Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Water cooling system help! Lots of things! Calculations and everything!

  1. Aug 22, 2012 #1
    I live in Dubai, and the water for my whole house is stored in a water tank on my roof. This is a problem as during the summer months the temperatures reach up to and a bit beyond 50 degrees in the day, this heats up the water in the tank alot so we don't have any cold water during the later part of the day. I wanting to implicate a passive water cooling system on my house where a pipe comes out of the tank and into the ground (I live in a bungalow so it's not too much of a problem) and down into the ground where it's cooler. I'd then run this pipe around underground which will cool down the water, then bring it back up into the house. I am desperately looking for help though. I want to investigate all of the variables and try to come up with best system for my house. I'll need to investigate things such as thermal conductivity of the soil, the equations for the pump that I want and what the best pressure/water flow would be as I'm not too familiar with water systems like this. I'd love detail and how I'd carry out the experiments, its a really open project and I'd like to measure/calculate everything. All the help would be useful.Thanks alot!
  2. jcsd
  3. Aug 26, 2012 #2
    I think what you are looking for that's exactly a CHILLER system, (like used in central airconditioning). It's not a big deal. Go and search for a water chiller system you will automatically foigure out it. But as the water gets heat up, you can also use vapor absorption cycle. It will be well suited for your system.
  4. Aug 27, 2012 #3
    Thanks very much for the reply! You're right that it is similar to a Chiller system, however, from what I've researched, chiller systems use an actual chiller and compression tank etc. and use this cooled water to cool the air etc. However, with mine I don't intend to use the chiller etc. Just a simple pump and piping down from the water tank, into the ground and back up into the house to cool the water. Some of the physics behind this chiller system would be very useful to me though? Like what you have to measure to calculate the temperature change between the two identities (the cool ground and the water in the pipes), how to calculate how long the pipe needs to be in the ground for and at what water flow rate? All these would be verrry useful if possible? Thanks a lot again!
  5. Aug 27, 2012 #4
    Why not just insulate the tank?
    Or if you are installing a pump, relocate the whole tank underground.
    Last edited: Aug 27, 2012
  6. Aug 27, 2012 #5
    For the same reason that people using passive heating (in the opposite way to me e.g. in England) don't just insulate the pipes etc that their water come from, the water coming in is already hot, and insulation would still let the water heat up anyways since it is left in the open for the whole day. And because it was to my belief that installing a tank underground would require frequent checks to ensure nothing was seeping in or out.
  7. Aug 27, 2012 #6
    Relocating the tank sounds like a no go then.

    Insulation would mean it heats up, but even a small thickness of insulating foam can dramatically reduce the heat transfer whilst it's sat in the tank. Reducing your heat transfer requirements for cooling, if you subsequently use any active system.

    Plus insulation is a very, very cheap way of improving the situation a little bit. I believe it's where you'll gain the most °C/cost benefit.


    Obviously is the water is pumped in at a similar temperature to the surroundings insulation is pointless. But even if it's pumped in at a lower temperature, you'll gain a large benefit from reducing heat transfer into the tank.

    As you are using the ground as a heat exchanger, a certain length of pipe will be needed to get to a target water temperature, depending on starting temp. It may be worth costing up, the difference between installing an extra length of pipe vs a bit of insulation.
  8. Aug 27, 2012 #7

    jack action

    User Avatar
    Science Advisor
    Gold Member

    I don't think water flow is a variable here. What you need is a «cooling tank» in the ground. I say «tank», but really it would be a pipe to dissipate the heat faster.

    I would first ask myself what is my daily need for cold water. Once you know that, it is easy to calculate the size of your «tank» (piping) that is in the ground. This way, you know the water you use has at least spent 24 hours in your underground «cooling tank».

    For example, using numbers from this site, I roughly calculate that an average household needs 60 L of cold water. (Of course, in your case, you don't need to cool the water that is used for the outdoor, the clothes washer, dishwasher or toilet).

    Using a 3/4" (19mm) pipe diameter, you need around 175' (53m) of pipe into the ground to store 60 L of water.

    Just put your numbers to adjust to your needs if they are different.
  9. Aug 27, 2012 #8
    As chiller system will be lot's of complication and costly also, so go for simple cooling tower which will transfer the extra heat to atmosphere by using latent heat of evaporation of water. Anyway I don't know what is the ambient temperature and humidity of your surroundings and how much cooling do you need. Water tower simple in construction, good aesthetic if made properly, and cheap also.

    But anyway I am getting how you transfer the heat to ground form piping. With what do you exchange the heat. Don,t you think it will trap the heat and after sometime it will stop working.

  10. Aug 30, 2012 #9


    User Avatar
    Gold Member

    I do not know why you say that. Underground cisterns have been in use for centuries, way back to Roman times and before. You could use concrete or a plastic tank and fill the unit at night when the water temperature may be lower. The water in the tank should stay at a constant temperature similar to that of the soil temperature and will not fluctuate to the extremes as your tank on the roof ( which receieves solar irradiation in addition to the heat transfer from the warmer air ). Depending on your setup the water could be classified as potable ( meaning drinkable ).

    One site

    You could add a pump to withdraw the water from the cistern.

    An underground cistern may be something that you might find interesting to suit your needs.
    Of course, the cost of this method and the others addressed you would have to consider.
  11. Sep 1, 2012 #10
    you should make your tank cubic with padding so that there can be an even volume and it will be harder to heat up and then paint the tank white so it can reflect any photons that hit without absorbing them it or if you have a roof over it paint the roof white. Then you won't need that expensive of a cooling system because your water won't be getting as hot as it is now.
  12. Sep 4, 2012 #11
    Just an FYI Anaximenes, the sphere has the smallest surface area by volume for any given shape. If you want to maximally reduce heating spheres are the way to go.

    With that said, if you want to maximize cooling, then a sphere isn't the way to go. Though, considering we are talking about the flow of water, neither are polygons.

    I agree with 256 bits, a concrete "sump" buried will help out a lot. How you get the water out depends on where/how the water enters the sump, how deep it is, and where the water has to go. The size of your pump (which shouldn't be very large at all for most domestic needs) will depend on these previous conditions, as well as what devices are being fed, and whether or not you need to send the water through some sort of softener/filtration system for treatment and/or potability.
  13. Oct 25, 2012 #12
    Any possibility of using a cooling tower like the ones they use at power plants to cool the process water? A cooling tower makes use of water evaporation to bring about the cooling. All this could be done outside, without pumping under the ground. You pump water to the top of the tower, and run it down a very large surface. You blow air over the surface with a fan. The water that doesn't evaporate gets cooled. Do they make small scale cooling towers for home use?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook