Synchronizing (2) feeders that are out of phase with each other 5

[bdn2004]

I work at a plant that is considering putting in a 4160V auto transfer scheme on their cooling water pumps. They want to do this with a main-tie-main-tie-main type configuration ( 3 - 4160V feeders). The problem is that one of the feeders is 60 degrees out of phase with the other two.

Is this possible? Are there more questions I need to ask?

 

[slavag]

No Problem Bill.
I agreed.
Regards.
Slava

 

[bdn2004]

Thanks for the break! This has now become a quest...

What else do I need?

 

[slavag]

Hi. I think Dave(DPC) posted right reason:
"Normally for this application at a power plant, one of the transformers is made with the 30 degree phase shift in the opposite direction. This is a common setup. But without looking at a one-line, I'm not sure this is exactly your situation.

If you already have the transformers, you may be able to reconnect them, as David Beach mentioned.

You might have a look at this: FAQ238-1154: Can I parallel a DY1 and DY11 transformer?"
It's also common mistake in power plants.

Other issue, for this size of power plant , Becwith relay HBT is good solution.

Regards.
Slava

 

[stevenal]

So from the 138kV system to the tie on the left, there is a wye-delta feeding a delta-wye; each providing a 30 degree shift. The overall shift from the 138 system is likely to be 60 degrees, but could be zero if non-ANSI windings or unconventional connections were used.

On the right, there is a single delta-wye, causing a 30 degree shift from the 138 system. So from left to right you are 30 degrees out of phase. I believe your choices are to add a transformer, or to do a high speed transfer.

 

[Marmite]

I don't think it's safe to assume any phase shifts. Suppose the delta star transformers are Yd1 but connected Yd9. You would need to look at all primary and secondary connections and all the rating plates to get the transformer vector groups before you can even think about drawing a phasing diagram.

Regards
Marmite

 

[stevenal]

Marmite,

I don't think we need a phasing diagram to see that it won't phase. Every delta-wye or wye-delta connection will shift by an odd multiple of thirty degrees. Combination of two odds on the left makes an even multiple or zero, against the odd multiple on the right.

 

[davidbeach]

If your drawing is accurate, the two ends of your "new feeder" are n*30[°] apart where n=1, 3, 5, 7, 9, or 11. There is not a wiring change that can solve that problem. You need one more transformer or you look at open transition transfer.

For fast transfer you can have n be any of those values by selective transformer rewiring, but not 0. If the new connection would be normally open and used when the normal source is unexpectedly lost, I'd arrange things so that the standby source is 30[°] lagging the normal source. That should allow successful fast transfer while the motor bus falls back 30[°] or a bit more. Return to normal will be a bit more difficult to achieve, but it can be scheduled and certain motors off-line.

On the other hand, if you need to go both directions equally, I'd pick n to be either 5 or 7 so the motor bus falls back 150[°] in one direction and 210[°] in the other direction.

 

[slavag]

Sorry. It's standard power plant configuration and not needed any aditional transformers. Maybe problem if used
only Y/D1 or Y/D11? I check tommorow primary connection scheme, but from my home archive. Standard power plant configuration and it's work : one RAT transformer 161/6.6kV
and four UAT ( two per generator) 20/6.6kV
work with live transfer w/o any problem. All synchronisation logic on the 6.6kV I provided before two years. See attached
Regards.
Slava

 

[slavag]

Not, Not, I must check, maybe isn't last version of drw.
Seems RAT is Y/Y according to vector diagram.

 

[davidbeach]

slavag, look at it this way: Start at the 138kV grid, go down to the motor bus through T1 & T2 plus T3 and T4. The motor bus is an even multiple of 30[°] shifted from the 138kV bus. If ANSI standard transformers are used the motor bus lags the 138kV bus by 60[°] but any other even multiple of 30[°] is possible, including 0. Now start at the 138kV grid and go through T7 to get to the motor bus. Now there is an odd multiple of 30[°] between the two.

There is and always be a phase shift between the two paths unless another transformer is added.

 

[slavag]

Yea , you are right,
now, I would like check my old scheme, maybe I forgot change in the drw RAT to Y/Y or ... after two years forgot something. tommorow at office I check AS MADE drw.

 

[waross]

I agree with Steven and David on this.
Rewiring will achieve phase shifts in multiples of 60deg.
A wye delta cnnection gives pase shifts in multiples of 30deg.
The T1,T2,T3,T4 group has two wye deltas in cascade, resulting in a simplest case phase shift of 0deg or 30deg + 30deg = 60deg.
Your T6 and T7 transformers are a single stage of wye delta transformation with a phase shift of 30deg. Not compatible.

David, How fast is a fast transfer? I am visualizing something in the order of a cycle or at most one or two cycles. In any event fast enough to complete the transfer before a motor can drop more than a few electrical degrees out of sync.
Anecdotally I had a situation with about 30 30HP propellor fans with two fans on each breaker. When the group was shut down by tripping out the breakers, the pair of motors appeared to circulate the back EMF between them. It would be two or three seconds before the back EMF decayed enough for the contactor coils to drop out.
To the point, with the number of 1000 HP motors involved,I would be cautious of transferring so close to 180 deg out of phase.
However, seeing that a transfer will be done by tripping one breaker and then closing another, rather than an ultra fast transfer on a single set of travelling contacts, I would rather monitor the back EMF and inhibit closing until the back EMF on the bus has decayed to a safe level.
But, is there something about fast transfers that I am about to learn? What I know as a fast transfer is so fast that it is done with a sync check relay.
I would rather pull up 30 deg than take a chance closing on residual voltage at 150deg or 210 deg. You may be close to 200% residual voltage.
Respectfully

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

 

[davidbeach]

Bill,

I've watched the Beckwith relay do it's thing at one of our plants, where we are also faced with a 30[°] shift between the two sources and the fast transfer there took about 6-7 cycles (issue trip, confirm open, issue close) and the motor bus fell back about 40[°]when going from the leading source to the lagging source. So the motors were about 10[°] behind the source being connected to. That was far better than the 19 cycles or so when transferring from the lagging bus to the leading bus and we had to drift back 330[°] to come into alignment. Lost a big fan at about 15 cycles. We will be rewiring a transformer to shift it back 60[°] so the standby source moves from leading the normal source by 30[°] to lagging the normal source by 30[°]. In our case we will be swapping A and B on the 115kV side and B and C on the 4.16kV side.

The relay that is controlling all of this is watching bus voltage and both source voltages, knows breaker close time, calculates rate of change of phase angle, and issues a close command far enough ahead of desired close point to hit it as close as possible. In our case the 40[°] shift in motor bus voltage happens in the fastest possible operating time so we can't do better.

 

[waross]

David
Thank you for your reply and the time you took to give me a detailed explanation.
Respectfully

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

 

[slavag]

Folks.
I'm so sorry, my home drw was wrong ( copy-paste). I check now my AS BUILD drw's at office, RAT is Y/Y 12. Sorry again.

Fast transfer.Possible, it's work. But it always stress for motors. If used fast transfer in fault time, is O.K. but for maintanance is not so good.

From my point of view better add one Y/Y transformer from 138kV as dedicated RAT ( or standby) connect to one common bus (3'd bus in drw with new feeder red line connection) and from this common bus with two CB's provide connection to each motor BUS in parallel with T3 and T4.

Or more simple solution: add new CB from T3 bus to 3'd bus as standby supply.
T3 and T4 feeded from double bus generator bus and of course possible reverse feedeng from 138kV grid.
Regards.
Slava

 

[slavag]

Hey. I forgot something important.
Stars to David and Stevenal.
Best Regards.
Slava

 

[stevenal]

Why should a proper fast transfer stess a motor? Aren't you simply starting a motor that already up to speed? It will simply slip until it adjusts to the new phase angle. Seems a lot less stressful than a cross the line start. I'm assuming induction motors. I can see there might be a problem with synchronous motors.

 

[slavag]

Stevenal.
What synch motors used HBT is big problems, once we check this with some German company ( provided such device). Syn motors is in out of step situation and protection must tripped motor.
About stress for the induction motors. In all case you have
phase angle difference and it's reason for some overvoltage on the motor ( not big, but...).
Maybe we need firstly told about terms.
Fast transfer for me is open and close command in the same time w/o any check of CB posotion. It's possible only in case of kvazisynchronasing.
This case is not possible, 30-60deg phase angle isn't kvazisyncronasing.
Our case is 1st phase concidence. David saied about it.
6-7 cycles is good result, in our tested we see also closing command after 300-700ms ( one case was about 1.1sec). you need min 120-150ms for the learn voltage/freq graph. maybe Beckwith relay have another opttion' I don't know. Voltage and freq are decied and in closing time in all cases you have some current pick.
Please see our test reports ( but w/o currents)
Regards.
Slava

 

[slavag]

Stevenal.
Another problems with this type of trnasfer, of course, I say only about manual transfer, not emergency.
1. Pick current will trip infeed.
2. T4 in maintanace and now T3 tripped, all load now on the new feeder, I assume it request some load shedding, after few operations, you would like back supply to T3 or T4, again stress for the system with risk to unwanted trips.
Always with fast transfer you have chance to some fault with stanby or main CB. Is not small power plant, and for my pinion solution must include not only fault condition, maintanance condition too.
At system. was I show, customer used very simple fast transfer. Via main CB contact ( CB open = NC contact) send signal directly to reserve CB close coil ( of course via LOR and additional conditions). Was work about 25years.
Now we provide same logic, but via new digital relys for more safety. Time of transfer about 100-120ms. Think about my feeling, when we tested it. Each down of unit, provided by reverse power trip, system is tested on live.
What I can say, stress for the system, overvoltages/undervoltages, protection starts, tap changer operation etc, once was unwanted trip.
But, is only my opinion.
Regards.
Slava