Considering a simple geothermal radiator... Thoughts?

I want to use a geothermal cooling loop for my watercooling rig. My reasons are simple: Less mess in the house and less heat being blown around the house. South Texas summers are brutal and I don't want my computer to fight my A/C.

I plan on burying the loop 4' underground. I know that deeper would be better, but I would not be able to dig down deeper. This depth should be more than adequate for my purposes.

I have two designs for the radiators.

The first is a simple helix design. All 1" copper pipe and fittings.


The second is a bit more complicated and has more copper. Again, all 1" pipe and fittings. In this design, water cascades down from the upper tier to the lower tiers.


Which would be better? I would assume design #2 because of the added copper and surface area. However, that would be a pain in the butt to build.

Does anyone have experiance with this? Any thoughts?

Thanks,
Mark

Build smaller models of each and run tests. IMO #2 looks to be a little better, but i am no expert.

The second design would be the common sense pick of the two. It has more pipe, therefore more surface area. As long as you are running a decent amount of pressure through the system, it should be good.

I'd re-think this, because:

1) 100 ft. of 1" copper pipe is not cheap!

2) You completely lose the portability of the computer

3) It's not a trivial amount of work.

4) You didn't show any calculations to indicate the materials and design were going to do the job.

For instance, why 1" pipe? Why 100'? Won't the top part (in and out) sections of the pipe freeze and burst in a Winter cold spell?

Seems like you're going to spend a lot of $$$, and work, for a system you have no assurance will work right, AND it completely removes any portability from the computer that you hope to cool with it.

I'd consider water cooling systems that set up INSIDE or adjacent to, the case to be cooled. They do the same job, cost less, and never freeze during the Winter.

I have NO experience with geothermal cooling, but I've buried a lot of 3/4" and 1 1/2" copper water piping.

Good luck with your project, whatever you decide.

Adak

I too have been toying with the idea of geothermal cooling loops for my computers.

However I planned to use slinky coils for my buried cooling loop for several reasons:
1. Flexible (easier to properly bury)
2. No problems with corrosion at the soil/loop interface
3. It is an industry standard product for home cooling.

My knowledge barriers are:
1. How many feet of buried loop do I need
2. How much pump do I need for the sytem

Adak said:

I'd re-think this, because:

1) 100 ft. of 1" copper pipe is not cheap!

2) You completely lose the portability of the computer

3) It's not a trivial amount of work.

4) You didn't show any calculations to indicate the materials and design were going to do the job.

For instance, why 1" pipe? Why 100'? Won't the top part (in and out) sections of the pipe freeze and burst in a Winter cold spell?

Seems like you're going to spend a lot of $$$, and work, for a system you have no assurance will work right, AND it completely removes any portability from the computer that you hope to cool with it.

I'd consider water cooling systems that set up INSIDE or adjacent to, the case to be cooled. They do the same job, cost less, and never freeze during the Winter.

I have NO experience with geothermal cooling, but I've buried a lot of 3/4" and 1 1/2" copper water piping.

Good luck with your project, whatever you decide.

Adak

Click to expand...

Well, hopefully he understands that geothermal means the loss of portability. Of course, geothermal is a good amount of work; between burying the pipe and actually assembling it, there is a lot to be done. As for why he chose 1" pipe, my only guess would be to increase surface area. But this does mean that he is going to need a rather powerful pump. As for the pipes freezing during winter, this isn't possible. At a certain depth(5' I think) the ground maintains a constant temperature of 50F. From winter to summer there is no change in temperature of the ground. The only problem would be the necks of the system. If he were to bury this in his backyard, that could pose a problem. But some insulation around the pipe would avert a potential disaster. Or, he could bury this in his basement provided he doesn't already have a concrete floor in place.

MarkS,
Might want to look around for calculations on depth and temps etc... 4' might not be deep enough.

But if it is what are your limitations on digging? If you have 100' of copper tubing it'd be better to make a huge long pass. Instead of packing a lot of surface area into a small space, span it out as much as possible. Do 4 lengths of 25', do a double pass with each pass having 2 25' in parrallel to keep restriction down. That would disappate the heat into the ground more efficently than having 25 4' passes.

Course you might not be able to dig a 25' long ditch Which is completely understandable.

If you are going to buy all that copper you might want to look at just buying some baseboard heater pipes and burrying them. Would be much easier.

I chose 1" pipe because of the surface area, but also because there would be less head loss and I would not need as large of a pump. The total length was what I ended up with after designing the radiators with the dimensions I had in mind. I am not concerned with portability. The computer will stay where it is.

Will this enter the house in a basement? The top loop has to be below frost level for you area, and the entrance and exit point in your ground loop should be below frost level as well to avoid freezing in winter. Also condinsation could be a problem at times. With that much of a run at that depth you will need a strong pump (Iwaki class) This wont be a cheap project to undertake.

learning curve said:

Will this enter the house in a basement? The top loop has to be below frost level for you area, and the entrance and exit point in your ground loop should be below frost level as well to avoid freezing in winter. Also condinsation could be a problem at times. With that much of a run at that depth you will need a strong pump (Iwaki class) This wont be a cheap project to undertake.

Click to expand...

Frost? What's frost? Seriously, I live in south Texas. I can count the number of days below freezing each winter on one hand with fingers to spare. The key here is to insulate the pipes very well and keep the liquid flowing.

Nope. No basement. The entrance and exit pipes will be above ground, but wrapped with thick layers of insulation and there will be a protective covering (aluminum box or simular) over them. I was planning on using an Iwaki 30 RLT that I am currently using on an aquarium. I think I'll need to up that to a 55 RLT or possibly a 70.

I priced 1" copper tubing at Home Depot. I'm looking at about $21 for a 10' length. I can probably find that cheaper. What I might end up doing is using PVC. There is less heat transfer with PVC, but it's cheaper and easier to use and I can do a MUCH longer run for far less.

I think it's important to understand that I'm not looking for sub zero temps or anything close. Regardless of how deep the loop is buried or what materials I use, the temps will drop. I'm looking at holding the system temps at around 65° - 70° F. I'm going to use a temp controller to monitor the coolant leaving the computer. When it gets below a set point, a valve will shut and the flow will be diverted away from the computer without shutting off the flow.

MarkS said:

Frost? What's frost? Seriously, I live in south Texas. I can count the number of days below freezing each winter on one hand with fingers to spare. The key here is to insulate the pipes very well and keep the liquid flowing.

Nope. No basement. The entrance and exit pipes will be above ground, but wrapped with thick layers of insulation and there will be a protective covering (aluminum box or simular) over them. I was planning on using an Iwaki 30 RLT that I am currently using on an aquarium. I think I'll need to up that to a 55 RLT or possibly a 70.

I priced 1" copper tubing at Home Depot. I'm looking at about $21 for a 10' length. I can probably find that cheaper. What I might end up doing is using PVC. There is less heat transfer with PVC, but it's cheaper and easier to use and I can do a MUCH longer run for far less.

I think it's important to understand that I'm not looking for sub zero temps or anything close. Regardless of how deep the loop is buried or what materials I use, the temps will drop. I'm looking at holding the system temps at around 65° - 70° F. I'm going to use a temp controller to monitor the coolant leaving the computer. When it gets below a set point, a valve will shut and the flow will be diverted away from the computer without shutting off the flow.

Click to expand...

Obviously this is not a whimsical idea and you have put considerable thought into it. I have seen other posts where the OP has a picture of their WC rig and in it is a single 90deg elbow. A lot of these posts earn quick suggestions to the tune of: "Lose that elbow, it is bad for the loop." I am wondering if 20+ 90deg elbows would be a deal-breaker in your case? Perhaps if the turns were smoother, like a wide 'U' at each end it would help. Just a thought.

I too am unsure about the right goethermal cooling capacity. I am prolly 2 weeks away from my ground loop cooling tests.

One thought on your designs is that having too much copper in a smaller area may not work well depending on the thermal conductivity of the soil. But i have no idea what your (or my ) soil's capacity is.

Here in my 46 degree latitude, i am contemplating going shallow for my copper trench.
My thinking is that in the winter, i will get cooler temps and no fear of freezing since it will be heated and moving. In the summer, i have a shaded location for the trench, and may experiemnt with a supplemental chiller, or a natural evap system by laying a soaker hose at a small trickle to moisten the ground and use the grass as an evap. Keeping the soil moist i would think would aslo help the thremal conductivity around the copper. Of course cost/availability of cheap water might well be the limiter to this.

On the copper, look at 3/4" copper coils that is used for water supply lines to houses. Home depot had the cheapest prices for it here in 30' lengths. The advantage is few joints, wide smooth bends (it bends easily). Don't get the thick stuff used for gas.

Take advantage of your property ... my line is going to a tree/house shaded section which also gets rainwater from the eavestroughs. For me going shallow should not be too bad (roots, clay soil, etc). For your climate, it may be too hot to take advantage of these factors. Am guessing that if you dont have a well shaded area, you will need to go deeper than me. even with shade you may still need to go deep if your average temps year round are high. And if you need to go deep there is probably no opportunity for an evap experiment even if you have abundant chaep water.

Just a few of my, as yet, untested thoughts.

The second option wouldn't work properly, water takes the path of least resistance so no/very little flow would go to the outer two pipes on each level.

What would work is if you had the inlet and outlet of each level in the corners diagonally opposite each other, that way each route through the pipe is of equal length so the flow would be even across the whole level.

1. its been done before and does work. te person owns thier home so can dig up thier back yard.
i dont think the amount of work is really a factor.
the amount of pipe on the other hand is.

you want a continuos spiral inward spiralling plate levels. a little hard to describe but instead of 90 bends or elbows its a wide bend that spirals in on itself and declines in the last close revolution and then spirals outwards as the decline levels out on the next lower level.
this will give more surface area without going so deep.
also you will not need 100 feet of copper.
but you will need a stronger pump than normal. for instance a T1 2 or3 velocity.

the biggest benifit you can get out of this is complete silence by keeping the pump outside and not heating our case up with exhaust air.

bending the pipe takes a bit of skill or you can have it done for you but your loop will not need to be that large and i would definately advise a LARGE res with large opening before the pump for changing and refilling the loop. you will have to increase the amount of antifreeze in your loop durring the winter.

I would think that a more spread out rad would work better, but you should do the math. You should be able to guess roughly how much heat you are pushing out, should be able to find the thermal conductivity of dirt somewhere on the net, then all you need is to figure up the area/volume your rad is and plug the numbers in and compare. I would give formulas but I'm to tired now, maybe in 6-8 hrs.
btw, wet sand/dirt conducts better than dry sand by a factor of 10.

I was thinking of getting under my house (always cool) and burying a few 55 gal drums and just circulating water through them without a rad. Haven't done the math yet so I don't know if it would work, just a thought.

Source

MarkS said:

I chose 1" pipe because of the surface area, but also because there would be less head loss and I would not need as large of a pump. The total length was what I ended up with after designing the radiators with the dimensions I had in mind. I am not concerned with portability. The computer will stay where it is.

Click to expand...

Are you going to dig this by hand or what? As pscout said, its better to use more area and not pack the radiator into a small space.

And even with 1" copper pipe, with 100' you'll have a huge amount of restriction. Even with just a straight length, let alone tons of 90 degree bends. And dont get a md30rlt, get a md30rz. RL is for max flow, RZ is for max head/pressure.

A geotherml system isn't meant to provide very cold temps. It has the advantage, though, of maintaining a constant temperature year round for very low cost. Geothermal, as a friend of mine puts it, keeps his house cold in the summer and warm in the winter. This is because of the constant 50 or so degrees(fahrenheit) that is found about 5 feet or so deep in to the earth. I think that if you are going to use a large pump with a generous flow rate, that angles aren't AS bad, but I would say that just buying a coil of copper tube and then stretching it to make a large spiral would be a good way to go. You may want to post a couple more pictures of some plans, or what you get accomplished as you go.

ahh treenode you found some good stuff

Reflex said:

The second option wouldn't work properly, water takes the path of least resistance so no/very little flow would go to the outer two pipes on each level.

What would work is if you had the inlet and outlet of each level in the corners diagonally opposite each other, that way each route through the pipe is of equal length so the flow would be even across the whole level.

Click to expand...

Agreed, I noticed that also.