Doble Testing a Bushing: C2 reading higher than nameplate

[Bob_Woodward]

Does anyone have an explanation for high C2 readings when Doble testing transformer bushings?

The test was on a three winding transformer. All C2 values were consistent but approximately 40 pf higher than the nameplate value

[SecondGen]

Does anyone have an explanation for high C2 readings when Doble testing transformer bushings?

The test was on a three winding transformer. All C2 values were consistent but approximately 40 pf higher than the nameplate value

Increased capacitance could be short circuit in the layers. Attached is from Doble Test Procedures, section 3-25.

[Bob_Woodward]

Increased capacitance could be short circuit in the layers. Attached is from Doble Test Procedures, section 3-25.

Thanks for your reply, although this can occasionally be the case seeing this result consistently across 8 different bushings suggests to me that there is either an issue with the test procedure or another external factor influencing the results. It would be highly unusual for all bushings to experience a short.

Can you speak to any other possible explanations?

Thanks!

[SecondGen]

Thanks for your reply, although this can occasionally be the case seeing this result consistently across 8 different bushings suggests to me that there is either an issue with the test procedure or another external factor influencing the results. It would be highly unusual for all bushings to experience a short.

Can you speak to any other possible explanations?

Thanks!

If you are experiencing that many failures then I would say investigate the test procedure. Often times its very difficult to get a good connection on the C2 tap if you don't have the proper accessories. Its very easy to short to ground. I'd recommend using a small alligator clip jumper from the C2 spring to the HV hook. A rubber blanket can be used for insulation.

[rofo42]

Does anyone have an explanation for high C2 readings when Doble testing transformer bushings?

The test was on a three winding transformer. All C2 values were consistent but approximately 40 pf higher than the nameplate value

The name plate refers to the factory testing. It is my understanding that this value will always be lower than installed values.

The fact that they are all consistent tells me that you are seeing the difference between installed and not installed values.

Just a thought/guess.

[jblunt92482]

Dirty glass can have an effect on your power factor readings. If the insulators are porcelain or glass, then make sure they are clean before testing.

Without pictures, numbers, or data, it's hard to really go into further details on this. I know that the Doble lid works good for keeping the HV hook off of the transformer if you have it up there.

Also, I don't know if you are using FRANK or not, but I know that I've had issues with it with CCVT testing. Just because it tells you to investigate doesn't necessarily mean anything is wrong.

My 2 cents anyways.

[Realtime]

I think I ran into something similar many years ago. The bushings were ABB Type GOH on the LV side of a GSU. The HV side were probably O+C and tested fine. This was commissioning on a new unit, so I only had the nameplate data to go from. ABB was contacted and I got some kind of explanation that the factory tests are in an oil immersed stand, while my test had the LV winding of the GSU connected. Basically the gist of the letter was don't worry, its fine. I submitted that letter with my final report.

[Bob_Woodward]

If you are experiencing that many failures then I would say investigate the test procedure. Often times its very difficult to get a good connection on the C2 tap if you don't have the proper accessories. Its very easy to short to ground. I'd recommend using a small alligator clip jumper from the C2 spring to the HV hook. A rubber blanket can be used for insulation.

Thanks for your suggestion! I used a rubber blanket and was able to improve my test results

[amlawren]

Does anyone have an explanation for high C2 readings when Doble testing transformer bushings?

The test was on a three winding transformer. All C2 values were consistent but approximately 40 pf higher than the nameplate value

I assume the bushings are rated <69kV. People tend to forget that bushings in this class don't have set C2 capacitance & that your measurement can be different from nameplate when installed in an apparatus, compared to testing in air.

If your capacitance values were consistent across all three, then that will be the baseline for future maintenance tests. If future tests measure 10% off from the original benchmark value, there is a cause for concern. In that case, hopefully the technician behind you didn't add an extra ground connection to the flange, attempting to make the measurements closer to nameplate and throwing the real capacitance measurement off.

[Meggerman]

Hello Bob, you are not alone in experiencing the inconsistent readings for C2 across various bushings. I attended a presentation from ABB at a conference in New Brunswick Canada in 2019 - specifically on C2 testing, construction and value of the C2 test. In the presentation, David M Geibbel - Technical Director ABB Alamo TN presented evidence and statistical data showing that C2 was not a reliable indicator of health of a bushing - in other words, would one "condemn/change" a bushing based on the C2 result. David clearly indicated the answer to be NO. As well, when we asked ABB Bushing Headquarters in Sweden for a statement on C2 testing, they formally wrote back to state that C2 is not considered in failing or passing a bushing.
C2 testing is subject to many influences which are not seen in the C1 testing. Comparison to nameplate PF is contentious leaving only capacitance as a comparison. Trending, although useful in C1 testing is both difficult to keep up (as my experience, most users do not maintain the needed trending history)
In my opinion, C1 testing is your best determinent for health, and one MUST compare to nameplate PF and temperature correction is mandatory if at any extreme temperatures above 25C or below 15C.
There are new methods of testing bushings where the bushings are tested at both 60Hz PF and at lower frequency - down to 1 Hz PF with clear evidence of detecting issues very early in the degradation stages of the bushing - and one can compare the other phase results to determine if one specific bushing is changing.
Hope this is useful to you.