12 Feet Under: 1000 Square Feet of Geothermal PC Cooling

[shazza]

I admit, I've been reading this thread wondering if this was real - all I can say is "WOW." Kudos to Romir.

[Romir]

I admit, I've been reading this thread wondering if this was real - all I can say is "WOW." Kudos to Romir.

Thanks! I waited until almost everything was on hand before starting the log for that reason. I wish I spent more time on the initial write up though. This will deserve, and get, a re-write and summary in the end anyway.

Today's update:

The geothermal piping was easily uncoiled and straightened out at lunch time this afternoon. The (SDR11) wall was thick enough that kinking it by hand seemed extremely unlikely.

Half of the piping loosely pulled to where I took the picture. When I let go it sprung that far away from me.


A short while later. The bend at the very end will be handy for exiting the ground. The rest was straightened out a bit more.


The corner of the house where the rest of tubing vanished around. This is the general location of where the geothermal trench will go.


Tomorrow morning we're renting a ditch witch to replace a shorted out power cable under a gravel road at work. There should be enough time left to dig my return trench too. It'll fill in some after next weeks forecasted rain, but re-clearing it with a trench shovel will go quickly. I'm still debating how much to spend on the pipe insulation. At least the run to the pit is shorter now. There's more cooling coil length, less line temperature loss or gain, and less insulation outlays.

I should start zip tying the coils together this weekend once the return lines lengths are finalized.

The 1-wire monitoring components have been selected and ordered. I hope they're all in stock. This data logging is going to cost 3/4th as much as the loop! The ground water moisture sensor, control board for it, and the indoor humidity sensor weren't cheap.

The planned monitored items, and currently purchases sensors are:

Water in to house temperature - inline (water in from ground)
Indoor humidity and ambient temperature - (dew point monitoring eventually)
PC area water out temperature - inline (PC heat dump monitoring)
Utility closet water out temperature - inline (post pump, water out to ground)
Ground soil temperature by coils - in dirt (monitor the soil temperature to compare to the water ins)
Ground moisture sensor by coils - in dirt (learn how much moisture is there and how it affects the temperature)
Outside temperature - air (outside ambient to chart against the buried sensor)

I'm not sure if its worth placing temperature sensors at the ends of the buried pipe's insulated area. It would measure the insulated pipes effectiveness but the results also wouldn't be in-line temperatures like the indoor water sensors. Only one chance to do it... there's still time to order more of the sensors from Maryland.

Another temperature sensor buried twelve feet down would be nice as a control against my heated grounds one. Sending one down away from the coils in a pvc pipe my future plan. Extending the trench to install another one would be simpler though. I'd also like to monitor the temperature at the bottom of the hill and possibly install a moisture sensor there too.

[Philwong]

WOW!!! Insane stuff man...

Phil

[Romir]

Today's update:

After a bit of debate and re-reading the VT geothermal page, the coil's trench is once again planned for the bottom of the hill. It won't be as deep because it'll be constantly sinking and caving in. We're going to have to dig and lower the coiled piping down within minutes of digging each 5-6 foot section of the trench.

Having a moisture and temperature sensor with the coils will objectively reveal how this project works out. I'll plan on burying two temperature sensors near the top of the hill for comparison purposes. One sensor should be at the same depth, and the other can go "12 feet under". Something should be... If that location has more suitable temperatures during this summer and the next winter, look for a second coil loop project one year from now.


There was time to borrow the rented ditch witch for the return lines. Here it is trenching down the hill, going up is safer.


This is dry sandy so far. There wasn't much debris near the top of the hill.


Half way down a tree stump was encountered. Trying to go around it to the left was futile.


The finished trench. It looks short from this perspective. The soil was more moist every inch of the way. The pit will connect to the end here and be dug towards the right.


The side view.


The neighbor's golden retriever performed a surprise inspection of the operation.


The trenches depth is about 2.5 feet with a 4 inch width. I'll have to clean it out once it rains next week. The thicker pipe insulation from Grainger won't arrive in time.

Time to take some measurements and then head to Lowe's again.

[sirheck]

Looks good and good luck.
I think you might need more than one gallon of distilled water though.

[Shocker003]

Well i don´t understand how you intend to get good temps with this cooling technic. This technic is being used to keep houses warm during winter in Germany. I will be following your worklog, all the same my respect for taking a different path in cooling your PC.

[rge]

Once you go about 6 feet underground, most of the planet with normal terrain has ~same temperature between 10C and 15C. So if your cooling a PC, it can work because you can cool the water down to 10-15C and with enough surface area, keep water in that range during cpu load...much better than my 6x120 rad with water temps 30C at load. If outside ambient temps are 0 C or below 0 C in winter, that 10-15C temp underground is then a source of heat for a house (with supplemental heating to get higher to 22-24C whatever is comfortable). And animals that dig underground may be cooling themselves by escaping 40+C ambient temps in summer, or heating themselves by escaping -20C below temps in winter.

[Kanzy]

This is truly something else, go for the record, Good luck Romir, looking forward to it.

[CptDreadFlint]

Jesus Christ called. He wants his "holy mother of God" back.

[Romir]

Well i don´t understand how you intend to get good temps with this cooling technic. This technic is being used to keep houses warm during winter in Germany. I will be following your worklog, all the same my respect for taking a different path in cooling your PC.

Thanks. Look at the x = 12ft line on this image. That's basically the condition my piping will sit in and cool my loop to. It shouldn't get warmer than 18c or so at the peak on Oct 29.


Today's quick update:

There's not much to report today. I finished the reservoir and connected all the utility closet component's fittings with some sealant. They're all ready to be strung together.

That sealant is marvelous for liquids and gasses.


Bottom fitting of the assembled reservoir. The piece of piping was JB welded there to create a safe spot to put silver in the bottom of the res.


Look carefully and you can see the outline of that thin pipe.


The finished reservoir. It'll be mounted by steel bands to exposes studs in the utility closet.


Couplers and barbs mounted and sealed to the pump.


Smaller 1/2 barbs connected to a bushing on the King 5 gpm flow meter. I plan to mostly use this for testing purposes near the pc and probably won't keep it in the loop all the time.


Its going to rain tomorrow so I don't have anything planned. Coiling and zip tying the piping together is the next big step now that the trench size has been finalized. Keeping it under 20 feet (the plan all along) will allow the excavator to remain in one spot and dig the entire run. We're still targeting 10-12 feet, even in the wet muddy soil. The coiled loop will be ready to drop down in under a minute from the final muck lifting pass.

On a side note, I feel like an idiot for mis-reading this diagram.

.

The WET soil lines are the left most in both groups. I thought they were the outside ones which didn't make much sense. So at 12 feet, wet soil has a temperature rise equivalent to light dry soil at 20+! If this is accurate for my location, the ground temperature won't go above 18 degrees celcius ever.

Winter ambient temperatures in the summer, near silently. That was the lofty goal from the start.

I'll believe it when I see it.

(please be true, please be true, please be true)

[c22]

TBH I saw a project identical like that in polish forums, I will try to find it out again just for pics. But all I can tell is - his temperatures were outstanding for a totally "passive" cooling.

[Romir]

I only have time for a quick update tonight:

Before it started raining I was able to get in a little practice at coiling the piping. Starting 1/3rd into the length is rough. Every loop requires turning over a 200 foot length of straightened pipe. Its very tempting to scrounge up a 4x8 sheet and two 2x4s to create a center channel to work in. Holding everything together and sizing the loops equally is awkward. This has become a waiting game so the inefficiency of this doesn't really matter. It's more engaging than my daily cardio at any rate.

I settled on a package of 24 6' Black Tubular Polyethylene Foam Pipe Insulation. If "This item normally ships in 5-8 days" holds true, then it will be the project's bottleneck. The 1/2" ID Norprene tubing shipped today, along with the impulse buysmall blazed plate heat exchanger with 1/2 barb fittings. At $45 shipped with the barbs it was too hard to pass up. Hobby boards is also processing my order now and I haven't heard from the vendor with the temperature sensors yet.

Above the ground loop testing is still on for this week. Wed & Thurs rain/sleet/snow be damned. The coiling and under the house piping can both be finished in a couple late afternoons. Insulating and protecting the holes into the house from insects will take a bit of thought.

[scook9]

And here I thought I was hardcore taking my laptop on my front porch in the winter haha......the groud temperature should be more then enough to keep your system cool. You also have to take into account the ability of the groud to sink ALOT of heat.

Are you at all concerned about freezing air temperatures in the winter? It is getting into the single digits a few nights a year here in northern virginia (not sure what part you are in) so if there is a part of your piping that is exposed to air that should be a concern.

[Romir]

And here I thought I was hardcore taking my laptop on my front porch in the winter haha......the groud temperature should be more then enough to keep your system cool. You also have to take into account the ability of the groud to sink ALOT of heat.

Yeah, this cheap coil of HDPE piping is going to thrive in the deep, wet ground. There's so much contact area and acceptable head with the .86" ID.

Are you at all concerned about freezing air temperatures in the winter? It is getting into the single digits a few nights a year here in northern virginia (not sure what part you are in) so if there is a part of your piping that is exposed to air that should be a concern.

All of the lines leading out are going to be R4 insulated into the house. Keeping the line out of pit will be especially importance in the summer. In any case, the temperature probes will eventually be programmed to alert me at a certain threshold.

My water pipes are buried at 18 inches in some sections and the pex isn't insulated under the house. I'm used to paying attention to those low or sub 20s nights. There should be enough left over R4 insulation tubes to cover all of my accessible pex piping. Unfortunately, the iron ground pipes will still freeze over without running some water on those really cold nights.

[scook9]

This will be an awesome project, I am looking forward to some results and temps. I doubt you will be able to overload this system haha

[Vinas]

You might want to look into a 80mm or 90mm cooling fan for that iwaki. They run hot, and IMHO putting a fan on them prolongs the life. The one I have has been running 24/7 for almost 3 years now and runs A LOT COOLER with the 80mm fan + shroud I made for it. Before then I was getting some heat dump back in the loop... The only reason I put the fan on is because once it's running for 6 hours or so it will get pretty damn hot to the touch!!!

Looking good, I'll be following your progress! How deep did that trench end up being? Did you use the caterpillar?

[Mabyboi]

looks amazing!and XTREME!

[Romir]

You might want to look into a 80mm or 90mm cooling fan for that iwaki. They run hot, and IMHO putting a fan on them prolongs the life. The one I have has been running 24/7 for almost 3 years now and runs A LOT COOLER with the 80mm fan + shroud I made for it. Before then I was getting some heat dump back in the loop... The only reason I put the fan on is because once it's running for 6 hours or so it will get pretty damn hot to the touch!!!

Looking good, I'll be following your progress! How deep did that trench end up being? Did you use the caterpillar?

I'll be sure to do that when I build the noise dampened and ventilated box for it. I have a PA160 I've thought about blowing the chilled air off of onto it even. That seemed a bit unnecessary though. I've read about that too hot to touch being the norm but like you're saying, cooling won't hurt.

Originally I wanted to keep the pump outdoors below the computer area, but it's only supposed to be operated in 32-104 ambients.

That ditch witch dug trench was for the return lines up and down the hill to the wet ground where the real trench will be. I'm told that getting it to 10-12 feet still shouldn't be a problem in the muck down there.

[Romir]

This will be an awesome project, I am looking forward to some results and temps. I doubt you will be able to overload this system haha

Right now the ground temperature at the planned depth is going to around 50f, so it will be appear wonderful right off the bat. It'll be a while before we see the most important temperatures later this summer and early this fall when the air and then ground temperatures peak. In this deep, wet soil, it'll keep its cool.

Even with conservative estimaes, the 1500w heat element test I'm going to do shouldn't saturate the loop's cooling capacity. Two will though.

[rge]

Would be interesting to learn about its cooling capacity...could you post the numbers when you do the testing...love this project.

[Romir]

Would be interesting to learn about its cooling capacity...could you post the numbers when you do the testing...love this project.

Will do. I'll be following what AC_Hacker discovered.

http://ecorenovator.org/forum/projec....html#post2787

[scook9]

Wonder what this will do for the resale value of the house......"comes watercooling ready!"

lol

[Blauhung]

Will do. I'll be following what AC_Hacker discovered.

http://ecorenovator.org/forum/projec....html#post2787

This project and the stuff that guy is doing are awesome. I need to start accumulating some tools.

[scook9]

like a giant backhoe? lol

[Romir]

Daily update:

It was mostly overcast all day so the piping was extremely uncooperative in forming coils when I got home. Its going to rain and snow tomorrow, followed by a high of 41 on Thursday with 18mph winds. The coiling will have to wait until this weekend when the wind calms down and the sun will be out.

With nothing else to do, I shimmied around the crawl space to measure the runs length. At 35 feet or so each way, its a bit longer than expected. There won't be enough left over R4 pipe insulation, so I'll have to buy some 3/8th wall R1.9 stuff. I'll use it in the "vertical" run into the middle of the house. The better insulation should go on the "horizontal" run along the side of the house where there will be more wind. I never found strong enough siding to seal off this hill side crawl space. The heavy winds here knocked a tree onto the previous incarnation of this house. I might as well get on with placing cinder blocks and reap the energy savings.

Finishing the crawl space plumbing is now my top priority. I picked up three 1/2" compact valves and four 3/4" ones from harbor freight tonight. With those in hand and the measurements recorded, I'll be able figure out everything else I need to buy at Lowe's tomorrow.

The in-line temperature sensors will somewhat hold up 100%ing the in-house sections of the loop. They're going to be mounted in 1/4" copper tubes, then inserted into 1/4" compression fittings, which will screw into a T. Actually, that's a horribly expensive way to do it fitting wise. It makes securing the water temperature sensors on the inside of some insulated tubing much more tempting!

Finishing the house section of the loop this week and then doing the coils this weekend seems like a realistic plan. I'm really looking forward to testing the flow of dual ddc 3.2s pushing the complete loop (minus the future pits depth of course).