I am planning to pump water to the roof of my house in order to cool a solar panel array during warm days. I'm trying to determine the best way to route the water as the water source is on one side of the house and the array on the other.

Is it more efficient to minimize the height that the water will move, but introduce some 90° bends in the plumbing, or minimize the bends in the plumbing at the cost of an increase in height? What is the general trade-off in pressure for these scenarios?

To illustrate, I can pump the water straight up, over the apex of my roof and down the other side. Assuming the pump can handle the extra elevation of about 3-4m, gravity will assist in moving the water back down the opposite slope of the roof.

Alternatively, I can bury a length of pipe going from one side of the house to the other, introducing some 90° elbows, then pump straight up to the desired elevation.


A rough estimate shows that I would need to raise water at-most 10m/33'. I don't need a high flow rate since I plan to use micro-irrigation drip sprayers; these things will require at most 1 gallon per hour in flow, but typically much less.

  • 3
    Why do you want to do this? The solar arrays don't give a dang about the temperature (hot or subzero). OTOH, there are commercially available "heatpump" systems which circulate fluid so as to pass the excess heat to a heat exchanger or water heater. Is that what you're after? Sep 28 '16 at 14:21
  • 1
    My solar panels generally lose 0.5% in output per degree Celsius above 25°C. On a typical sunny day, I am losing 10-20% of my system's max rating due to air and roof temperature. Since my state government pays me to produce electricity at about $0.50/kWh, I would be foolish to not capitalize on a low-cost cooling system. Sep 28 '16 at 14:59
  • 2
    I stand corrected on the existence, but am not convinced the panels reach the temperature you report. There are fixes, such as designing a system to run at higher voltages (too late in your case :-( ) . So, time to go back and get an accurate measurement of production loss (perhaps a temporary cooling system on half the array) and of the power required for either a fluid loop or some local cooling fans. Obviously, if the cooling system uses more power than the added generation, it's a loss :-) Sep 28 '16 at 15:10
  • 5
    I'm not clear on the plan, but if your intent is to spray water on the surface of the solar panels you may create another problem. Water contains impurities such as calcium carbonate which get left behind as the water evaporates. These will build up over time and will be hard to remove. This itself may ultimately have an impact on efficiency.
    – Mark
    Sep 28 '16 at 18:02
  • 4
    I live in Oregon and it gets warm enough here that cooling the panels can increase your output, I helped a friend install a 5kw system he wanted to cool his array, I did warn him about water scale on the glass but he did it to find out running a mister spray enough for peak performance ended up increasing his water bill so enough to offset his savings. Next that he put a small pool behind his garrage and a tiny pump running off the inverter and found there was a tiny savings but it added two much maintenance, he was using a micro controller to cycle the water based on temperature. No water now
    – Ed Beal
    Jul 20 '18 at 13:40

Head is a term used to identify the vertical distance between the pump output and terminus. This is where we figure out how hard the pump has to fight gravity. Another thing to remember is the rule of thumb is every 90 degree elbow is equal to 20 feet or 6m of pipe. How much water you want to move and your head distance will determine pump and discharge pipe diameter. So we need to know head distance, volume in lpm or m^3/hr to get an idea of where to start. And FYI, even though gravity will help the water fall, you will need a bigger motor to start against the head pressure meaning a bigger pump than you really need. Go horizontal.

  • Does the rule of thumb for 90° elbows imply added friction or elevation for 6m of pipe? Some of the elbows will route water vertically; I can't avoid that. But if I route horizontally, many of the elbows will also be routing horizontally too. Sep 27 '16 at 21:03
  • 1
    The rule for elbows is about friction loss. Is this pipe going to be open ended? You have not adequately described the scenario, to properly calculate you need to state how much water you need at the end of the run. Example, if I have a sprayer that I need to flow 10gpm for effective spray pattern, I calculate head + friction loss for pipe and fittings to determine pump discharge requirement. Note, get a bigger (centrifugal) pump than needed and throttle it back as necessary. Sep 28 '16 at 4:42
  • 4
    I suspect that "rule of thumb" only applies to reasonably fast flow rates. If the OP's just cooling things, his flow rate probably is slow enough that bend loss is far less than that. Sep 28 '16 at 14:22
  • I've edited the question above to include a rough estimate on height and flow rate: ~33' at at most 1 gallon per hour. The flow rate will likely be less since I'm planning to use micro/drip-irrigation sprayers. Sep 28 '16 at 15:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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