Trafo inrush current. 2

[slavag]

Hi.
We work now in one substation.
Before two weeks customer energize two new transformers,
60MVA each one.
We check now ATS on the 22kV side of this SS, for test we send trip to HV CB and after energize it again.
I check DR file of trafo protection and fast trend.
Current was crazy, peak about 4xInom for the 0.6sec, but currrent 0.8-1.2Inom was about 15sec. Noise of trafo was crazy too. We check second trafo --same, we disconnect one trafo and check inrush of one trafo only--same. Something crazy. We send this information to customer and stop all works.
Customer ask mnf. Answer of mnf ( in few words):
Before trafo energizing, you must move TC on 33 position ( higer position, it's clear), energized trafo and after 2.5hours move TC to needed position and after this connect load.
We asked, maybe first time, answer : " NO, always".
It's something strange for us, we heard this first time.
From our point of view, something wrong with design of trafo.
What do you think about it.
Thanks in advance.
Best Regards.
Slava

 

[slavag]

additional info. HV lines aren't weak.
5km from power station and 2km from very big SS.
Trafo;s are 170/22kV YnD11 with zig-zag trafo (connected directly to the trafo) on the delta side .
Regards.
Slava

 

[dpc]

Slava,

Is this an on-load tap changer?

Seems odd to me.

 

[slavag]

Hi Dave.
Yes, it's on-load tap changer.
For end customer and us ( we are responsible for protection and SCADA) it seems crazy. First energazing, maybe. But always.. isn't possible!!!!!!!!!
Best Regards.
Slava

 

[dpc]

But why would it make that much difference anyway? Is the incoming voltage too high?

I agree that doing this on first energization might be plausible, but not for every energization. Is there a language barrier?

 

[slavag]

Hi Dave.
"Is there a language barrier?"
Possible too, I didn't see mnf. answers and instruction.
Responsible person of this mnf. isn't seems as good professional ( sales manager).
Incoming voltage are nominal, normal work is 17 position.
I will ask end-customer send me mnf. answers.
It's intresting trafo, w/o cooling funs. ( place is ME)
Thank you.
Slava

 

[dpc]

I guess I would try to talk with the manufacturer directly if possible.

Based on the current you were seeing, I'd suspect some problem with the transformer. But the manufacturer thinks it is normal?

Do you have the transformer test reports?

 

[Bronzeado]

Hi Slava,

This a interesting phenomenon. Is it possile to send to us the waveforms (voltage and current in contrade format) and a single line diagram of the system?

I have a clue, but I would like to analyse first the waveform and system configuration to give an opinion.

Best regards,

Herivelto

 

[slavag]

Dear Dave and Herivelto.
Thanks a lot.

Dave. I haven't test report. I haven't option provide something official ( customer do it). I'm only subcontractor
with very small part of work in this project. ( small but very important ). More, trafo mnf. is our competitor, and...
OK, it's ME, it's all.

Herivelto.
Please find attached data.
one file, it's first DR with 1.4sec of recorder lenght, trafo in 17 position.
second file is 5sec ( for this I used only 3 current channels) of recorder lenght.
trafo in 30 position.
I haven't recorder for the 10-15sec, it's fast trend on the SCADA display w/o recorder and with low resolution.
OLD is attached in zip file, sorry for the resolution ( my mobile scaner isn't best).
170 GIS, 22kV AIS and both trafos placed in the same building, distance between trafos about 20m.
Best Regards.
Have a nice weekend
Slava

 

[Bronzeado]

Hi Slava,

Based on the information you sent I would say that the behaviour of the transformers is OK.

The slow rate of inrush current decay is normal when a transformer is energized nearby other transformer. This is due a phenomenon called "sympathetic interaction" between transformers.

I will upload a paper "Harmonic analysis of magnetizing transient currents associated with transformer energization", which was published in the VII Encontro Regional Latino Americano do Cigré - ERLAC, Puerto Iguazu, Argentina, may 1997, where the phenomenon of sympathetic interaction is explained.

I hope this paper may help you.

Best regards,

Herivelto Bronzeado

 

[Bronzeado]

Slava,

Sorry. I was not able to upload the paper. Its size is almost 5Mb.
How can I send it to you?

Herivelto

 

[slavag]

Dear Herivelto.
Thanks a lot.
I'll send my e-mail to you ( I know your ).
We had cheked inrush w/o second trafo, more or less same inrush.
Mnf. send to customer answer, that work with 33 position need only first energazing.
Customer would like order test trafo services from third side.
I little worry , next step of this SS connection of 50MW gaz turbine, and probably so long inrush ( for example, in case of one - phase AR on the line) will impact on the protection.
Best Regards.
Slava

 

[Bronzeado]

Slava,

Please, could you send me the complete system one line diagram, since the generator until the customer´s transformers, with the exact configuration of the transformer energizing.

The sympathetic interaction occurs due to the DC voltage drop cuased by the inrush current in the series resistances in the circuit. So, transformer in series (such the generator step up transformer) or in parallel to the transformer being energized may saturate due this DC voltage drop, triggering the phenomenon.

Best regards,

Herivelto Bronzeado

 

[dpc]

Slava,

I have seen some very long inrush periods on new transformers when first energized on site. The transformer field guys talk about the core getting "settled in" after been moved around, but I don't have a scientific explanation.

So if this is a one-time problem, it may be something that is acceptable. I'd make sure the no-load current after it settles down is within specs, if you can measure it accurately.

Cheers,

Dave

 

[davidbeach]

Unloaded transformers "ring" for far longer than loaded transformers, the load provides damping. New transformers are generally first energized without load and will therefore have a very long inrush.

 

[slavag]

Hi.
Thanks again for your help.
I agree with all you, first energizing is long inrush.
First, second... I puted attention on this one weak after energizing and after sevral on/off of HV CB.


I now back to home.
Today I connect to end user, some crszy situation.
He check again , bad situation, inrush tens seconf, but not so high level.
He sent me also answers from trafo mnf."
---------------------------------------------------------
1) It's not a normal situation for a primary distribution transformer to be submitted several times a day to inrush situations; please take into account that, for every inrush, we could have (depending on the instant of the HV circuit breaker closing) inrush current values up tu several times the normal service current (calculated value 6.4 x In with the OLTC in the nominal position), with the consequence of repeated severe mechanical stress on the HV winding (fatigue stress) and consequent danger for the transformer integrity.
Of course the transformer is designed to withstand the mechanical stresses due to the overcurrents (e.g. in case of short circuit), but it doesn't mean that the transformer must withstand e.g. 3 short-circuits/day!
Other types of transformers (like the transformers feeding arc furnaces for steel fusion) are subjected, as a normal service conditions, to 10÷30 inrush operation / day, but they are conveniently designed to withstand these heavy service conditions.
Otherwise, I think that the continuous interruptions of electric supply are a problem also for the continuity of the service of this factory; so, the customer has to contact the utility supplying the energy in order to stop these problems as soon as possible.

2) It's not necessary to wait for two hours after every new energizing operation before closing the MV circuit breakers; this step is suggested only for the first energizing operation, in order to verify the behaviour of the transformer.
It's always valid the suggestion of putting the OLTC on the position giving the maximum ratio (in this case pos.33); this precaution it's necessary to reduce the amount of the inrush current (calculated value of 4.9 x In), specially in this case with a lot of switch-off / switch-on operations.
So the procedure of energizing is the following:

- Put the OLTC on the position 33 (maximum ratio);
- Close the HV circuit breaker.
- Put the OLTC on the position corresponding to the actual value of the HV.
- Close the MV circuit breaker and adjust the voltage on the MV side to the correct value by operating on the OLTC.

Unfortunately, the AVR TAPCON can't go automatically on the max. ratio position in case of HV failure.
-------------------------------------------------
OK, I first time see such recommendation.
I understood only one, this type of trafo isn't type tested and isn't tested by SC situation.
From my point of view it's some bad design.
Best Regards,
Slava

 

[slavag]

Dear All.
I would like add to this topic intresting document, for part of you it's clear, for part of you it's something new.

Best Regards.
Slava

 

[slavag]

OK. Small update.
After several.... first energazings.
Inrush about 20sec. According to diff protection 2-d harmonic block od 87T about 14-24sec. With two trafos, with single trafo . Third party company checked this trafo, results come soon, according to first data, constraction of trafo it's a problem.
Best Regards.
Slava