Trying to find a company to evaluate our generator. 2

[justatech]

Hello all

I've got white marks on the end turns on one of our generators. I think it may be corona. It is a Brush BDAX 7-290ERJT, 71176 KVA. We have been going through Brush for the last week, had a rep from them out for the last couple days and no answers. I think the white marks are an issue and want a second opinion.

So does anyone know of a company that that can take a look at our generator in California? And hopefully do repairs as neccesary.

Thanks all.

 

[jraef]

What part of California?

"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[justatech]

Bay Area, San Jose area.

 

[rmw]

http://www.natlmech.com/about.html

Look over their website. Ask for Rudy. They go far and wide.

rmw

PS: my $0.02: you got corona, sorry.

 

[justatech]

Just a few pictures of 'nothing'.

 

[justatech]

The one I thought I got the first time.

 

[rmw]

Corona isn't fatal. Actually it is fairly common. Fixing it is fairly straight forward. But, you need someone who knows what they are doing to fix it. The problem is that the air gap isn't sufficient at that point and corona is the result. Beware of someone who only wants to just gob on more tape and varnish and thereby further reducing the air gap.

rmw

 

[justatech]

the problem is there is corona, or atleast white marks, between several windings in a row. from what I understand it should be between phases not between 4 or 5 windings in a row. We are more concerned about that, and brush says its nothing.

 

[electricpete]

from what I understand it should be between phases not between 4 or 5 windings in a row. We are more concerned about that, and brush says its nothing.

Typically pd tends to occur where there are larger phase to phase and phase to ground voltages, often at the line end coils. Typically not at every coil. If you have pattern that occurs on every coil and does not seem to correlate with voltage, then consider non-corona causes. One non-corona cause would be chafing due to excess vibration. Another non-corona cause may be some kind of external contamination.

=====================================
(2B)+(2B)' ?

 

[electricpete]

What voltage is your machine?
Is the location of the mark approx 1" from the slot?

If > 6kv and at that location, sounds like degadation at the overlap between the slot-section carbon/semicon and the grading.

=====================================
(2B)+(2B)' ?

 

[electricpete]

Also, it may help to know if this is air-cooled or hydrogen cooled.


=====================================
(2B)+(2B)' ?

 

[rmw]

Brush (England) makes nothing but air cooled. A 7-290 may or may not have been built in England - some were/are made in the Czech Republic - but the 7-290 is an air cooled machine.

rmw

 

[justatech]

It's 12kv and it's about 5 or 6 inches from slot. The rep from brush said it's at the edge of the shielding/semiconductor. Official Brush report said nothing.

 

[electricpete]

I'm not familiar with the term "shielding" in this context. For me: Semiconducting = carbon tape or paint is used in the slot and only and inch or so outside. Grading (silicon carbie paint or tape) overlaps that just a little, and extends maybe 6" out before it stops. That must be the outer edge of the grading tape. I have never heard of burning at that location. Sounds like you are doing the right thing to bring someone in OEM... warranty?) help to evaluate.

Meanwhile, just thinking out loud, since grading layer is supposed to slope the voltage gradually from ground voltage at core end to line voltage at outer end, maybe the effective resistance of that grading layer is way too low for some reason, resulting in outer edge of grading being too near ground.

=====================================
(2B)+(2B)' ?

 

[ScottyUK]

The bigger BDAX-9 operating at 15.5kV suffered continuous problems with low / medium levels of PD and repairs were carried out on at least an annual basis, almost always in the endwinding and phase break areas. I'll see if I have any photos. PD normally leaves a pure white deposit - that looks fairly yellow. The winding looks darker, maybe a change in manufacturing process and/or location, or perhaps the whole image is just dark?


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[justatech]

I think 'shielding' was just an example used to explain some of the layers in the insulation to me.

Some of the spots look like, from what I understand, classic corona. White to a little yellow, starting at a point and a 'V' pattern away.

Other spots just look like a patch of white, almost pure white.

Electricpete, thats the best idea I've heard so far.

 

[controlsdude]

I thought corona is a problem with insulation breakdown and moisture issues. If your in the bay area do you have problems with salt moisture in the same room as this machine? Why it would be yellow, contanimation of some kind in the air?

 

[electricpete]

I think there is a lot of evidence pointing to the fact that the pd is occurring at the outer edge (endwinding edge) of your grading tape/paint. Here are 3 factors that point in that direction:

1 – It does not appear to be a tight area. Typical pd due to coils too close occurs at tight areas where coils cross over each other, where jumper is adjacent to coil, where blocking bridges coils etc... none of these appear to be the case.

2 – It is about 6" outside the slot where the end of the grading would be. And your OEM rep said something like it was the end of the "shield" .... can't think what else he would be referring to.

3 – It appears to be a generally circumferential pattern (*although perhaps you can check this better for yourself than we can check from the pictures*). Traditional pd due to close spacing does not follow a circumferential pattern. But circumferential pattern is expected at edges of stress treatments if they malfunction in a manner to create voltage difference at that edge.

The more typical malfunction of a stress treatment occurs at the junction of the semicon = carbon and the grading = SiC.. It gives a circumferential band about 1" outside the slot. This degradation can progress in 2 stages (as well detailed in "Electrical Insulation for Rotating Machines" by Greg Stone, Ian Culbert, etc al):
Stage 1: Burning occurs due to surface current flowing from endwindings to the slot. In the endwindings, the circuit to the conductor is completed through the winding capacitance. (I think endwinding conductive contamination will tend to increase this capacitance by the way). So we have current flowing and it is relatively small, but the heating occurs at a local high-resistance area which is the interface where the current must transition from one layer to another. Eventually the heating from this current burns the stress treatment open to create an open circuit, which leads to stage 2:
Stage 2: A voltage difference appears across the open circuit created by stage 1. Surface partial discharge occurs due to this voltage difference.


Again all of the above is well documented for the interface from semicon to grading, but I have never heard of anything like it at the outer edge of the grading (the endwinding side). We could reasonably guess that there are 2 possible mechanism for damage at this location (outer edge of the grading) that resemble the 2 possible mechanisms of stage 1 and stage 2 above. The two possibilities are:
1 – burning due to current which is passing from the outer surface of the endwinding, possibly through a coating / paint layer, down into the grading layer. This type of burning could discolor the coatings to create the yellow or other colors.
2 – surface partial discharge at the outer edge of the grading. It is harder to come up with a reason for this since it is not really analogous to stage 2 above. The best I can come up with is what I had suggested in the previous post.... if effective resistance of grading is too low than you could end up with a surface voltage different at the outer edge of the grading which creates partial discharge.

If it were one of our critical machines, I would call Dave Scherer of IPS LaPorte for his opinion. We have paid him several times to travel to other repair shops to provide his opinion on various visual partial discharge symptoms to help us figure out what we should do.

=====================================
(2B)+(2B)' ?

 

[rmw]

I am not an EE but I understand the concepts. Pete has hit on what the "shielding" is, although I am not familiar with that term. As I remember it, it extended out about ~4" from the edge of the iron.

I can tell you that Brush England speaks a different English than USA folks. Same language, different words or different meaning for the same words. I can't remember what they called this conductive tape. If I heard it I would remember it. Shielding doesn't ring a bell.

Ask your "rep" what those metal things on the end of the rotor are. If he says 'end caps' he is a Brit. If he says 'retaining rings' he is an American. If he says 'end caps' with an american accent, be suspicious of his/her background.

In either case, as I am acquainted with the rep that used to cover that area for Brush, I can state that this 'rep' must be a fairly recent addition to the Brush crew.

See if you can find out who did the report from the home office.

What vintage is this BDAX unit?

rmw

 

[justatech]

The rep from Brush was actually a sub-contractor out of Canada and he has been doing this for Brush for about 15 years. I think he thinks there is an issue, but he isn't allowed to say anything to us directly. The Brush report for our generator says there is no problem, we just need to pull the rotor and clean it.

The moisture isn't really an issue here.

The generator was manufactured in 2002, and been in service since 2005.