Looking for advice on DIY chiller

"Roughly 2' of the Ti will be out of the water and not transferring heat to the sump. "

THis has changed.... due to the configuration of the "stuff" already in my sump, all of it will be submerged. This is part of why I am using epoxy to ensure no saltwater gets into the fittings.
 
I would recommend surrounding that Copper coil with some mesh or something to keep random stuff from the lake getting caught up on it. And a Nylon rope to get it back if you need to.

Otherwise it looks great, and if you need to increase capacity you can just add an other manifold.

And if you insulate the pipe when you burry it(just some simply expanding foam stuff should work fine) and set a small pump tp keep things moving every few minutes, you wont have to worry about it freezing.
 
I am debating upon insulating the buried portion. Much will depend upon how deep I can get it. As for increased capacity.... we over-built this tremendously....at least I believe so :)


The loops are in place and filling. This is in the sump:
Sumploop.jpg


The lake loop is in too, but is about 13' down and my camera doesn't like to get wet.
It's filling now, and I will test-fire in the morning. Dummy me, I should have bought another bag of clamps! I put a fill line with valve and vent line with the same in the sump room. The vent is the highest point in the loop. I also put 4' of clear vinyl from the pump outlet to the Ti loop, so I can "see" the water flow.
Wish me luck!
D
 
It will work splendidly my friend! Great job!

If you are going to insulate anything, just insulate the return line from the lake, bury it and call it a day :)

It's going to provide you with a massive amount of chilling :) Your thermal concerns are solved. :)
 
Very nice work danno14. Its nice to see somoeone actually building one of these Geothermal chillers. Please keep us posted with some specs once you get this thing into full operation.
 
I have the pump plugged into a single stage controller set for a 1 degree swing. I will loosen that up (2-3deg) as I become more comfortable with the setup.

I find myself wishing for a hot day :)
It's been in the 80-90 degree range for the last 4-5 days, and today has decided to be cool. Figures :rolleyes:

I didn't design in an effective way to measure all that would be nice to know, but here are a few details:
The water temp after it has passed through the sump exchanger is 70-71f.
The pump is rated at 53w, but draws 65 when it kicks on. Then it drops to 35 while running. (Yah!)
I have not yet measured the time to draw down tank temp, but am planning upon it (2 degrees = X minutes).

The flow is visible through the vinyl tube when the pump is running, and I am glad I added that feature.
I was concerned about trapped air bubbles in the exchangers, and they still may not all have been worked out. A small one makes its way to the bleed-off valve every so often. The flow seems mild but pretty good....and it works :)
The biggest fear yesterday was the lake loop leaking. It ran from 2am last night and I needed to top off with about 1/2 cup of fresh water. I attribute this to air working its way out.


FWIW, I am calling this system a lake loop geothermal reef tank chiller (words to help the search engines)

That's all for now. D
 
Last edited:
the only info you need to guage capacity is as follows...in order of importance:

In/out temps for the Cu manifold
In/out temps for the Ti manifold
Flow rate of coolant
Length, diameter and wall thickness of Cu piping
Lenght, diameter and wall thickness of Ti piping
 
I was referring to the "as built" measurements. The calculations were performed prior to establish a design and shopping list :)

Accordingly:
In/out temps for the Cu manifold don't have this
In/out temps for the Ti manifold don't have "in" temp
"out" temp is 70-71f with tank temp at 78f
Flow rate of coolant pump is rated at 13.7gpm. unknown what actual is
Length, diameter and wall thickness of Cu piping 6 pieces 8-10 feet long (each), 3/8" diameter, .032 wall thickness
Lenght, diameter and wall thickness of Ti piping 4 pieces 6' long (each), 3/8" x .022

As mentioned and pictured, the Cu ant Ti lines are run in parallel.
 
Get that copper out of your system. It will kill your system, hands down. Use TI throughout the system ONLY. That is definetly 'HOKIE'
 
<a href=showthread.php?s=&postid=12879455#post12879455 target=_blank>Originally posted</a> by Reefer Brian
Get that copper out of your system. It will kill your system, hands down. Use TI throughout the system ONLY. That is definetly 'HOKIE'

Re-read the thread. The copper is only in contact with the lake water and the water that is circualted in the CLOSED LOOP system. The only metal in contact with his tank water IS titanium.
 
<a href=showthread.php?s=&postid=12879455#post12879455 target=_blank>Originally posted</a> by Reefer Brian
Get that copper out of your system. It will kill your system, hands down. Use TI throughout the system ONLY. That is definetly 'HOKIE'

It is bothersome when somebody enters a conversation without taking the time to become familiar with the concepts and context of the conversation... it is even more bothersome when the entrance into the conversation is an insult.

The system described in this thread is far from "hokie", it is well thought out and based on sound principles.
 
<a href=showthread.php?s=&postid=12859829#post12859829 target=_blank>Originally posted</a> by danno14
I was referring to the "as built" measurements. The calculations were performed prior to establish a design and shopping list :)

Accordingly:
In/out temps for the Cu manifold don't have this
In/out temps for the Ti manifold don't have "in" temp
"out" temp is 70-71f with tank temp at 78f
Flow rate of coolant pump is rated at 13.7gpm. unknown what actual is
Length, diameter and wall thickness of Cu piping 6 pieces 8-10 feet long (each), 3/8" diameter, .032 wall thickness
Lenght, diameter and wall thickness of Ti piping 4 pieces 6' long (each), 3/8" x .022

As mentioned and pictured, the Cu ant Ti lines are run in parallel.

how are you getting the "out temp" for the Ti manny? and is that after its gone through the tank water?

if your using a temp probe just swap it around to get the in temps and that should be all I need to get some kind of baseline.
 
If you know the flow in GPM, the inlet water temp (at the input side of the indoor Ti coil) and the outlet water temp (at the output side of the indoor Ti coil) then you can easily calculate the BTU capacity of the Ti coil for the "summer" water temperature.

Because the winter will bring colder water, the "winter" capacity will likely be more.

You can take it a step further and do the same calcs with the outdoor copper coil or staggered at the input of one and output of the other to get the mean BTU capacity of the SYSTEM with regard to both the Ti tank coil and the earth that the feed lins run though.
 
Cooling loop

Cooling loop

how are you getting the "out temp" for the Ti manny? and is that after its gone through the tank water?

EnderG60- I have a bleed port with valve on the "out" flow side of the Ti loop. It lets me put a probe into the water flow and measure the "after the sump" temp of water as it has just exited the Ti. Perhaps I will put another such port in the inlet side, but other than to get measureable numbers I don't really want to (more joints). I am more than comfortable with the capacity calculations performed on the system as it stands, once I get the wrinkles ironed out.

Meuser & Bean- Thanks for the defense.... I felt it was not worth the time or effort = the reply I posted :)

All- Thanks for the kudos...... My project pales by comparison to what I have seen many on here accomplish!

Now the latest- I noticed a level drop in the loop and decided to pressurize it. I hoped it might be from air bubbles....Sure enough, the pressure caused a big-time drain. Of course, the leak was in the lake. Pulled my mask on, went for a swim, and pulled the HE (heat exchanger) up. The problem was a weak glue joint on the PVC manifold, that 80psi made apparent. It's now fixed, and setting up overnight. I will fashion a cover for the assembly and drop it back in tomorrow. This time I will use scuba gear to place it as well.

A note for others- If you decide to try something similar, make sure you come up with a way to get the air bubbles out of your HE loop. Mine is NOT the most conducive for doing so. My plan is to work them out with the HE under water, then attach the polypipe. This way the pump is not trying to "push" the bubbles out from the loops, and should ensure good flow/all tubes filled.

An easier way might be to loop the tubing (like my Ti) instead of coiling it, then sloping it downward from the manifold in such a way as to allow the air to move up naturally. Space underwater and positioning preclude this as the "best" for me, but I may end up going this route if my idea doesn't work well enough. The same would be important for earth/buried systems too.

More as it comes up!
Thx, D
 
Last edited:
Looks great! I wouldnt be too concerned about flow being a factor in the cooling with that small of piping, beyond a minimum point of flow the cooling wont go up with the more flow you add. It isnt really convective like evaporative cooling, the water just acts as a heat transfer medium but since its enclosed, the faster you go, the less time the water will cool in the lake before it returns, so the water just acts as a really good conductor. The system's cooling ability will depend more on the flow and ambient temp of the lake you are cooling the copper in, as well as how much copper in the first place. I would say much past 300gph is worthless, maybe even just 100-200 would be enough, any more, and you are just adding pump heat to the system. The only catch possibly being how well insulated the pipes are between the lake and the tank... you might lose all your cooling potential if the pipes heat up the water on the way... making the cooler useless. Also, a water cooled pump will work against you... something like a small DC motor would rock here.

Something else you might consider is just a tank of water on your basement floor (not as far away), or just copper pipe right on the concrete floor. I used to cool a 500w computer like this... just ran about 50 feet of copper pipe along the basement floor... the copper was cooled by the concrete. A water tank (or even a thin pond liner) just aids in the transfer of heat better from the concrete to the pipe.
 
I wouldnt be too concerned about flow being a factor in the cooling

I agree... I also think that the heat added by the pump is not an issue. It's drawing 38 watts while running, so if we assume (incorrectly) that all 38 are converted to heating the water, this loop had BETTER be able to offset it!

The HE is back in the lake, and the system is once again running. Final placement on the bottom will occur once I am certain all is as it should be. The test right now is to find out how long it will take to drop the system temp from 78.8 f to 77.0 f. Total of around 400g to cool.
 
Last edited:
Back
Top