bus duct between transformer and frequency converter

[engelectrical15]

Hello,

I would like to learn your opinion on the folllowing.

We have a frequency converter (VFD) for a machine drive application (VFD rated power is approximately 20 MW).
The frequency converter is fed by a transformer converter outdoor located. Teansformer primary voltage is 33 kV.
The transformer is a multi-winding phase shifting transformers for 36 pulse VFD/transformer topology at input. Transformer turn ratio is 1 to 18 (quite high).

The current design is based on bus duct connection between frequency converter and its transformer. The bus duct shall cover distance of less than 30 feet.

The question is what is your experience on this type of connection (bus duct) given the high turn ratio transformer and resulting high number of connections between transformer secondary and VFD ? are there any technical limitations or issues you could identify?

Any orientation is appreciated. I can try and provide complement of information as you would require.

 

[engelectrical15]

Any help ?

Thanks

 

[jraef]

36 pulse? So you are planning on having 18 separate bus bars between the drive and the transformer? Seems a bit excessive if you ask me. I don't know of anyone who does this even for 18 pulse drives, most people use cable. What is the reasoning behind doing it with bus duct? And what is the drive voltage? 18:1 on a 33kV primary looks like 1833V to me, so something isn't right with your numbers. But assuming 10kV for the drive, you are probably looking at less than 100A on each secondary circuit. Why do you need bus bar?


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[engelectrical15]

jraef, thanks for the feedback.

bus duct was specified primarily on reliability ground.
When you mention that you don't know of anyone who does this, what is the principal reason / technical limitation for them not doing this with a 18:1 and 36 pulse topology? Risk of failure of the connection (mechanical failure such as bending problem, overheating of bus bar, etc) ?
Are there any concern on safety later on being this a quite big power rating?

I will check anyway all the voltage and power I've given are correct;

 

[engelectrical15]

Something I would like to revise in the original data.
The transformer primary side is 11 kV and the transformer is rated 15 MVA. The other info are unchanged.

Appreciate if someone can help me on the issue ?

"If you want to acquire a knowledge or skill, read a book and practice the skill".

 

[jraef]

I just think it is serious overkill to use custom made bus bar arrangements for something so simple. The "reliability" factor is not there if you ask me. To use bus, you have just as many connections, maybe more. And by the way, I was wrong about the number of connections, because most of the large multi-pulse MV drives now are using the "cascaded H bridge" topology, where there are low voltage "power modules" for each phase and pulse in series strings to get to the required output voltage, each with it's rectifier. So that means 3 conductors PER MODULE, per pulse, and if the output is 10kV, I think you might end up with have 108 conductors, not 18! Look at the attached image of an 18 pulse 4kV drive. The yellow conductors on the right of the transformer are the ones you are talking about, but you would be looking at at least twice as many if the voltage is 3.3-4kV, more if higher.

I think if would be difficult (if not impossible) to have hard bus connections to all of those connection points in the VFD power modules, so you will likely end up with cable from the ends of the bus bars to the rectifier inputs anyway. All you are doing then is exchanging longer wires for shorter wires and adding another connection in each one, which increases the possible failure risk.




"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[engelectrical15]

Jraef, first off, thanks a lot for the response. great help indeed.

The rating at the drive is 5.0 kV. Think the topology is not H drive in my case but old fashion topology (based on power cells). Possibly the number of connectors are about half of what you described but still this remains a lot.

Looking at your picture, the transformer and drive are at same location / nearby. In the case I am concern with, the transformer will be located externally.

Agree with you that most likely the problem is how to fit all the bus bar connections to the power module which normally is a quite compact thing with small dimensions like in your picture - so if at all feasible, it will be crowded and dense.

What I still can't figure out, is the failure mode which can be involved. I mean when you fit so many connections I guess you do need some special or custom made design. In this case and in your opinion, what would make the failure risk higher when compared to a cable type approach of connection ?

"If you want to acquire a knowledge or skill, read a book and practice the skill".

 

[jraef]

Locating the transformer of a multi-pulse drive away from the drive itself is always problematic. So I already pointed out how I thought the failure risk is increased, IF you end up making the final connections to the power modules via wire. If you just used wire, you have two connections; one at the transformer terminal, one at the drive terminal. If you use bus bar, you have a connection at the transformer to the bus, then at the other end of the bus to a piece of wire, then the wire to the drive module terminal. Each connection is going to represent some resistance and heat, expansion / contraction etc., which all increases the risk of failure. The fewer the connections, the better. Making the individual bus bars connect directly to the drive power module connections and keeping all of those complex bus bends from toughing each other would be an extreme challenge even to the best of bus bar benders.

Are you attempting to locate the transformer outside so as to avoid the heat rejection? If so, the two times I have proposed this have resulted in the cost to make it work being so extreme that it was untenable.

Have you already purchased the drive in question" There are alternatives that could give you similar performance as far as harmonic mitigation, but without the multi-pulse transformer. That way you can use a standard transformer and standard bus duct.


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington