to produce a meaningful solution without the grounding transformer, I get 1/333 as much current through my load without the grounding transformer whether or not I ground the secondary neutral.
Without the grounding transformer, there is no ground source on the secondary. The transformer will not support a phase-to-neutral connected load without a ground source.
 
I think I need to back up one step and start over.

I tried to over simplify the situation I've run into in the past and tried to make it as clean and simple as possible. The known problem was a single phase CPT connected phase-phase on the secondary of a wye-wye transformer with only the secondary grounded. This caused significant voltage imbalances.

So, now lets move the load in jghrist's diagram, from the middle phase to the bottom phase. In this case, I believe you will see the voltage imbalances, but I'm going to wait until I have time to get back into ATP before going further.

Obviously, if you add a grounding transformer and have only load from one phase to neutral, you will have voltage stability. I'll see where it goes with other imbalances and report back.
 
So, now lets move the load in jghrist's diagram, from the middle phase to the bottom phase.
My load is from the bottom phase (chosen for topological convenience) to the neutral.
 
Sorry, not clear on my part. Connect the load so that it connects from the middle phase to the lower bottom phase. Obviously connecting it from the middle phase to neutral wouldn't change anything.
 
Yep, and my experience is with that situation you won't have H0 at the "neutral" point and therefore the voltage across X1-X0, X2-X0, and X3-X0 won't all be the same magnitude.
 
Yeah, I see what you mean now. Without the grounding transformer to establish a neutral point, or having H0 tied to the neutral of a grounded system, the X0 and H0 voltages will move to halfway between X2-X3 and H2-H3 respectively for an X2-X3 connected load. The voltage to neutral of X1 and H1 will become sqrt(3)/2 times VLL.
 
Well this gets interesting.
You have shown a 480/277 volt primary and a 120/240 volt secondary with a phase to phase load on the secondary.
By the fact that it is one load, it will be balanced on the two phase windings.
Now, the 277 windings act in series across one primary phase and so each receives only 480/2 = 240 volts.
With 240 volts on the 277 volt winding the secondary windings generate only 104 volts instead of 120 volts. However, the secondary windings are in series and in phase, so they sum to 208 volts. The third phase winding will see 416 volts on a 277 volt winding and try to develop 208 volts in a 120 volt winding. In the real world, the third phase windings will probably be up on the knee of the saturation curve. The disproportionate increase in the current will be pulling the neutral point away from the straight line and so reducing the voltages and currents somewhat but we may still expect over-voltages on the lightly loaded phase(s).
Now if we add a line to line load on a second phase it gets even more interesting. I expect we will have vectors in from the three corners of the delta supply with the ratios of 2, 2, and 1. We will now have equal, but above 120 volt voltages on two secondary windings and a third winding developing 50% of the other two voltages. This neglects the nonlinearity of the magnetic circuit.
Am I visualizing this correctly fellows?
Respectfully

Bill
--------------------
"Why not the best?"
Jimmy Carter
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor