Optimizing Cooling Efficiency: Factors to Consider for Coiled Tubing Length

[origen87]

How will water flow inside a copper conductor effect the flow of current in the conductor and vice versa?The idea being to cool a tube 1560 metres long with an i^2r loss of 7kW.

 

[vk6kro]

The water will reduce the total resistance slightly if it contains dissolved salts. It will conduct some of the current.
This would be a good effect and not a problem. Compared with copper, the conductivity of tap water would be trivial.

However, the water will have to be electrically insulated since it will have the same voltage as the copper pipe.

This would be difficult, but probably easier if you cool then recycle the water rather than get new tap water all the time. Environmentally it would be better to recycle the water, too.

Have you thought about insulating the water?

 

[Q_Goest]

Hi vk6kro,

The water will reduce the total resistance slightly if it contains dissolved salts. It will conduct some of the current.
This would be a good effect and not a problem. Compared with copper, the conductivity of tap water would be trivial.

However, the water will have to be electrically insulated since it will have the same voltage as the copper pipe.

This would be difficult, but probably easier if you cool then recycle the water rather than get new tap water all the time. Environmentally it would be better to recycle the water, too.

Have you thought about insulating the water?

Thanks for that. Right, the copper has to be electrically insulated from the rest. That shouldn't be so hard if there was a plastic section at the inlet and outlet of the pipe.

This follows a discussion I've had with origen87 https://www.physicsforums.com/showthread.php?t=346227".

 

[Bob S]

Two things
1) If the flow of water is too slow, the water will boil, and the boiling will significantly reduce the flow rate unless you increase the pressure substantially. It's better to increase flow rate to prevent boiling.
2) Calculate the Reynolds number, and the hydraulic pipe equations for pressure vs. flow rate to make sure you have enough pressure. Small dia pipes require lots of prsssure to maintain adequate flow rate.
Bob S.

 

[vk6kro]

Thanks for that. Right, the copper has to be electrically insulated from the rest. That shouldn't be so hard if there was a plastic section at the inlet and outlet of the pipe.

No. The water itself has to be insulated because it is a conductor.


Good one, BobS.
I wondered if convection cooling alone would be enough with 7.5 KW spread over 1560 metres.
That is only 4.8 watts per metre of pipe.
Still, it must be a problem or he wouldn't be asking. Maybe the pipe is very small.

 

[vk6kro]

You could probably use distilled water. This would avoid the insulation problems.

I calculate that you would only need about 75 litres (about 20 US gallons) to fill the pipe and maybe 3 or 4 times that to pump the water back to a cooling apparatus.

 

[Bob S]

Could this copper tube be a magnet coil, or is it long and straight? I have used de-ionized water in magnet coils. How many amps per square cm of copper?
Bob S

 

[vk6kro]

Just reading the other thread, the current is 200 amps and

1 inlet n 1 outlet will dD=8.65mm and ID=8mm.6 such layers of coil are wound on top of each odr,between these layers we are planning to have NOMEX paper covering.And the hollow copper will also be covered with PICC insulation.

So, that is pretty thin pipe. I make it 8.5 sq mm of copper in cross section.

He doesn't say what voltage is involved overall, but there would be 37.5 volts across the pipe.

Yes, sounds like a big magnet.

 

[Phrak]

If the water is circulating, I think it will need to be broken by a fountain--unless you can live with the corrosion due to electrolysis.

 

[origen87]

thanks for your responses.
well its 433V and arnd 140A,its a coil,but den as it is a thin cu tube,how can i insulate the coil from dis water?
and will continuous expansion and contraction of water help it remain at low temps so that it can be recycled in the tube?
i am planning to take current out of this tube, so the water to be recycled has to be taken out before. how can we prevent water(with current) from entering the expansion-contraction vessel?
and could anyone help me with the expansion contraction unit?

 

[vk6kro]

Distilled water is quite a good insulator.

I have a friend who is water cooling a valve amplifer which has 1200 volts on it. He just runs plastic pipe to it and has a normal earthed water pump circulating the distilled water. He has measured the leakage and it is negligible.

Eventually the water will become slightly conductive as it absorbs Carbon Dioxide. This won't happen unless the water is exposed to the air, but I guess you could check the pH of the water periodically.

So, the problems will be mainly plumbing ones and you have some expert help already on your other thread.

 

[origen87]

thanks a lot to all here,the information you provided was of immense help, i will try working out more and come back 2 u...thanks:)

 

[Bob S]

Here is a site that has an on-line pressure drop calculator. (There are others):
http://www.engineeringtoolbox.com/hazen-williams-water-d_797.html
I entered 5100 feet length, 8 mm dia, and 1 gal per minute, and got a pressure drop of 686 psi. That's a lot of psi. How many gpm do you need for 7kW? In our lab, we usually have several parallel water loops to limit the pressure drop to about 50 psi.
Bob S

 

[origen87]

hey i tuk length(ft)=25721.78...flow rate=.6023332gpm,,,,and with a new diameter of 17.78mm=.7inch...we get pressure drop=28psi...flow rate was calculated from Q=mC(T2-T1)...temp change was assumed from 100-25...12kW heat was to be absorbed...from that calculate flow rate,,correct me if i am wrong.do we have any other option than changing the dia of the pipe?

 

[Bob S]

The design criteria, often used by magnet coil designers, is to

a) keep current density in copper to less than 600 amps per cm-square.

b) keep pressure drop of (de-ionized) water to less than about 50 psi by paralleling water channels. This may require cutting the coil into sections (pancakes) and splicing in the water connections.

c) if the current is ac (pulsed), consider the eddy current generation in the copper due to the pulsed magnetic fields.

Bob S

 

[origen87]

so increasing diameter is the only solution?

 

[Q_Goest]

Can you bring the water out every few hundred meters and put cold water back in at a higher pressure as shown on attached? That way you can accommodate the original tube material while reducing pressure and temperature drop through each section. I've thrown some numbers for pressure and temperature on this just for clarity, but they can be any value.

 

[Q_Goest]

Hi Bob,

Here is a site that has an on-line pressure drop calculator. (There are others):
http://www.engineeringtoolbox.com/hazen-williams-water-d_797.html
I entered 5100 feet length, 8 mm dia, and 1 gal per minute, and got a pressure drop of 686 psi. That's a lot of psi. How many gpm do you need for 7kW? In our lab, we usually have several parallel water loops to limit the pressure drop to about 50 psi.
Bob S

I had a look at the calculator. It uses the Hazens-Williams formula instead of Darcy Weisbach, which is ok and gets you in the ball park but doing the equations for this particular geometry, this calculator underestimates pressure drop by about 30 to 50%.

The real problem with using this calculator however, is that if you put in the actual length of tube, you're not accounting for the additional restriction presented by the turns of the tube. In this case, the tubing is coiled which means the equivalent length can be many times greater than the straight line length. Without knowing the actual radius the pipe is coiled at, we're shooting in the dark on pressure drop. I'd suggest using an equivalent length of up to 4 times actual length for a coiled tube. It's probably better than that but for the purposes of sizing a line, it's best to be conservative on the high dP end.

 

[Bob S]

Hi origen87-
I wish you would tell us more about the use of this tubing. We do know that the current is 200 amps. Therefore by the 600 amp/cm^2 rule, the cross section should be about 0.33 cm^2.
Do you want round or square tubing? Square tubing is better for coils because it stacks better and is more resistant to the tube collapsing in bends.
Do you want insulated hollow buss bar? Magnet wire usually can be supplied with nylon insulation, ready for winding into a coil.
What is your specific application? Is it a magnet? What kind?
Bob S

 

[Q_Goest]

I'd suggest using an equivalent length of up to 4 times actual length for a coiled tube. It's probably better than that but for the purposes of sizing a line, it's best to be conservative on the high dP end.

I just checked to see how much the equivalent length changes with coiled tubing and found the equation in Crane TP #410 (attached excerpt). Looks like the equivalent length is much less than the 4 times I'd quoted. The attached provides a formula to calculate that depending on your geometry.