Heat Pipe/Pump

[telecomguy]

I have an underground cooling application and was thinking of driving a heatpipe into the ground (similar to an earthground rod) in order to sink the heat of my electronics to the earth cooled rod.

It seems like this would work since homes are now cooled with heatpumps in a similar fashion. But I would like it to be simply driven into the ground. Does anyone know where I could find this off-the-shelf? or will I have to go work the design from scratch?

TIA,
<tg>

 

[IRstuff]

What's the thermal conductivity of the ground?

TTFN

FAQ731-376

 

[telecomguy]

The location is going to mostly sand so this site indicates about .44 BTU/hr-ft-F.

http://www.geo4va.vt.edu/A1/A1.htm#A1Sec3b

I'm looking at dissipating 50W with a max ambient in the electronics enclosure of 65C. Looks like I may need a big hammer...

<tg>

 

[telecomguy]

Maybe I jumped the gun, or else I haven't had enough coffee yet.

.44 BTU/hr-ft-F x 1.8 F/C * .293 W-hr/BTU ==> .232 W/ft-C

Assuming surface temperature gets to 120F (49C) in Las Vegas, and looking at the sub-surface gradient table on that site, I'm guessing soil temperature is in the mid-30C range about 10ft down. I'm looking for L.V. specific data, but for my napkin calcs I'll use that.

So I'm looking at about a 30C delta, or about 7 W/ft (30C * .232 W/ft-C). That means I need about 7ft in that sub-10ft soil to dissipate my 50W (which I now find out is more like 30W). I know there are alot of assumptions here, but it's better than the number I came up with on my first attempt.

Where am I going wrong?
<tg>

 

[Compositepro]

A heatpipe will move heat upwards or horizontally. The liquid moves by gravity or capillary action. If you need to move heat down very far you will need a pump. Heatpipes are used on the Alaska pipeline to keep the soil frozen around the foundations. Heat moves up in winter but does not move down in summer.

 

[IRstuff]

Heatpipes can be designed to go against gravity.

The issue, to me, is how the ground is going to absorb the heat that you plan on dumping, given that dirt is typically considered to be an insulator.


TTFN

FAQ731-376

 

[Compositepro]

Capillary action has a pretty limited height that it can raise a fluid against gravity.

 

[chicopee]

What an intersting question and I was thinking could thermopiles be used in this application .
By the way, homes are not cooled with heat pumps, homes are heated with heat pumps.

 

[telecomguy]

It is my understanding a heat pump is capable of extracting cooler temperatures from below and cool homes.

http://www.energysavers.gov/your_home/space_heating_cooling/index.cfm/mytopic=12640

I'm thinking in general terms, a copper rod sunk into the ground is going to be essentially at a constant temperature (cooler temperature). If the heat source is bonded to this cool rod, it will transfer the heat down the rod and be dissipated into the earth. I guess I was thinking a heat pipe could be a more efficient transfer device, but I understand the concern about a traditional heat pipe (capillary action/gravity). Maybe that was misleading.

Maybe it's time for an experiment with a styrofoam cooler and a resistor board with heatsink bonded to the ground rod under my electric meter. Could be a total bust.

Thanks,
<tg>

 

[telecomguy]

Found this study. Looks like there's a solution here...

http://linkinghub.elsevier.com/retrieve/pii/S0360544201000366

 

[IRstuff]

You apparently choose to ignore the SIZE of the exhanger loop. Your last citation stipulates a coolant loop of 67-m length, not a "rod" sunk into the ground.

Likewise, your first citation shows an example cooling system with 400-ft depth thermal wells.

http://www.energysavers.gov/your_home/space_heating_cooling/index.cfm/mytopic=12650

A copper rod sunk into the ground will NOT stay at constant temperature, if there's a heat flow. Your quoted soil thermal conductivity is comparable to styrofoam insulation. You might get lucky if your ground is wet, which would substantially improve the thermal performance

TTFN

FAQ731-376

 

[telecomguy]

30W-50W vs cooling a house is slightly different. But I agree, this may be just a pipe dream (pun intended)...

<tg>

 

[chicopee]

telecomguy, a heat pump extracts temperatures from under ground (wells normally) which are warmer than the low pressure side refrigerant allowing the compressor to raise the refrigerant pressure and temperature (high pressure side of the refrigeration cycle) which is used to warm up your house instead of rejecting that heat to the outside which an air conditioner would do. The heat to your house is created by the compressor and is normally refered as the heat of compression.

 

[telecomguy]

I'm obviously twisting terminology, though I still read this to mean a heat pump is capable of cooling your home... "As with any heat pump, geothermal and water-source heat pumps are able to heat, cool, and, if so equipped, supply the house with hot water. "

I'm looking for a passive solution to cool a 30W-50W device that is currently in the ground inside a plastic enclosure (ie. low thermal conductivity to the earth around it; think plastic water meter box). I prefer not to dig it up and replace the enclosure with a conductive enclosure. So I'm thinking I need to remove the heat via a heat spreader. The question is, where do I go with the heat? I could pipe it up to the surface, but now I've got a protrusion above ground which is undesirable or a heat spreader that will get covered up by dust/dirt or damaged. So my next thought was to sink to the earth. I'm not trying to cool a 3000 sq. ft. home, just a little electronics box, so I don't believe it should require a 67m cooling line. But since I want this to be passive, a cooling line with a pump is not an option and a heatpipe is the wrong answer based on the physics, so I may be stuck with the solid copper rod as the leading candidate.

Now the question is, what is the efficiency of the rod driven into the ground and attached to the thermal load. I just need to suck the heat out of the enclosure. If the earth can accommodate me, and help out a little, I'll be most appreciative. Not looking to defy any of Newton's laws, just trying to please the customer.

<tg>

 

[davefitz]

Without putting too much time into it, it is common and easy to drive a 10' long x 3/4" copper rod for home lighting grounding in to the soil.( $10 at home depot + 1 hour labor). It would seem to be simpler to just try it than to calculate it.

Heat pipes normally are configured to have the hotter body at a lower elevation than the cooler body, to enable the use of natural convection to drive the fluid motion, but in you application this is not practical. One can also use wicking action in zero g ( space ) applications, but I am not sure the wicking action can overcome gravity in your application.So a heat pipe per se is out of the question, and a heat pump is way too expensive for a 50 W load.

If you absolutely need to calculate it rather than simply do it and see what happens, the gen Y method is model it on a finite element program and let it rip. The older group might calculate it as a pin fin conductor inserted into an infinte sink , substituting (1/contact resistance) for convective heat transfer coeficient .

 

[Compositepro]

Grounding rods are copper plated steel.