Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Water Cooled Busbar Design

Status
Not open for further replies.

Rhodrons

Mechanical
Feb 11, 2005
18
0
0
FR
Does anyone know of any guidance available for the design of water cooled copper busbars? I want to use copper tubes to run at high current (>5000 Amps) but low voltage (20 Volts) to feed a resistance heating furnace. I can find lots of information about the current carrying capacity of flat copper bars but nothing about water cooled copper tubes. I have seen photographs of water cooled busbars and they have the benefit of being flexible to cope with the movement of the furnace in service. Alternatively a copper braid inside a water cooled pipe might work but I still would not know how to do the design sums for the required copper area.
 
Replies continue below

Recommended for you

At that voltage, you can just make the bus cross-section big enough to handle the current in air-cooled mode assuming you have room.

Aluminum smelters use aluminum bus running at 40,000 amps and higher with no forced cooling of any kind.

5000 A shouldn't be too big a problem - unless you have space limitations.

Water-cooling would be a huge complication and a huge cost.

Large isolated-phase bus duct is routinely forced air-cooled, but it is running at 15-30 kV and is insanely expensive compared with a slab of copper or aluminum.



 
Conductor ampacity is determined by temperature. The I^2R heat loss in the copper tube needs to be carried away by the cooling water to keep the conductor cool- around 80C at the desired current. Keep the conductor temperature below 100C or the water will flash into steam and blow your bus apart.

You will need to use demineralized water to minimize stray current in the water lines. Demin water can cause corrosion problems. At 20 V, there may not be a problem with losses through the water.

Pick a size of copper tube. Calculate the resistance and the losses at your desired ampacity. Assume the cooling water will remove all heat, ignoring the heat los to the surrounding air. Calculate the heat transfer from the tube to the water using heat transfer coefficients for a heat exchanger. Assume some cold-water entry temperature and some hot water exit temperature <100C. That will be enough data to calculate the water flow required. Then build a mock up and test it.

If you use demin water you will need a water cooler (radiator or a refrigeration unit or external cooling water) and circulating pumps. If you use a garden hose in a once through system, you run chances with variations in conductivity.

Larger diameter copper tubes allow more water flow with less pressure drop and less resistance. Try different combinations until you get the best combination.

Use non-conductive tubing to connect the copper bus to the water piping.

I wouldn't use braid inside a cooling tube. Too many complications.
 
Check the lastest issue IEEE IAS (Industry Applications Society) Magazine there is an article in there about water cooled bus bar design.
 
Water cooled conductors are commonly used in induction heating equipment.

There are three basic types:
Hollow conductors, water cooled cables and flat busbars with pipes soldered or welded to it.

Current ratings in the range of 10 kA zo 100 kA are not uncommon in that business.

So, 5000 A as in your case so is a small installation to me.

With water cooling, removing the heat from a straight conductor is usually not a big concern. Suitable design provided you can operate at current densities 5 to 50 times larger than in a air cooled conductor. Usually the current density has to be restricted for economic reasons to keep operating cost low. In AC applications at line frequency or higher (as common in iduction heating) inductive voltage drop and local heating at edges is an important concern.

 
Status
Not open for further replies.
Back
Top