GSU transformer unique PF tests

[rseggerman]

While testing a 200 MVA, 345 / 18 kV, delta wye generator step up transformer, "poor" power factor test results were noted in the CL tests at 10 kV. The tests were repeated at lower test voltages and the power facor results were worse. Does anyone have any theories on the cause? A summary of the test voltage and power factor results are noted below.

kV CL (power factor %)
10. 0.84
5.0 1.33
4.0 1.79
2.5 3.02

 

[Ceast]

Doesn't look right. Not enough detail though. Totally isolated bushings? Everyting de-termed? Testset type? Humidity? Oil sample?

 

[ausphil]

Depending on your test set arrangement, your measurements can be influenced by other electric fields in the yard that you are testing. We have found that in some yards, we cannot get reliable readings below 20kV.

This is not an issue if you are testing at a frequency just off your power system frequency, as the influence is removed because you are not measuring the power system frequency component.

Also, don't forget backward trends (ie power factor decreasing with increases in applied voltage) in power factor can be indicative of larger problems (eg delaminated foils in bushings).

Have you guarded out everything that you should have, or are you measuring all insulation paths between your test terminals?

Ausphil.

 

[rseggerman]

Some more data on this.
1) The transformer was isolated for this testing
2) The test set in use was a Doble M4000 (note that this tests at frequencies just above and below nominal of 60Hz)
3)The transformer was tested in Nov of 2010. During this time slightly elevated power factor test results were observed on the low voltage bushings. The CL power factor was suspected to be more than just a problem in the bushings
4) The low voltage bushings were replaced in March of 2011.
5) The transformer CL power factor tests in March after the bushings were replaced are nearly identical to the November 2010 results. (We fixed one problem; howevever, another is suspected.)
6) Note that two different test sets were used and two different contractor performed the tests in Nov 2010 and March 2011.
7) The test results were temperature and humidity compensated

 

[unclebob]

I always bring a capacitance/DF bridge (DB100) with me when doing a Doble test. Helps to confirm bad results or bad setups.

 

[prc]

unclebob, can you explain a little more.How it helps to confirm bad results.
A high PF of bushing will not affect the winding PF during measurements.

 

[backpack96]

I would suspect a bad connection as the power factor decreases with voltage. How did the current and capacitance compare with the test voltage? There was a paper at the 1989 Doble conference that was written about dry type transformers but it covered topics such as: tip-up, tip-down, and flat characteristic, but it did mention about core ground problems. Has DGA testing been performed? If so what are the key gasses? Has it been overheating?