residual problem from filling a transformer with oil without vacuum 3

[electricpete]

I was asked for urgent input about something but only given some vague details, So of course, I'll turn around and do the same to you guys!

This is 25kv / 13.8kv 3 winding oil winding transformer around 50MVA.
Normal oil preservation is nitrogen blanket.

At beginning of maintenance period, oil results and doble results are sat. [not sure if they took megger]

During outage electrical testing bushings were found bad. Transformer drained, bushings replaced. Gas blanket or dry air was kept on transformer at all times.

There were concerns about whether certain gaskets could withstand vacuum, so during refill (the first time) no vacuum was drawn.
[ul]
[li]Above some voltage threshhold, that is bad practice due to concern for getting bubbles in the winding. I know it is bad practice for 345KV transformers. Is it bad practice at 25kv transformers?[/li]
[/ul]

After refill, proper insulation resistance readings were not obtained. So oil was drained, vacuum drawn (successfully), circulated oil under vacuum for some period (not sure how long), and refilled under vacuum. Insulation results still bad. I think the temperature-correcteed insulation resistance are all in the neighborhood 80 megaohms to 120 megaohms (I will try to get all results for all windings to ground and to each other).

My underlying question: is it possible there are residual effects from that first refill (if indeed it's improper to refill 25kv transformer without vacuum)? ... Or should any defects/bubbles introduced during that process have been erased by the draining, circulating, and refillig under vacuum?

 

[electricpete]

We have some more details. During processing the inlet oil was 70C and at the end of processing outlet oil got up to 63C. So the core / winding is still hot and we are applying a big temperature correction factor. The plan is to run the pumps and fans to help equalize temperature and cool it down a lot closer to the reference temperature.

 

[wcaseyharman]

What is your acceptance threshold for insulation resistance? My recollection (I could be off here, please check) for minimum IR of general electrical equipment from NETA is 1 megohm per KV +1, which at 120 megohms is more than sufficient. How did the Doble results look?

 

[electricpete]

There are a lot more details coming in.

But for the time being, I'd rather focus on the questions asked in initial post (not what's going on with this transformer, but whether fill without vacuum is acceptable at this voltage level and whether it might have residual effects after having subsequently drained, drew a vacuum, circulated hot oil under vacuum and refilled under vacuum).

 

[prc]

1) No residual effect from oil filling without vacuuming. There are many manufacturers doing oil filling without vacuuming for 33 kV. But then the important point is you have to leave a pretty long settling time ( I will suggest a minimum of a week or oil filtration through a machine with a vacuum pump) before energization so that all air will get dissolved into the degassed oil used for filling up.

2) Don't bother with IR values much as it more depends on the quality of the oil ( ie the resistivity of oil)

3) Once I burned my fingers during testing some 11 kV class CT_PT metering units, that were oil filled without vacuuming. Then I insisted on vacuum oil filling the units.

 

[electricpete]

Thanks PRC. That answers my question well.

Just an update for info - We ran pump/fans and let the unit cool to ambient and now the corrected insulation resistances are all above 250meg-ohm. They are all still far below historical (trend) temperature-corrected values which are all in the range 900 megaohms to 2500 megaohms for the low voltage windings. BUT a complete power factor test gave good results consistent with historical results. To my thinking the power factor test has more diagnostic value and less inherent variability so it seems reasonable to accept the lower (compared to historical) insulation resistance results even though I can't fully explain why they're different.

 

[waross]

Just a thought, Pete. Is any of the old oil still available?
How about some comparison tests between the old oil and the new oil, or did the same oil go back in?

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[electricpete]

Hi Bill. The oil was reprocessed but you're absolutely right that the oil is still an important piece of the insulation puzzle. I forgot to mention that we drew a sample yesterday and expect results today and will check dielectric strength and water.

 

[waross]

Something to think about;
If there is any air trapped internally it may take quite awhile to absorb into the oil.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[prc]

electricpete, you are absolutely right. Power factor is a better parameter for winding condition monitoring than IR values. IR value mainly depends on the resistivity of oil, not the dryness of paper. Forty years back, in India, I used to struggle to get 50-100 Megohms for new EHV transformers. Those days we were using locally made acid-cleaned oil with low resistivity. Today the same transformers give 15-20 Gigohms as the oil used is highly purified ( made by Isocracking the crude, removing all polar compounds) oil with high resistivity. I am attaching an old paper that I prepared on this subject.

 

[electricpete]

Thanks PRC, interesting paper (from a well respected author!). That makes good sense that the oil conductivity could have a big effect on the insulation resistance test (even if oil conductivity is not necessarily a big factor for insulation reliability).

We don't test oil conductivity. Our oil lab tests dielectric (ASTM 877 and 1816), water, and "liquid power factor". If the oil resistivity was playing a role in decreased insulation resistance measurements, would you also expect to see an increased liquid power factor?