Parallelling transformers during switching

[TWD82]

I have a substation transformer question and would like to get other people’s opinion or current practice on the subject.

We have several substations with 115kV to 12.47kV transformers that feed a set of switchgear. Some of these stations have 2 transformers, which is where my question comes from. In stations with 2 transformers, there are also 2 sets of switchgear, with a 12.47kV tie breaker between the sets of switchgear. This breaker is normally open so that the transformers and switchgear are completely separate, except that they share the same 115kV bus.

During switching operations to take out one transformer the procedure goes like this: Both transformer LTC’s are put in manual. Using a single phase PT on the low side of each transformer the LTC’s are moved up or down to get their voltages to match as close as possible. Once this happens the low side tie breaker is closed, which allows the low side main breaker for the transformer to be opened and the transformer de-energized.

The question was brought up awhile back if it would be better match percent regulation (steps) on the LTC’s instead of matching low side voltage?

The transformers in question are either 20MVA or 33MVA and may or may not be the same impedance (usually very close), but are the same MVA rating when in the same substation.

My thought on this is that before the tie breaker is closed the transformers could have very different loads on them, causing one to be at (for example) boost 10 and the other at boost 2 in order to get their voltages to match (using our current practice). But when they are tied together this is going to cause circulating current because all of the sudden their load is pretty much equal, their high side voltage is equal, but their LTC steps are different.

If using the method of matching percent regulation the voltages on the low side of each transformer could be very different when closing the tie breaker. Is this initial voltage difference an issue? This would minimize circulating current and I think this would be the preferred method of tying the together but I’m curious how do other utilities address this situation?

 

[opmgr1]

The utility where I work do not match regulation (tap position) for the exact reason that you mention, there will be a voltage difference enough to result in a short circuit. Instead what we do is to match voltage which is a form of impedance matching. The fact that the turns ration are now similar when the secondary breaker is closed the voltage differece at the tie point is so small that the circulating current as a result of the voltage difference can be ignored . Parallel any other way and you may be in some serious trouble!

 

[stevenal]

We match tap positions for the reason you state.

 

[Robert789]

If no load is dropped during the switching operation, I think that matching voltage prior to switching will not result in high circulating current (nor will the transformers share load equally after they are paralleled). With the voltage matching method, I think that the each transformer will continue to serve approximately the same percentage of the total load that it was serving before the tie. The unit with the higher no-load voltage will need more load through it's impedance to bring the voltage down to the other unit's level.

That being said, if you use the voltage matching method and the load goes away for some reason while the units are in parallel, significant circulating current may develop.

When paralleling substation transformers (closing a tie switch), I usually match the LTC/regulator tap postion. The only times I have deviated from this is when placing a mobile substation into service that had a slightly different turns ratio (I used regulators to balance out the turns ratio at the tie point).

To separate a parallel arrangement, I have at times placed bus regulators on different steps prior to opening non-loadbreak tie switches. In this case, I adjusted the regulator positions to the point that each transformer was carrying approximately the same proportion of total load as it would carry when separated. My reasoning here is that doing so would minimize the amount of arcing when the tie switch was opened.




A separate, but possibly relevant story...

Several years ago, a contract linecrew replacing a pole mounted regulator called me to say that the new regulator wouldn't work. I arrived at the site and went up in a bucket truck to examine the regulator. I saw the regulator was in the neutral position and bypass blade open. I then turned on the control and tried to manually operate the regulator. It stepped one time and then the line burnt down. The crewleader said "that's what it did to us too.". There was a mechanical jumper running around the opposite side of the pole.

The circuit formed by the regulator, its jumpers, and the mechanical jumper didn't have enough impedance to limit the circulating current. The phase conductor burnt in two between the clamp of the mechanical jumper and the deadend shoe. The mechanical jumper kept the conductor from falling to the ground, the line remained energized and no protective device ever operated (this circulating current isn't seen by any upline device).

Robert

 

[waross]

The magnitude of the circulating current between transformers of different voltages is related to the % impedance of the transformers.
As a transformer is loaded, the phase angle changes because of the quadrature or reactive component of the voltage drop in the transformer windings.
When two transformers are with different loads and different tap positions are connected in parallel, even though the voltages are set the same, the phase angles may be different. This is because of the reactance of the transformer winding, not differing load power factors.
This gives rise to circulating currents which are limited by the transformer impedances and may be relatively large.
If a heavily loaded transformer is connected in parallel with a lightly loaded transformer when both are set for the same voltage, the added circulating current may overload the heavily loaded transformer.
With the taps on the same step, you may be closing on different terminal voltages.
With the taps at the same voltage setting, you may be closing on a phase angle difference and initiating a circulating current.
My knee jerk reaction is to say;
"Circulating current is bad. Avoid circulating current!"
But, once I got over the initial reaction, I would probably try for a compromise that minimized distribution circuit voltage disturbances as much as possible.


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

 

[slavag]

Hi.
Thanks to all for the opinions. Bill special thanks for you.
Up today we was provide OLTC matching only, max 1 tap position difference for avoid reactive current circulation.
But never connected two trafo with very serious difference in the tap positions, 3-4 tap positions not more.
What will a best solution for the 7-9 tap positions difference.
We think a long time about some combination of tap steps and voltage matching. Mavbe someone now about this way.
I heard from one designer about some average tap positions and he requested from us provide this automaticly...we was against, will provide only manually. But in all cases we think about some algorithm.
Best Regards.
Slava

 

[waross]

A suggestion.
As soon as possible after the tie breaker is closed, the taps should be set to the same position.
However, the transformers should be the same or close to the same % impedance. If the transformers are widely different % impedances, it may not be possible to prevent the transformer with the lower % impedance from overloading with any combination of tap positions.
After the transformers have been first paralleled and one transformer then taken out of service, the tap position on the remaining transformer may be expected to go higher when placed back on automatic control due to the increased load on the remaining transformer.
With this in mind, I suggest that any tap adjustments prior to closing the tie breaker be done on the OLTC with the lower setting.
I wouldn't want to set tap positions lower when an anticipated switching operation will result in a higher setting being required.
Of course, if the OLTCs are being used for head end compensation of voltage drop on long distribution feeders, you will have a different set of conditions and compromises.

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

 

[slavag]

Bill, after bus-tie closing system will moved in Master/Follower logic. Isn't long feeders, factory 600m long cables.
Slava

 

[TWD82]

I appreciate all the responses. It seems there is no cut and dry answer to this one. From this discussion I gather that if we are matching voltage we also need to make sure that the transformers are not more than a couple steps off, and that if matching steps we need to make sure the voltages are not too different. This should be possible in most situations, but what about when putting the transformers back in service? In this situation one transformer could be nearly at full load and the other transformer has no load on it. Matching taps creates a large voltage difference, and matching voltage means that the LTC taps will be quite different.

It is easy for me to quantify the circulating current, and it would be possible to make a decision on the acceptable difference in LTC steps based on transformer loading or relay pickup or something like that. But I am having trouble quantifying the effects of a difference in voltage across the tie breaker, when using the matching steps method.

The tap position difference is fairly straight forward. I can say that in a substation with 20/27/33MVA 10%Z transformers we get roughly 30 amps circulating current for every step the LTC’s are different. But, for every 50 volts difference in transformer voltage across the tie breaker, what happens? How large can the voltage difference be before it is no longer acceptable to close the tie breaker? This confusion is what’s keeping me from jumping on board with the matching taps method. Closing a breaker across two different voltages seems like it would cause a large current flow, is this just a transient until the transformers begin to share load?

I like the point made by waross about making the tap adjustments to the LTC with the lower setting. This is not something we have made practice of, but its is a good idea, because yes the single transformer raises its tap once its put back in auto, due to the new larger load. It only makes sense to not lower the tap on a transformer in manual when you know it will be raising its tap when put back to auto.

 

[isquaredr]

The practice in our company is to match volts, the only thing to add would be to tap the transformer to be switched out to unity pf, this will prevent voltage dip when all the load is transferred to the 1-transformer.
We have many sites with 2-transformers supplied from different sources, the same tap method just would not work as they naturally run 2,3 sometimes 4 taps apart (tied system).

 

[waross]

What is the voltage change per step?
Divide your voltage diference by the voltage change per step and multiply by 30 Amps per step.

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

 

[apowerengr]

Matched transformers (matched base MVA, matched no-load tap position and roughly the same impedance) means match the LTC tap positions for parallelling. The LTC's were running in auto before you decided to switch - assuming you have LTC control settings which don't vary the regulated bus voltage too much based on load variations means you'll have roughly the same bus voltage loaded or unloaded. You have to decide which step the match should occur at - usually somewhere between the present tap positions of the two transfomrers when operating in auto before switching begins. Do your switching, remove the parallel condition and return the LTC controls to auto - they'll take care of the rest. Matching step position means you'll get load and voltages matched when the parallel occurs - this takes care of your circulating current concern.

 

[opmgr1]

If you are matching tap position and not voltage, what happened when there are two transformers with different voltage per tap position? Theory indicated that greater level of circulating current will result from matching tap position than from matching turns ratio.
I have heard of instances where persons were paralleling two substations and match tap position rather than turns ratio and it resulted in a catastrophic failure of the breaker at the tie point and the clearing of the substations from the grid.

 

[stevenal]

Correction to Jan 27 post: We match ratios. For most, this is equivalent to matching taps. The one mobile transformer that has unusual ratios has a laminated spreadsheet on the door showing which steps most closely match all the others in ratio.

 

[dpc]

This has been an interesting discussion. Appears there is no perfect solution to this question. My knee-jerk reaction was that voltage should be matched. Now I'm not really persuaded one way or the other.

 

[slavag]

Hi.
Actually, we have all neede for calculation information in the SCADA, maybe a best solution is insert formula for on-line calculation of circulation current and according to result, decide, when send command to closing bus-tie.
Icirc=dU/((uk1/I1)+(uk2/I2)), Uk,I and dU online parameters.
Uk will calculated according to tap position.
Best Regards.
Slava

 

[waross]

That is a valid and important observation, opmgr1.
Thanks

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

 

[apowerengr]

op, even with different tap % values the amount of circulating current resulting from a couple % difference in voltage between two paralleled transformers is not going to fail them. I've seen people design systems using one LTC power transformer in parallel with a non-LTC power transformer to regulate the parallelled bus voltage. Certainly not a good or recommended idea but it results in much more circulating current than that which we're talking about. LTC parallel balance controls for automatic operation of parallelled LTC power transfomrers plan to run one-step off and don't normally trip for out-of-step operation with greater amounts of circulating current. Beckwith controls has a good discussion of parallel balance and discusses this topic further.