DIY cooler using well water

[BeanAnimal]

PEX is somewhere in the neighborhood of 0.35 W/mK

Ti is somewhere in the neighborhood of 15 W/mK

So about 42 feet of pex has the same thermal conductivity as about a foot of PEX (given the same diameter)

A few small lengths of Ti tubing from McMaster connected together with john guest fittings would outperform a big coil of PEX. However, as there are joints there is always a risk that the freshwater could leak into the sump. Passing the tubing through the sump walls with unisleals is an option...

This project is fairly straight forward and does not need a lot of engineering, but will take a bit of though with regard to safety and temperature control.

 

[dkeller_nc]

The parts I cannot find for doing this scientifically are info on the heat transfer per material and length of coil and flow rate? I'm sure it's out there I just can't find it.

You may have difficulty with this. Calculating the required heat removal (and therefore the minimum flowrate of water from the well) is straightforward, as BeanAnimal noted.

However, correctly estimating the actual heat transfer that can be achieved for a given material (titanium, for example), geometric configuration (shell and tube, coil and fin, simple coil, etc...), mixing in the cooler, etc... is vastly more complex.

In the first calculation, you're simply using thermodynamics to get the amount of heat removed required to keep your tank at a certain temperature.

In the second calculation, you must use heat transfer (conduction and convection, with a tiny amount of radiation) for a complex 3-dimensional shape to estimate the efficiency of your set-up and whether it will work to move the amount of heat you calculated. This is definitely done in my field, but it either involves using standard components whose characteristics have been empirically tested, or it involves finite element analysis modeling.

That's why if you're doing it as a "backyard" project, you calculate the heat removal requirement, multiply by 3, set it up, and take some measurements.

 

[BeanAnimal]

For those what want to "do the math"... They could vastly simplify the real world complexities (and arrive at that multiply by 3 answer in the long wrong) by only looking at conduction (Fourier's law) and ignoring the complexities of the surface properties, associated currents, convection and radiation... as the big elephant is the conduction with a submerged heat exchanger that has both sufficient internal flow and external flow around it...

That is, we simply calculate the surface area of the tube, look up the materials Thermal Conductivity and then apply the Fourier equation. Assume that most of the other variables reasonably cancel out, multiply by a little fudge factor (say 1.5) and call it a day...

If it were me and this were a permanent solution, I would buy a titanium heat exchanger coil and attach it to the well pump or for that matter the large holding tanks. Control the pump with a temperature controller. Add a spring check valve to prevent convective circulation through the loop, and call it a day. If the Ti coil cost too much then passing Ti tubes the who way through the sump through uniseals would be the next best option. They could be connected together on the outside with john guest style fittings. In that way a leaking fitting will not contaminate the tank with fresh water...


Have fun

 

[Pokerman11]

All and all it looks like this is a workable solution. Having a well water temp of 47D, this gives me a lot of cooling power.

Heat ex changer is TBD, going to find a real DIY solution. Broken chiller parts if I can find one seems good, if not perhaps a junk yard looking for something that does not react with seawater, like some titanium tubing from a who knows where. In the end even PEX will work if I throw in enough of it.

One final part to figure out is the controller. I have ready access to 12V sprinkler valves. A few years back I set up a well water system to water cool my overclocked computer. There it was easy, as when the power rail was on the valve was open . That was a way to REALLY overclock, but that's another story.

Any suggestions on how to DIY rig a temp controller to turn on/off the water?

 

[zachts]

Ranco temp controller and a 12v wall wart or transformer, whatever you've got around, just don't skimp on the controller. I'm pretty sure all Irrigation valves are 12 or 24 volt AC not DC by the way.

 

[Pokerman11]

Yes know sprinkler valves are AC. I have a landscaping lighting transformer just sitting around. Current thought is to use that Rancor control you mentioned to turn on the lighting transformer to energize the valve.

or just wing it and manually adjust, or simple timer that turns the value on during the lighting hours. Going a little more will cause my tank heater to kick in.