Concrete Slab water cooler loop - Hooked up!!!

Thanks pugs

Ive got hold of one of these pumps and will have it this arvo so look out for some results this weekend I wanted a better quality pump but oh well Im sick of waiting.

MAXI 106 3200L/H WATER PUMP

AQUA ONE

- Suitable for both Fresh and Saltwater Aquariums.
- Completely Submersible.
- Quiet and Efficient.
- 3200L/Hr.
- 80Watt.
- Max.Head Height 3.5 meter.
- Outlet size:19 mm + 21 mm.
- Dimensions:14H x 10W x 17.5L cm
- Warranty: 3 years

 

I agree with pugs, i've been following this thread from the start.
Who cares what the bloke does with his house?
To be honest, its an awesome idea.
Gigs is taking extreme PC cooling to a level most of us havent even thought about, and for those of you that had thought of it, he's doing what you didn't do.
If it does fail, who cares? It's a bloody cool experiment.

 

i would use a different brand i have a few fish tanks and i have stoped using Aquaone due to been noisy and they some times just stop working for no reason.

the Enhim would be a lot better choice

 

yep agree aqua one are elcheapo chinese I will get a better pump down the track

 

This is much the same as burying a jerry can 6 ft under and running copper pipe to it - which for the record works incredibly well.

Looks good!

 

Hmm, don't have my heat transfer text handy, but from memory, I used a value for concrete of 1.37 W/m.K, although that is quite likely to be for a different concrete mix (with high aggregate content or something).

The big difference between water & concrete, though, is convection.

The heat carried away by convection in the water will far exceed the conduction through the water itself (i.e. you're replacing a warmed-up bit of water with a cold bit, and keeping a high thermal gradient, thus dramatically increasing the heat transfer from pipe to fluid).

My rough calc above suggested 20 degrees between pipe & concrete, so you might be looking at 35 deg for the fluid rather than 25, but otherwise it should work just fine. That was for a steady-state calc, so you'd be able to pump 150w into the slab at that temp for a loooooong time. The specific heat of concrete is something like 750J/kg.K, so putting 150J/s into it will take a long time to significantly increase the temperature of half a tonne of concrete, even ignoring losses to the ground beneath!

Of course, if someone cared to do a less rough-and-ready calc (i.e. take into account temperature diff across the pipe wall, temp gradient along the pipe, temp gradient in the concrete, transfer to the soil beneath, etc etc), they might find I'm out by a bit...

I'll be interested to see what your test results are - if for no other reason than to see how close my rough calc was!

Hey, I did!

 

Thanks Walshy, Im not great at maths, Im more of a jump in head first kind of a guy as lots of people have pointed out

Ive got a pump and pipe sorted so any ideas for a good way to run some tests over the weekend? I also have a 150 watt fish tank heater and the wifes kettle @ 2400 watts

 

I got a chance to run a quick test tonight. I will post pics and run some more test tomorrow with a more realistic heat source (150 watt fish tank heater).

The experiment:

Pumped tap water through the system with kettle as the reserve. So in total it was about 2 to 3 litres water volume. Let it run for about 10 minutes when the temp settled to the same as the slab which was 10 deg.

Turned the jug on (2200 watts) and the water temp increased up to about 45 deg in 1hour. The next 30 minutes the temp increased only by another 5 deg to 50 deg. So I can see that with 2200 watt load:

.5 hour interval 25 deg increase (water @ 35 deg in kettle)
1 hour interval 10 deg temp increase (water @ 45 deg in kettle)
1.5 hour interval 5 deg increase (water @ 50 deg in kettle)

The return water sat at about 3-5 deg lower than the water in the kettle.

I stopped the test at 50 deg because the pipe was starting to kink as it warmed up and I didnt want to run to much hotter water through the pump.

I switched the kettle off and literaly watched the degrees drop on thermometer.

After 10 minutes the temp had halved to 25 deg and by 15 minutes was down to 15 deg.

The numbers look good to me so far. Tomorrow I will run a much longer test with a 150 watt heat load and see how it pans out over 5-6hrs (my typical time spent with the computer on).

 

Looks like you will get very similar results to JSE's Earth Cooler GiGs excellent stuff buddy... Did you plan on hooking multiple pcs up ort was it just the one? appologies if it was mentioned previously i don't recall...?

KiM

 

i think its a waste of money with no guarantee it will work or guarantee it will make much difference in temps.

also you might regret having a copper pipe running along ur foundation later on in life

 

Gigs isnt looking for a guarantee that it will work, he is just trying something that i spose alot of people have thought about at one stage or another during watercooling.

Big thumbs up to Gigs progress and results so far.

I would say going by your quick test, 150w should be no issue at all.

 

here are the results for today:

• test start time 9am
  
• Water temp at start 10 deg (after 10min of running the pump with no heat load).
  
• Water volume in the revserve 2litres, water volume in the loop .5 to 1 litre. Total volume 2.5 to 3 litres.
  
• Heat load 150 watt fish tank heater. This will simulate very closely the heat load I will be putting on the system as i7 creates about 130 watts of heat.

The test system is still running but I don't think it will change from 17 deg that it has been running for the last 4 hours+


Click to view full size!


So what does this test tell me / what can I expect?


Firstly, Tasmanian agricultural areas (soil) have a mean annual temperature of 11-13 deg C. So I can expect the concrete foundation to sit at about 14-16 deg year round. When the house is built adding a few deg for ambient heat in the house. I have shown that this system will hold 150 watt load at about +7 deg above concrete temperature.

So I can expect water for my PC cooling loop to be about 21-23 deg C year round. I was hoping for a bit lower in the range of 16 deg and I'm sure this would be possible with a longer loop.

My current water loop (standard setup with 120.3 radiator and CPU block) sits at about 10 deg above ambient air temp giving me 29-31 deg C coolant temp.

After all those variables are taken into account. I can expect to have chilled water 8-10 deg C below my current water loop which should give me quite a lot more head room for overclocking and best of all NO NOISY FANS blasting air through a heat exchanger.

A big thanks to the people that posted constructive comments and offered help with this project


Click to view full size!


Click to view full size!


Click to view full size!

To the people that said this would fail:

http://www.youtube.com/watch?v=rX7wtNOkuHo

 

Very nice mate, looks fantastic.

i think the only thing that will stop your overclocking is the efficiency of the waterblock or the cpu itself

 

what were the atmospheric conditions that day? what was ambient temp, wind speeds etc.? as being exposed out on the slab all these things may have added extra heat transfer mechanisms into the balance which may not exist in the final rig.

just for a easy test you should try running the test loop not actually connected to the concrete slab cooler (just run water through the ~3m of plasic tubing (or whatever length was used in the test rig), the pump and the heater/reserve jug)

that will give you a rough idea of what impact the exposed parts are having on the cooling. which should allow a better estimate for final loop temps.

 

Like any of that will make a huge difference.

In the spirit of the plane on a treadmill question.
The plane will take off, vis a vis, the computer will stay cool.

Thats all that matters.

 

Tassie has 4 seasons in one day and thats what we had today

Was sunny and warm one minute, next minute it was raining and windy. During the test the air ambient temp varied a few deg C but didnt impact on the temp of the loop.

Thermal mass of concrete FTW!

 

EDIT:-...DOH....just scrolled up DAMN... ignore previous post...how the f**K i miss that :-S

ROFLMFAO @ YOUTUBE link to hehehye

GiGs we new this would be the case but it is good to see how well. it has panned out.. I would think you could put several pcs on the loop and still be way under your Rad performance...the only thing i would have changed as mentioned previously was the length of the loop, Something i WOULD be trying now though is slowing the waterflow down :: wink :: give it a shot see how that effects the temps....either way though HUGE Props for going through with it AND keeping your cool when drips continued to tell you how it wouldnt work, unlike 99% peeps here who repeatedly throw around the line "I'm definitely doing that myself" (or similar) never to be seen or heard from again....you stuck to your guns, Well done GiGs

Best of luck finishing the house construction to buddy i hope this is as trouble/hassle free as it can be...

KiM

 

Thank god for some completely non-anally retentive realism.

 

Its a bloody good idea. I do like the idea of houses having cooling points.

 

One step closer to being green