Saturated CT or not ? 1

[collies99]

Hi all:

Would appreciate some comments.

Tested a CT which does work when performing primary injection at 100 amps with some errors, so it does perform transformation. However I cannot obtain a typical X1-X5 saturation curve, and ratio testing is off of nameplate values. These are C200 rating andI have included 2 EZCT summary reports, one typical (and 29 was tested as per typical) and the odd CT result of CT in question.

Thanks for any feedback.

 

[DTR2011]

Did you megger the CT secondary? If the megger results indicate anything less than infinity(to ground), it may be a pinched secondary wire.

My guess is that if the CT meggers OK, it is something in the test leads, or clips. I'm assuming this is an ANSI CT with a GE or Marathon (or similar) black 6 point shorting terminal block.

The EZCT leads come out to a 4mm, bannana plug connection and something is used to connect to the CT terminal block. Check your clips and see if connection is being made to the shorting / grounding strip in the terminal block. It appears that @~10V, you have ~1 amp. I believe with the EZCT, the 1 amp is going to ground and not the neutral (or neutral is internally grounded and going to ground that way).

We use these beefy 4mm to fork lug adapters, often also used in relay testing. The adapters sometimes disappear and something else like the cheap non insulated alligator clips are used. I think I've had similar results in the past only to find my alligator clip was making contact with the shorting bar.

 

[stevenal]

I saw very similar results when an arrester bracket caused a ground loop through the window of a bushing CT on a gas breaker. Any chance you have a second path through the primary window?

 

[collies99]

Stevenal:

This CT is from GIS gear, isophase and CT is external. Cannot imagine another primary path.

dtr2011:
No grounds on sec windings, all 5 leads are isolated, voltage drop test per tap as per ratios of winding.

would a 20A sat demag this unit? We have tried up to 12.5A to no avail.

 

[scottf]

That looks more like a turn-to-turn or turn-to-core short to me...assuming it's not a connection issue as mentioned above. Doesn't look like magnetization.

If it is a turn-to-turn or turn-to-core short, then you should be able to isolate it.

 

[stevenal]

I have no experience with GIS, but it sounds like you may have grounded metal going through the window of this exterior CT? If so, all you need is a ground path around the outside of the CT to complete the loop. Shorted turns on the secondary would have the same effect, giving you three windings instead of two.

 

[DTR2011]

So the power cables going through this external GIS CT are grounded?
This is a donut type CT?

I had a very strange CT issue in GIS 15 years ago, so maybe my details are a bit foggy. Long story short the CT was at the junction of a G&W Oil/SF6 cable termination. At the junction of the GIS and termination, one of the grounding clips was touching the GIS, creating a ground loop.

Probably something similar in theory to what stevenal is describing. When we performed primary injection on this circuit ~60% of the primary current was going through the GIS and not the CT.

Locating and isolating the grounding clip magically solved the issue!

 

[collies99]

I took a picture of the bus structure showing the external CTs. The housing is aluminum and CTs are encapsulated. Grey epoxy ? Phase-C is most accessible and my issue is phase-C. I will revisit in in 2 days, but I have some breaker timing to do first in another GIS switchgear, they are at 34.5 kv.

scottf:
I like your thoughts and it may line up with the summary test report. Is it possible that a turn-to-turn fault exist in inter-taps X2-X3. This is the only windings that shows a much lower ratio compared to the nameplate, the others are higher and I can surmise that as the effects of the short.. That would explain the higher X1-X5 overall ration. 1-2 turns are shorted out ?

The stumper still is the large excitation current being drawn as one tries to excite the core. It is a C200, but at 100V excitation voltage, the current is already at 10A. I may try with an EZCT 2000C which can do 20A and if it can go above 200V perhaps I can get a showing of a knee point.

Is the order of magnitude increase in excitation current an indicator of turn-to-turn shorts ?

http://files.engineering.com/getfil...64-9665-5949ec9831ad&file=WP_20160219_009.jpg

 

[ScottyUK]

GIS is a sub-group of isolated phase bus and its behaviour is similar in terms of where currents flow. Look for a metallic loop around the CT, perhaps quite a large one, and probably including the IPB outer casing and supporting steel.

 

[stevenal]

Excitation testing is performed with all the windings open circuited. With either a turn to turn short or a ground loop, you have a winding that is not open. You will not see a knee until the situation is resolved.