Theory and Apllication of electrical tests on transformers

[lal786]

Hi all,

TTR, mag, impedance, DC winding resistance, SFRA, FDS, RVM, PI, LVI. Have I missed any?

What do they measure and what can I glean from the results? I ask the question because yesterday I had a buccholz trip, the oil showed high levels of acetylene. Why must I waste time and money on electrical tests, when the transformer must be detanked and inspected.

Another incident, I had a delta-star-delta transformer supplying a 12 pulse rectifier bridge and VSD. The VSD popped (a multi-million dollar pop!). Transformer saw the fault for 2 seconds. So I tested the transformer, all the tests in the book were done. All tests passed, except the sfra. The sfra was sent to the doble specialist, who wanted to know if the transformer was in the tank when we tested it (funny guy!). Point is, the sfra was inconclusive and we are now detanking and inspecting the transformer (not sure what we will find). Guys, please don't focus on the SFRA test, i was just using it as an example. The two questions are: what are the electrical tests and what do they tell us? and why must we test if we know that there is something wrong!

 

[Eleceng01]

I have always found Myers', et. all, book "A Guide to Transformer Maintenance" (ISBN 0-939320-00-2)to be useful. There are other IEEE standards and such as well. Hope it helps.

 

[electricpete]

what are the electrical tests and what do they tell us?

It's a little bit broad. I'll get you started with a website.

http://www.google.com

and why must we test if we know that there is something wrong!

It's your choice. When I worked as transformer engineer at general office for a large utility, we would absolutely test a failed transformer. A few reasons I think it's a good thing, in order of importance:
1 - You might want to collect all the data you can while the evidence is fresh to understand the best you can what went wrong. I'm assuming you consider it a good thing to try to figure out why the transformer failed so you can prevent recurrence.
2 - You can build a case study which may provide helpful in evaluating future test results against teardown findings.
3 You might want to get a preview of the scope (and therefore cost/schedule) that liest ahead. Single phase rewind? Three phase rewind?
4 you might want to do a certain degree of independent analysis to validate the information that your repair shop tells you.
5 Chances are that you can send the transformer to be repaired and no need to test. But once in awhile, there might be an occasion when you don't need to send the transformer off. It would be embarassing to find that out from the repair shop and answer to your bosses why you didn't do the simple checks before you shipped it.

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

 

[prc]

lal786, excellent query.Today many advocate SFRA to find out any movements or disturbances of transformer internal body during tranport from factory to site, when this can be easily done by an expert by opening the manhole and looking inside.Many sophistictaed diagnostic tests are superfluous,adding little value.

But in the case mentioned by you I will advocate some simple electrical tests,though Iam sure the transformer failed. These are not expensive and time consuming.These are done to identify the cause of failure and also for locating the failue point which may not be possible unless you tear down the unit.This will help in repair decision making even before the disassembly of unit.From the list of tests mentioned-LV excitation current ( to find out whether the fault is interturn) IR ( whether any break down to earth)Resistance ( location of fault) oil BDV,ppm,FDS( to check any excess water went in- ie cause of failure)RVM (no longer considered meaning ful)SFRA /LVI only in case failure due to heavy shortcircuit current.

 

[davidbeach]

The electrical tests may also tell you to look for something else besides/in addition to what you think you are going to find when you detank the core. So you detank and find something that show at the surface and deal with that and think all is well and then find out that you barely touched the surface of the problem.

 

[Zogzog]

The tests that should be performed depend on the type and size of the transformer. NETA has detailed test procedures for each type.

Here is some reading to get you started.

TTR -

http://www.netaworld.org/files/ItemFileA1200.pdf

Basic overview of several tests

http://www.netaworld.org/files/ItemFileA1201.pdf

And one you forgot about, PD testing (See attachment)

 

[JTS44]

Field test results can be extremely valuable for diagnosing transformer condition, and are relatively inexpensive, comparibly. As mentioned, saving the costs of shipping a unit unnecessarily.
However, generally, acetylene=bad. So you may be in need of some major repairs.

 

[TurbineGen]

Lal786,

Most would test if they plan on repairing the unit and placing it back in service. Testing the transformer gives you a base for comparison. Right now you have acetylene which means you likely have arcing somewhere. Testing the transformer before it is untanked will possibly result in a failed test or two. When the transformer is repaired you will want to test the unit again and compare before placing it back in service.

The tests can also reveal that perhaps the problem doesn't require detanking. For example, a TTR test and resistance test might reveal a broken tap changer (if the unit is a tap changer). You wouldn't want to de tank the unit and plan to rewind it if the windings are ok. Also, it could be a loose bushing connection that a simple resistance test might pick up. The cost of testing a transformer is significantly less than the cost of untanking it in almost all cases.

------------------------------------------------------------------------
If it is broken, fix it. If it isn't broken, I'll soon fix that.