I'm planning an installation for heating water using solar "exchanger" panels (solar used to heat water directly, not through electricity). I don't want to bind the reservoir, panel and tap locations to the natural circulation cycle though (hot water traveling up etc).

So, in order for this to work, I'd have to force some very slow water circulation between the reservoir and the panels; a pump that takes very little power and provides very small but constant throughput: something with power requirements low enough to make this economically viable, maintenance-proof and preferably silent. On the other hand, the throughput can be very low (the water in the panel must be given enough time to heat up, so I guess even something like 10cc/min is satisfactory) and it will work only with clean, drinking quality water and very low pressures (just enough to have the water moving through pipes, no resistances or height pressure to overcome).

Do you have any idea what pumping mechanism could I use for that?

  • 1
    This is quite a mature market, so copying whatever the commercial products do is probably your best bet.
    – 410 gone
    Feb 6, 2013 at 12:15
  • 2
    Are you sure you should use water as the medium? No freezing all the year?
    – Peter Ivan
    Feb 14, 2013 at 7:59
  • @PeterIvan: The purpose is utility hot water, for bathing, washing etc. working in parallel with electric tank water heater, so definitely water. On hot days the water would run hotter than what the heater hysteresis level, saving electricity. During unexpected frost water heated by the heater would slowly circulate through the panel keeping it from freezing. Normally, we'd just empty the panel for winter.
    – SF.
    Feb 14, 2013 at 8:28
  • 3
    I'd strongly discourage such the scenario, it's just opposite of sustainable. During night you're going to lose more heat into atmosphere than you collect during the day. And the cost of electricity spent to heat the outside air is just wasting.
    – Peter Ivan
    Feb 14, 2013 at 16:27
  • @PeterIvan: I can stop the circulation for the night (save for situations of unexpected frost).
    – SF.
    Mar 8, 2014 at 3:11

4 Answers 4


Thermo-siphon works surprisingly well when feasible. If it is feasible to move the physical location of the reservoir placing it above the collector(s) will cause the water to circulate on it's own. Placing simple diaphragm valves in the reservoir will make the water circulate only when the panel temp is higher than the reservoir. If you can achieve precise control of the water level you can do the same with relative positioning of the inlet and outlet pipes using ambient air pressure as your valve.

This picture shows the most simplistic thermosiphon design. The design challenges in thermosiphon systems are minimizing temperature change in the transfer tubing between collector and reservoir, structural support for the elevated reservoir (water is heavy). If direct heating is used preventing freeze damage to the collector is a design concern, if indirect heating is used preventing conductive cooling during non-collection times should be addressed during the design phase.

Note: I'm not recommending any product from any commercial site hyper-linked above, just directing readers to the educational material, especially pictures/diagrams, on those sites.

  • Can you provide more info? The reservoir probably will be placed in the attic and filled to given level using the standard WC mechanism.
    – SF.
    Feb 20, 2013 at 23:53
  • 1
    Of course! Links and a little more info added. Thermosiphon systems are widely used. In the west prosperity has just allowed us ignorance. Here's a typical Israeli roof top: sunnergy.cn/uploadfile/newfile/01.jpg As you can see in this picture from a hilltop in Tel-Aviv such systems are common there: 1.bp.blogspot.com/_B51On_cmJmM/SNUT6-TK4HI/AAAAAAAAAIw/…
    – OCDtech
    Feb 21, 2013 at 14:40
  • A cheap sprinkler timer can be used to stop reverse flow, and corresponding heat loss, but is a no-no in sub-freezing temperatures (an occasional night at >25'F is ok), because the reverse flow is actually what keeps the collector from freezing during those circumstances. Thermo-siphon works well both open loop and closed.
    – OCDtech
    Mar 7, 2013 at 15:45

As soon as you have moving parts, you can no longer have it maintenance-free. Moving parts wear out eventually. On the height pressure issue your panels will probably be on your roof and your reservoir somewhat lower. We can probably disregard the height pressure for operational reasons but if you ever get air in the system you will want to be able to overcome that pressure. There are a few ways to do this (including of course using the height pressure from the municipal water supply to beat the heigh pressure of the house).

The simplest solution is a small electric motor to drive the water pump. I suppose if you want it off-grid, you could hook that up to a PV cell as after all you only need it running when the sun is out. Water pumps are pretty simple devices and there are only a few types out there. If you break apart a commercial model or two you should get a pretty good idea of how they work. The only hard part is keeping the motor sealed so water doesn't get in and short things out.

  • I guess moving parts = not maintenance-free, and electric motors are quite power-hungry. I heard about something like ion pumps... low power, low throughput, no moving parts, no clue about the rest.
    – SF.
    Feb 14, 2013 at 18:02
  • @SF. Do you mean ion pump as a low tech? You won't lose with today's standard commercial products.
    – Peter Ivan
    Feb 17, 2013 at 19:46
  • @Peter: From what I heard they are essentially extremely simple devices. Sure there are fancy materials and high voltages involved...
    – SF.
    Feb 20, 2013 at 15:18
  • 1
    That is true, an ionic pump has no moving parts, but must ionize the water. Ionizing has a sterilizing effect, which is good, but also tends to promote corrosion of plumbing components. The electrical components needed to step up to the extremely high voltages tend to waste a lot of energy as heat. You're also going to have a very hard time generating enough force to overcome more than a few inches of head without expending FAR more energy than you save from solar water heating. Chris is right, a cheap commercial pump will serve you better.
    – OCDtech
    Mar 7, 2013 at 15:52

Use a common (commercially available) electric pump. Home made ones can be problematic, as Chris writes.

The solar pumps are available in various power. Even with a very small pump you'd want to install a controller with (at least) two temperature probes on the panel and in the tank to turn the pump on and off.
I'm using a 45W pump with throughput of 2 litres/minute on a panel with area of 1.2 m2. During sunny summer it runs almost all the day and it's stopped by the controller if the heat gain is too low.
The pump installed in a polystyrene box to silence it and it works well.

You'll always have a high pressure (1 bar) in a closed system with varying temperatures. The panel can easily reach 100°C if the pump stops and the pressure will rise. An expansion tank is a must and water is not to go usable because it would boil out.


Another possibility would be a bubble pump system or Geyser Pump. It's very simple, passive and innovative - but you'll have to find some references to existing installations. There is a comparison of solar hot water systems to consider your requirements.
The system is underpressurized and the medium boils when heated in the panel easily. While bubbles move to a (some) higher place, they move the loop of medium toward the water tank. After doing their work, the bubbles condensate back and preheat the returning medium.

  • I'd prefer a slow sustained flow, and skipping any controllers, sensors etc. As low-tech as possible. I'd choose convectional circulation if I wasn't afraid of boiling or freezing.
    – SF.
    Feb 14, 2013 at 11:30
  • 1
    What do you mean "and water is not to go."? Jun 29, 2013 at 5:53
  • @HighlyIrregular: I edited the answer to clarify that water would boil out and that's why it's not usable.
    – Peter Ivan
    Jul 7, 2013 at 19:19

I have the same problem and I think you should build your system as like double jackrt thermosephon system and that's I will do know I hope it's work rite choice

  • 3
    Welcome to Sustainable Living! Could you explain this a bit more? What exactly do you mean with a double jack thermosiphon system?
    – THelper
    Jan 30, 2014 at 15:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.