Doble Testing of Dry-type Insulation 1

[scottf]

Does anyone have any information on Doble testing of dry-type insulation systems, particularly cast epoxy. I am most concerned with the test's validity and if deemed valid, the acceptable range of readings.

 

[mpparent]

Scott,

I recently went through a discussion on this. If my memory serves me, acceptable ranges are somewhere between 2 and 8% (I'd have to look it up to be sure). However, I've been told that in any instance the reading could be almost anything. In this case, it's more a trending tool than anything else, and shouldn't be the sole test in a commissioning/maintenance procedure. Be sure to also look at things like excitation, ttr, megger and the like.

Mike

 

[vjankov]

Dielectric dissipation factor is a global characteristic of the insulation system and have no meaning for diagnostic of epoxy insulation. The greatest risk for these structures are voids and they can be detected by partial discharge test. Although it is difficult and expensive to do it on-site and the only appropriate test is megger.

 

[scottf]

I share the opinions stated above. What I'm after is any papers or justifications of these opinions.

Thanks for the help.

 

[jbartos]

Suggestion: Visit

http://library.abb.com/GLOBAL/SCOT/...400016088/$File/1LUB000002-SPE_DryCC_ANSI.doc

for:
5.1.10 Induced Partial Discharge Test

http://www.esi.buffalo.edu/pms_berg_062200.pdf

for: low partial discharge behavior

http://www.ulb.ac.be/polytech/laborulb/pdmgen/pdmgenerator1.htm

http://www.furukawa.co.jp/review/fr019/fr19_21.pdf

for:
7.1 Test Conditions
Partial discharge measurement tests using high-frequency tuning were conducted on two occasions. The first, during the dielectric strength test, involved application of 1.1 E for 10 min to each phase in turn; the second, during the preliminary charging test, involved a 3-phase (batched) test with preliminary charging voltage for 3.5 hrs.
etc. for more info

 

[scottf]

jbartos-

Thanks for the links, but unfortunately they don't address power factor testing, but rather PD testing.

 

[jbartos]

Suggestion: Visit

http://www.usbr.gov/power/data/fist/fist3_2/vol3-2.pdf

for: Porcelain
Also, why not to contact

http://www.doble.com/index.html

directly?

 

[alehman]

NETA recommends these minimum acceptance tests for dry-type transformers: insulation resistance, PI, ttr, power factor, connection resistances.

They recommend power factor less than 1% for power transformers and less than 3% for distribution transformers. They also say results may vary depending on the type of insulation and suggest consulting the manufacturer for additional information.

 

[DanDel]

Doble tests for dry-types are valid, though not as much as liquid-insulated units.
For dry-type transformers, Doble recommends a 'tip-up' test, which is a comparison of separate PF tests at different voltages.

 

[electricpete]

I did some research on this a few months back, including a post on this forum. Doble did publish two articles on the subject. I'll try to dig up my notes tomorrow.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[scottf]

Thanks.

I'm on the manufacturing side and I don't find Doble to be of much help, unless you subscribe to their services. They commonly try to over-ride our recommendations on acceptable insulation PF levels to our customers.

I appreciate any help offered.

 

[ausphil]

We (testing staff) have had this argument with our network owners for years now, that you must be very careful in interpretation of ddf test results.

It all started (cue wavey lines down screen to indicate step back in time) when we could not get the same results from one day to the next on the same piece of equipment (11kV metalclad switchgear)

After analysis, we performed a large number of controlled tests on epoxy bushings in different temperature and humidity environments.

Because of the small size of capacitance of the epoxy bushings compared to a paper bushing (about an order of magnitude lower) the external effects play a greater part in the measurement circuit. So to get the equivalent ddf reading, your external IR (along the surface) must be much better than that for a paper bushing.

We found that temperature was not a significant factor, however rapid change of temperature could be (ie a step change could possibly form condensation on the surface, but stabilised out over time, the reading was within the expected range for the new temperature.)

The main factor in the differences was humidity, this could change the ddf up to 30X for a change in humidity from 50% to 90%.

Conclusion - if you have a stable environment in your substation (ie airconditioned, positive pressure to keep out dust/dirt etc) DDF could be used as a trending tool, irrespective of the absolute value. If not a stable environment, good luck.

Definitely PD, and even just an inspection of the epoxy can give a good indication as to its quality and the manufacturing quality. We had some gear that had holes from the outer surface through to the conductor. Needless to say that got replaced by the very red faced manufacturer.