Transformer energization harmonics 2

[KJLRSA]

I have a customer asking us to directly connect a 11kV 5MVAR capacitor bank with filter reactor tuned to 4.8pu to a 20MVA transformer breaker. The transformer is feeding a submerged arc furnace. The capacitor bank will therefore energize at the same time as the transformer.

They don't want to fork out the extra money for the additional capacitor bank breaker and control. I have a feeling the transformer energization harmonics will cause the tuned filter to amplify the harmonics below the 5th to a point where they damage the capacitors.

I have suggested another breaker so when the transformer is switched on the cap bank only comes in after 1 min to negate the first few cycles harmonics on the Tx, but as stated before money is a problem.

Has anyone had any experience with this sort of problem and can anyone tell me what the expected harmonic magnitudes are on transformer magnetization?

 

[alehman]

No such experience. You could try to run an ATP simulation if you have enough information to accurately model the system. Do you have particular information that suggests there will be a problem?

Alan
“The engineer's first problem in any design situation is to discover what the problem really is.” Unk.

 

[jghrist]

Is the capacitor on the secondary or the primary of the 20 MVA transformer?

Transformer inrush consists of high second harmonics and a high dc component. They probably won't cause a problem with a filter tuned to 4.8 times fundamental.

The inrush is on the primary, so if the capacitor is on the secondary, there will be no problem.

Arc furnaces have very high 2nd and 3rd harmonics, so the filter tuned to 4.8 may not be that effective. The 3rd harmonics will not be blocked by delta windings, because the phases are not balanced.

 

[alehman]

I was assuming primary. The 5th harmonic on inrush may be on the order of 15% where the 2nd could be 50%. Still may be worth looking at if on the primary.

Alan
“The engineer's first problem in any design situation is to discover what the problem really is.” Unk.

 

[KJLRSA]

Hi

The capacitor bank is on the primary of the transformer. Have being trying to model with Etap and LT spice but done have enough info regarding the inrush harmonics to transformers

 

[slavag]

Hi.
Is not seems good, direct connection of capacitor bank on the transformer primary.
From my point of view, is unpossible provide model of harmonics of transformer inrush current. Is depend on the constraction of this transfromer too.

Best Regards.
Slava

 

[MikiBg]

KJLRSA,
If you have so called "single tuned" filter, and looks like you do, it will actually amplify harmonics of order lower than filter's tuned frequency (during inrush or later), because of capacitive nature of filter on that frequencies. Filter needs to be tuned at the lowest harmonic in load spectrum. So, it looks to me that you have at least two problems, transformer inrush with cap connected, and incorrectly tuned filter.

 

[rbulsara]

Just thinking, large motors are routinely started with caps in parallel, this would not be much different than that.

I would at least provide fuses in cap circuit and then it will be safe to try it out.

Rafiq Bulsara

http://www.srengineersct.com

 

[TurbineGen]

I know this might not be the best solution, but perhaps you could connect them to the transformer via an airswitch? Airswitches are relatively inexpensive and reliable and also allow you to disconnect and isolate if necessary.

Depending on how many breaker operations the transformer is expected to go through in a lifetime, you will likely not need to add any attachments. The blade and arcing horn of an airswitch can take a fair amount of abuse so the arcing will only be a problem if it is going to be operated multiple times a year.

Also, you wouldn't need a separate cap or breaker control, since the airswitch takes several seconds to close and open. If directly connected to the local trip and close of the breaker, the airswitch would engage the capacitor bank after the harmonics have smoothed out and would open after the breaker has already cleared the cicuit.

If there is a concern regarding recloser switching, the might not be an option.

------------------------------------------------------------------------
If it is broken, fix it. If it isn't broken, I'll soon fix that.

 

[jghrist]

The attached graph shows the time varying harmonic current when energizing a 100-25 kV 30 MVA power transformer (173A base current). Phase X was the highest current. Peak currents were:

dc - 378A
fundamental - 352A
2nd harm - 166A
3rd harm - 86A
5th harm - 59A
7th harm - 46A

Horizontal scale on the graph is cycles from start of the record.

Hope this helps.

 

[rcw retired EE]

We are currently working on a similar problem where the filter bank is tied to the primary of the 6.9 kV isolation transformer on a 7 MVA, 12-pulse VFD drive.

We don't seem to have any problems when the breaker closes, but there are significant currents and voltages when the breaker opens. The filter's capacitors seem to discharge through the delta windings of the isolation transformer creating voltage surges on the secondary that damaged the drive input surge arrestors and SCR's. We have recorded surges to 1.9 p.u. Volts RMS lasting about 20 cycles. (We haven't been able to capture the peak or instantaneous voltage spikes.)

The solution is to add more surge arrestors and a contactor to disconnect the filter before the breaker opens.

 

[KJLRSA]

Thanks All... rcwilson, I didn't think that there would be a problem with breaker opening but you are correct, the cap will discharge through the windings.

jghrist, thanks your graph is going to help me develop a Etap model to fully test the system before we advise the client. I have a feeling that we are going to play it safe and tell the client it is not advisable to install without any form of contactor or air switch to delay the switching on and off of the bank.

 

[Bronzeado]

Hi folks,

One of the solution to avoid switching transients (harmonics, overvoltages, etc) is to use controlled switching. So, we can "kill" the inrush current just closing the circuit breaker at the right point on voltage wave and, also, open the breaker at the right point.

To do that, it is necessary to study and understand exactly the phenomena involved, and use a device (that is commertially available) to control the instant of breaker closing/openning.

Best regards,

H. Bronzeado

 

[alehman]

Bronzeado,
Theoretically this sounds plausible, but I can imagine many practical difficulties. Do you know of a control system that can accomplish this?

Alan
“The engineer's first problem in any design situation is to discover what the problem really is.” Unk.

 

[Bronzeado]

Alan,

Yes, there are some devices that can accomplish this.

I have used one of them to mitigate the inrush current of a 100 MVA, 230/138 kV and I have been successful. I will attach a paper of my experience on that.

Regarding the issue raised by KJLRSA, if the transformer inrush current is killed (or significantly reduced), part of the problem will be solved. Of course, as you said, there will be some practical difficulties but nothing that engineering could not solve.

Best regards,

Herivelto Bronzeado

 

[EP007]

Bronzeado, doesn't the residual flux change the best choice of closing point on voltage wave? Alehman, ABB and Mitsubishi. KJLRSA, the transformer energizing is rich in harmonics, even into double digit harmonic number. ETAP does not do that. The non-linear flux-current (inst values not rms) curve must be in the model. ATP and other similar programs can be used to obtain worst case. TurbineGen, there is something about primary switches on HV that can void transformer warranties, if I can only find it.

I'd be worried about OV, after the energizing, if transformer is unloaded.

 

[Bronzeado]

KJLRSA and coleagues,

Please, have a look in the paper “Harmonic Analysis of Transient Currents During Sympathetic Interaction”, 1996 IEEE PES Winter Meeting, paper 96 WM 288-1 PWRS, Baltimore, USA, Jan. 1996.
By BRONZEADO H. S., BROGAN P. and YACAMINI R.

Best regards,

Herivelto

 

[Bronzeado]

EP007,

Yes, it is necessary to measure the residual flux in the transformer to see what is the better instant to close the breaker. This is done just integrating the decaying voltage when the transformer is deenergized.
Then, we have to adjust the exact time to send the order to close the breaker, taking into account the arc time and breaker pole time delays.

Best regards,

H. Bronzeado

 

[alehman]

Thanks for the paper. Very interesting reading.

Alan
“The engineer's first problem in any design situation is to discover what the problem really is.” Unk.

 

[waross]

From google "point on wave switching"

http://elcon.se/Breaker_testing/Switchsync.htm

And from ABB, go here and download the attachment.

http://sites.google.com/a/pesthaila...ircuit-breakers/CBCHAPTER7.pdf?attredirects=1

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