Replacing aluminum outgoing bus bars with copper in 11 KV, 50 MW TG 1

[edison123]

We have a client with a large chunk of outgoing aluminum cylindrical (280 mm OD/260 mm ID = about 8500 sq mm cross section) busbars of a 11 KV, 50 MW, hydrogen cooled generator failing disastrously for about a length of 4 meters. All three phases and neutral busbars have failed and it was all a molten aluminum disaster in the pit.

I am planning to change these sections into copper bus bars of size about 200x20 mm rectangle. The generator current is about 3300 Amps.

Any one see any problems with this?

Also, I am wondering why aluminum outgoing busbars have been a standard for TG sets?

Muthu

http://www.edison.co.in

 

[ScottyUK]

Is this part of an isolated phase busbar? They are fairly sensitive to geometry changes because the magnetic flux levels are so high.

Regardless of bus type you should consider the bus bracings for fault withstand capability because the copper and aluminium sections will be have differently. The tubular alumium will be very stiff compared to the flat copper bending in the 'easy' direction, and discontinuities such as the transition from aluminium to copper tend to be weak points.

Why aluminium? Cost, weight, strength / weight ratio and ease of fabrication I suspect. In the case of IPB it allows for field welding of parts which would be virtually impossible with copper.

 

[desertfox]

Hi

You might encounter problems with differential thermal expansion between copper and aluminium when the section is replaced and agree with Scotty that the circular section is much stiffer than rectangular section.

Could the original problem be caused by the joints being incorrectly tightened?

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[edison123]

Thanks scotty & desertfox.

These are the photos of the failed busbars and the original ones.

As you can see, these round aluminum bars are anchored by 4x90 deg insulators against the outer enclosure.

As for the weight, since I plan to reduce the size (from 8000 sq mm for Al to 4000 sq mm for Cu), the weight may double, which is not a biggie in overall scheme.

The generator bushings have rectangular copper plates as terminals. They are using multiple thin copper sheets to connect the bushing terminals to the round busbars, which, as you say, is not a good idea because Cu/Al junctions as well as poor contacts. It is a big hassle to connect these thin sheets for all the nine bushings.

I take your point about round is better than rectangle for rigidity. But do we need that rigidity if the rectangles are properly anchored and expansion flexible joints are provided between sections?

Muthu
www.edison.co.in

 

[edison123]

Failed & Original busbars photos

Muthu
www.edison.co.in

 

[waross]

Any speculation as to the cause? Rodent? Dirt build-up and moisture? Bad connection?


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[desertfox]

Hi edison123

The rigidity of the busbars will be paramount during a short circuit so personally I wouldn't change the configuration, also you would be changing the busbars natural frequency so if there is anything nearby that generates vibration the bars might oscillate.
It would interesting to know how long these bars when replaced are lasting and investigate the cause.

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[ScottyUK]

That's a bit of a mess.

You should consider current crowding due to skin and proximity effects. Normally these aren't a problem at power frequencies, but at this size of conductor it becomes a factor which should at least be considered: how much of your copper will be active conductor, and how much will be contributing little to the process? Tubular conductors are the best shape for minimising problems due to skin effect, which I should have mentioned in my previous post. Proximity effect should be minimal as it is an IPB, except perhaps at the generator main terminalas and at the transformer bushings in the region beyond the main cross-bonds where the magnetic fields have a chance to interact with each other.

As it is an IPB you need to be aware of the image currents flowing in the outer shield, and that discontinuities such as tees and bends typically have imperfect field cancellation causing high external fields and heating in local steelwork. Watch your earthing too - the IPB shield should be earthed at precisely one place and the rest of the bus structure supported on insulators. The position of the main cross-bonds between the shields is critical to performance of the IPB and they carry the full image current so you must ensure that they are adequately sized and correctly positioned.

If you already know all of that then apologies.

 

[edison123]

Bill

I have no idea about the cause. The client doesn't seem to have done any root cause analysis. We were called in 3 months after the incident. So, no forensic evidence to look into. Strangely, none of the protection acted and the operator tripped the machine after hearing the bang.

desertfox

These are physically isolated busbars right up to the transformer terminals. So a phase to phase fault in the bus section is ruled out but the phase to ground fault is still possible.

scotty

I too thought skin effect will be an issue. But, I have seen 415 V, 3 MW generators (over 5000 amps stator current) using copper rectangular busbars without any problems in many sugar factories around here. May be I should split the busbars into 2x200x10 mm instead of 1x200x20 mm to reduce the skin effects? As for proximity effects heating the support structure, the size of the Cu bus (200 mm) is less than the present Al bus (280 mm) and so the proximity distance is increased?

I thought of using 150/130 mm tubular Cu bus but it creates the problem of connecting rectangular bushing terminals to the tubular bus and providing flexible expansion joints and also the difficulty of procuring such large Cu tubes.

I know you work with large generators. Are your outgoing busbars aluminum too?

P.S. I just checked with my sugar factory client. They are using 4x200x12 mm aluminum bus for their 415 V, 3 MW, 5300 Amps generators.

Muthu

http://www.edison.co.in

 

[ScottyUK]

The skin effect just makes the conductor less efficient, so if you can dissipate the extra heat then it should be OK. The problem with IPB is that it isn't normally ventilated so it just relies on convection within the cavity to dissipate the heat generated in the busbar. The large surface area of the tubular conductors helps with this process.

Yes, the IPB on our generators were welded aluminium. Copper of that size would have been very heavy and very costly. I'll see if I can find any constructional details later on - my PC is not behaving itself very well at moment.

You are replacing the outer shield with a similar conductor, not the just the inner bar itself?

 

[edison123]

Thanks scotty.

Yes, the heat dissipation has to be by natural convection to the outer covers.

The two outer covers are just ms sheets at ground potential, which I plan to keep.

Muthu

http://www.edison.co.in

 

[ScottyUK]

Hi Edison,

I'm quite surprised the outer sheets are mild steel. Normally these are aluminium, and the outer shield provides a path for an 'image current' which is of opposite polarity (antiphase) and very nearly equal magnitude to the busbar current. The image currents are key to the IPB having a low external magnetic field.

I have a set of drawings for one of these structures which might be of assistance - or not - but I probably shouldn't post them online. Is there somewhere I can find an email address?

 

[edison123]

Thanks scotty. It's g**ildotcom and name is amkeew123.

Muthu

http://www.edison.co.in

 

[ScottyUK]

Something on its way to you.