Build a simple solar water heater



I’ve seen a few different designs for solar water heaters and I wanted to share my own. It is quite an efficient design since every square inch of collector surface is in direct thermal contact with the water being heated. You can easily modify the design to any size you like. I made mine 8ft long by 22″ wide so that it can fit between the rafters in my attic. Tests showed that system output averaged about 530 Watts, heating 20 litres of water from 24 degrees C (75 degrees F) to 47 degrees C (117 degrees F) in one hour.

Aside: I’m in the middle of re-roofing my house and plan to build in a transparent section of roof in one area. Then I can experiment with different solar collector designs like this one and install and remove them easily from inside my attic instead of having to go out on my roof. It will make the plumbing easier too. The drawback is that if a collector springs a leak, it will leak into my home instead of into my gutter. See Build a solar attic for details.

Ads by Google

Warning – Don’t drink the water

It is not my intent for this design to be used for heating drinking water. The plastics and glues used will leach into the water so it’s a bad idea to drink any water that has been inside the panel. If you want to use this design for heating drinking water you should make a heat exchanger. Run the water from the collector through a coil of copper pipe placed in a tank suitable for drinking water. This collector design is also not meant to withstand city water pressure, but if you use a heat exchanger and an appropriate tank (such as a commercial hot water tank) you can use a collector like this to heat drinking water at city water pressure.


The collector is made from Coroplast (see which is a corrugated plastic sheet, commonly used for making signs. It has multiple square channels running lengthwise from end to end. When I first saw this type of sheet I immediately thought, “Wow, this would make an excellent flat panel solar collector if only there was a way to pipe water through all those little channels.” Several weeks later, a method of doing so occurred to me. If a slot of the right width is cut lengthwise in some ABS pipe (so the cross section looks like a “C”) then this pipe can be fit over the end of the corrugated plastic. The seams can be sealed to make everything water tight. The sheet can be painted black and voila… you have a flat panel solar collector.Because the whole collector is made of plastic, it is important that the temperature doesn’t get too high or it will soften and possibly spring a leak. 80 degrees C (176 degrees F) is about the limit. Don’t think it can get that hot? Think again. In practice the maximum temperature is difficult to guarantee. Water may stop circulating, or may drain out completely for a number of reasons and the panel will overheat. Therefore this may not be a practical design for residential installation but it is an inexpensive, easily built experimental system that produces as much or more hot water than commercially available systems. Mine cost about $60 in materials (about $4.00 per square foot) and about 6 hours of construction time.

The Concept

Tools and Materials


  • Table Saw
  • Hand Saw
  • Drill press
  • Power drill
  • 3/4″ drill bit
  • 1″ hole saw
  • Exacto knife
  • Tape measure
  • Screw driver
  • Digital thermometer
  • Caulking gun for adhesive
  • Coarse round file

Materials for collector

  • 1 – sheet of Coroplast plastic sheet (4’x8’x4mm) cut to 22″x90″ – $8.50
  • 1 – 4′ of 1 1/4″ ABS tubing – $6 (Note: Do not use PVC as it softens at too low a temperature causing leaks.)
  • 4 – 1 1/4″ ABS caps – $10
  • 2 – threaded 1/2″ hose nipples – $1.00
  • 1 – cartridge of silicone adhesive/sealant suitable for plastic – $3.50 (Note: since original publication I’ve found Marine GOOP to be better)
  • 1 – can of flat black spray paint – $5.00

Materials for frame

  • 1 – 1/2″ sheet of plywood (4’x8′) cut to 24″x8′ – $8.00
  • 1 – 3/4″ sheet of polystyrene (2’x8′) cut to 22″x87.5″ – $2.50
  • 2 – 2×3 x 8′ – $8.00 used
  • 1 – at least 4’x10′ of transparent plastic sheet – $0 scrap
  • misc screws and staples

Materials for tank / water circulation

  • 1 – cooler (or other water tank, preferably insulated) – $20 but I already had one scrap
  • 1 – 15ft of 5/8″ garden hose – $5.50
  • 2 – 1/2″ hose clamps – $1.50

Total cost of materials = $59.50


Build the collector

  1. Use an exacto knife to cut the corrugated plastic sheet to 22″x90″. When cutting lengthwise, be sure to cut in a single channel for the whole length.
  2. Cut the ABS pipe into two lengths, each 20.25″ long. Check that when a cap is placed on either end, the total length is 22″. I picked this width so it would fit between the roof rafters in my attic.
  3. Drill a 3/4″ hole in the side of two of the ABS caps. This will be easier if you pre-drill with a smaller bit and gradually increase the size.
  4. Enlarge the holes with a coarse round file until you can just thread in the nipples. I did not have a tap of the right thread, so I planned to just glue the nipples in place.
  5. Drill a 3/4 diameter semicircular notch in the end of each ABS tube. This is easiest if you clamp them end to end in a vise. Alternatively you could drill this hole in the ABS tube before cutting it, and then just cut through the center of the hole to make the notches. These notches fit around the nipple end when the ABS caps are in place.
    solar_thermal_collector_step_3_b solar_thermal_collector_step_3_c
  6. Using a table saw with a fence, carefully rip a slot down the full length of each ABS tube. The resulting cross section should look like a “C”. The ABS tube tends to compress as you cut, so that when you are done, the slot will not be as wide as the width of your saw blade. Feed each tube through the saw a second time to clean up the cut for a consistent width.
    solar_thermal_collector_step_3_d solar_thermal_collector_step_3_f
  7. Repeat the slot cutting process with the ABS caps, keeping in mind what direction you want the nipples to be pointing when the panel is fully assembled.
  8. Do a dry fit, assembling the ABS tubes, caps, and hose nipples. You may need to carve a bit out of the notch to get the slot in the tube to line up with the slot in the cap.
  9. Repeat the dry fit on the end of the corrugated plastic sheet. Carve up the ABS as needed to get a nice fit everywhere.
  10. After everything fits nicely, repeat the assembly, applying adhesive to all mating surfaces before assembly, and applying a bead of adhesive to all seams after assembly.
  11. Repeat for the other end of the corrugated plastic.
  12. Allow to dry for at least 24 hours.
  13. After drying, cut the garden hose in half and clamp the cut ends to the nipples.
  14. Fill the panel with water (just connect the garden hose to a tap on your house) and check for leaks.
  15. If there are any leaks, drain the panel, dry the area around the leak thoroughly and seal with more adhesive, allowing another 24 hours to dry.
  16. If you are interested in calculating the efficiency of your collector later, you need to know its volume. This is a good time to drain it into a bucket and measure the volume (including the hoses). Mine contained 7.2 litres.
  17. Once all leaks have been sealed, paint the surface of the collector black.

Ads by Google

Build the frame

You could use the collector as is. Just lay it out in the sun and pump water through it. However, much more heat can be captured by building an insulated enclosure for it.

  1. Cut one 2×3 to two lengths of 22″ for the ends of the frame. Screw the other 2 2x3s into the ends to make a rectangular frame.
  2. Wrap the transparent plastic around this frame to make a transparent lid to fit over the collector. In my case this is for test purposes only, since I intend to eventually install the collector between roof rafters underneath transparent roofing material which will provide a ready made frame.
  3. Cut the plywood to 24″x8′.
  4. Cut the styrofoam sheet to 7’4″ by 3’9″ and place it centered on the plywood. This will be the insulation for the back of the panel.
  5. Test fit the collector and drill 3/4″ holes in the plywood where the hoses will stick through. Make one of these holes into a slot by drilling two 3/4″ holes side by side and cutting away the wood between them. This is to allow for thermal expansion of the collector. Plastics typically have a high coefficient of thermal expansion. If you restrict the panel from expanding, it may warp and cause a leak.
  6. Now stack the whole works together: First the plywood, then polystyrene, then the collector, then the transparent cover.
  7. Secure the transparent cover to the plywood back with several clamps (or you can screw it on, but initially you might want to be able to remove it easily for access to the collector).

Fill the panel

Filling the panel in such a way that you get all the air bubbles out is easier said than done unless you use a few simple tricks.

  1. Lift one end of the panel and rest it on a chair or other object (I used my fence). Rest the other end on a couple blocks of wood so that the bottom hose will have clearance from the ground (remember I eventually want to install this on the underside of a roof, between rafters, which is why I made the hoses connect through the back instead of the sides).
  2. Mount your storage tank higher than the panel and stick the top hose in it.
  3. Connect the bottom hose to a tap on your house and turn on the water gently.
  4. Watch as the panel fills. When water starts coming out of the top hose, let it continue and fill the tank.
  5. As the tank is filling, temporarily tilt the panel so the corner where the top nipple exits is the highest point. This forces any air in the system to move towards the exit nipple where it will be expelled.
  6. Once you stop seeing air coming out of the top hose, return the panel to its previous position.
  7. Turn off the tap. Introduce a kink in the bottom hose to keep the water from flowing out. Then remove the hose from the tap.
  8. Keep the bottom hose kinked, and the top hose under water in the tank. Raise the end of the bottom hose above the water level in the tank and release the kink. Slowly lower the end of the hose until water starts coming out, then plug it with your thumb and quickly stick the end under water in the tank creating a sealed system with as little air in it as possible.
  9. Orient the hoses so that the bottom hose draws water from the bottom of the tank and the top hose delivers water to the top of the tank. Whatever you do, be careful to always keep both hose ends under water or you will “break the seal” and introduce air into the system which will prevent circulation by thermo-siphoning.


If you have removed all the air and have a sealed system and there is enough sunlight hitting the panel, it should start thermo-siphoning almost instantly.

  1. Turn the panel towards the sun and raise or lower the top end of the panel to better aim it towards the sun. One end of the panel must be raised higher than the other in order for thermo-siphoning to work. The storage tank must also be kept higher than the top end of the panel.
  2. Feel the top hose where it exits the panel. It should be hot if your setup is thermo-siphoning. The bottom hose should still be cool. If this isn’t the case, it probably means you have a vapor lock (air bubbles) somewhere preventing the water from circulating. Connect the bottom hose to your tap again and repeat the filling process, attempting to remove all the air bubbles.
  3. Once thermo-siphoning starts, use a digital thermometer with probe to measure the water temperature. By sticking the temperature probe inside the ends of the hoses, you can measure the inlet and exit temperatures of the collector. It took me about a minute after filling before I had my thermometer set up. At that time the inlet temperature was 23 degrees C (basically the initial temperature of the water) and the exit temperature was 50.7 degrees C (123 degrees F).
  4. Measure the inlet temperature over a period of an hour or so (or till the temperature stabilizes). The inlet temperature should always be the lowest temperature in the system. Measuring here will give conservative results when calculating the amount of energy transfered to the water.


See the image below for a plot of temperature vs time.

Thermo-Siphon Flow Rate

The hoses are setup such that the bottom hose draws cold water from the bottom of the tank and the top hose delivers hot water to the top of the tank. The water in the tank does not mix much due to the low flow rate. Therefore the water drawn into the bottom hose stays at almost a constant temperature (the original water temperature) until all the water in the tank has been drawn out and been replaced by warm water that has passed through the collector. Dividing the tank volume by the time till the temperature starts to rise gives a rough approximation of the flow rate through the collector.

Tank volume = 12.8 litres (Note: I filled it this much so the total water volume would be 20 litres)
Time to empty: 25 minutes
Calculated thermo-siphon flow rate: 0.8 litres per minute

Note that the thermo-siphon flow rate decreases as all the water heats up and the density imbalance difference between the tank and the panel is less.

Power Calculation

The temperature change I was able to achieve was about 23°C over a period of 1 hour. The heat capacity of water is 4.18 kJ/kg/°C. There were 20kg of water in the system. Given this information it is possible to calculate the average power that was actually input into the water:

Power = 4.18 kJ/kg/degreesC * 20 kg * 23 degreesC / 3600 seconds = 0.53 kW or 530 Watts.

Efficiency Calculation

The collector area is about 1.4 m2. Energy available from sunlight is about 1000 W/m2. Therefore the panel receives about 1400 W of incoming power when aimed directly towards the sun. The efficiency is simply the power actually extracted divided by the power available.

Efficiency = 530 Watts / 1400 Watts = 0.378 or 38%.

This is quite comparable to commercially available solar collectors. However, I’m doing this in my back yard with uninsulated hoses, a non-air tight panel, a single plastic pane that’s slightly opaque, an open topped tank and no pump. The fact that I can achieve commercial level efficiencies with this setup is a testament to the design and indicates there is plenty of room for improvement in the industry.

Why this panel design works so well

Most home brew and commercial solar collector designs I have seen use metal (usually copper) tubing to carry the water through the panel. Metal fins are attached to the copper tubing. The fins are painted black. The fins heat up and conduct the heat to the tubing. Metal is a good conductor, but the heat has to travel a long way through a thin cross-section to reach the tubing. In my design, I used plastic which is a poor conductor, but the heat only has to travel about 0.3mm through a very large cross-section from the front surface of the panel to the water. I’ll illustrate why this is better.

There is a property of any thermal system called thermal conductance that indicates how much heat (power) can be transfered from point ‘a’ to point ‘b’ for a given temperature differential. The formula is:

Thermal Conductance = K * A / L
K = thermal conductivity (a physical property of the material)
A = cross-sectional area through which heat must travel
L = distance heat must travel (the distance from ‘a’ to ‘b’).

Comparison of a typcial tube-and-fin collector to a corrugated plastic collector.

Lets calculate the thermal conductance of a typical flat panel collector.

Assume the panel is 2’x8′ with 4 copper tubes running lengthwise and fins sticking out 3″ on either side of every tube (6″ per tube x 4 tubes fills our 2′ width). Suppose the fins are 1mm thick and also made of copper. When the fins heat up, the cross sectional area through which this heat must be conducted to reach the tubes is 1mm * 8 ft * 8 fins = 19504 mm2. The average distance the heat must be conducted is 1/2 the fin width or 1.5″ = 38 mm. The conductivity of copper is about 0.4 W/mm/degreeC.

Therefore the thermal conductance from the collector surface to the water is 0.4 W/mm/degreeC * 19504 mm2 / 38 mm = 205W/degreeC. In other words, a 1 degreeC temperature difference between the water and the fin will result in 205W of heat transfer. But the panel is receiving something on the order of 1400 W of incoming power from sunlight. To transfer all that power to the water by conduction alone the fins would have to heat up to 7 degrees C higher than the water temperature.

This is assuming a 1mm thick copper fin which is better than you’re likely to find in most DIY tube and fin designs. For example, some DIY books I have read recommend making fins from aluminum cans (typical wall thickness under 0.15mm).

Now repeat the calculation for the corrugated plastic panel.

The cross sectional area through which heat must be conducted is the receiving area of the panel itself (2′ * 8′ = 1486448 mm2). The distance the heat must travel to reach the water is just the thickness of the plastic wall or about 0.3mm. The conductivity of plastic is about 0.0001 W/mm/degreeC. Note that it is over 1000 times lower than copper which makes sense since plastic is general thought of as an insulator, not a conductor.

Therefore the thermal conductance of the system is 0.0001 W/mm/degreeC * 1486448 mm2 / 0.3 mm = 495 W/degreeC. In other words, a 1 degreeC temperature difference between the water and the collector surface will result in 495 W of heat transfer into the water. To transfer 1400W, the panel surface only needs to heat up about 3 degreesC hotter than the water.

Of course in practice, not all of that 1400W goes into the water. The conductance from the collector surface to the water is in parallel with another conductance from the collector surface to the outside air. The relative values of those two conductances determines how much heat goes where (Aside: this is analogous to current in an electrical circuit with two resistors in parallel.)

Ads by Google


In spite of the much lower thermal conductivity of plastic, using a corrugated plastic sheet as a collector achieves over twice the conductance between the collector surface and the water when compared to a copper tube-and-fin design with 1mm thick fins.

Reader Built Systems

If you’ve tried building one of these solar collectors or something similar, feel free to let me know (leave a comment below). I’d be happy to post photos of your system and any data you are willing to share, or simply link to your website if you have one.

  • Alex Nuget has developed a similar design with a brilliant mechanism for “turning off the panel”. He uses black particles inside the panel which settle to the bottom of the panel if the water stops flowing. This prevents catastrophic failure due to panel overheating in the event of a malfunction in the circulating pump or some other component of the system. He calls his design the Particle Panel. You can find out more about it at
  • John Hearty built a drainback system following a similar design to mine but using black dye in the water rather than painting the panel black. He has shared some photos which I have posted here: John Hearty’s solar water heater.

53 comments on “Build a simple solar water heater

  1. As this article was originally posted on another site, I’ve included all the original comments below.

    Richard M. Izzo said … 13:28, 22 March 2008 (PDT)
    Are store bought solar panels made from copper and aluminum considering these materials would transfer alot more heat.

    Rob said … 14:49, 23 March 2008 (PDT)
    Most commercial solar thermal panels use copper due to it’s wide availability, ease of assembly by soldering, and it’s high thermal conductivity. As I demonstrated above though, higher conductivity doesn’t necessarily mean more heat transfer. It depends on the geometry. My plastic panel transfers much more heat than a similar sized copper tube-and-fin design would. If copper panels were available with the same geometry as the panels I made, they would be VERY efficient. Unfortunately tube-and-fin geometry is the norm.

    rick said … 02:57, 2 April 2008 (PDT)
    I have the plan to use old central heating radiators as collectors. Thermal conductance of steel is 0.08. what do you think about this plan ??

    Rob said … 20:25, 8 April 2008 (PDT)
    Hot water radiators, despite their name, are primarily designed to transfer heat by convection, not radiation. As such they don’t make very good collectors. The ideal collector geometry is a simple flat surface facing the sun. This receives maximum sunlight with minimum surface area to lose heat by convection. But if you can get a bunch of old radiators for free or cheap, they’re worth trying. In order to work well, you would have to build an insulated enclosure around them, with one transparent side facing the sun. Also, be careful about what you call conductance and what you call conductivity. These are two VERY different things. Conductivity is a property of the material, independent of shape and size. Conductance is a property of a system, which depends both on the materials used and on their shape and size.

    John said … 08:42, 11 April 2008 (PDT)
    Great idea! Thanks for sharing! I’m building one at the moment to heat my pool (10.000L). I’m going to use a pump though. Do you think it will be ok to have a flow rate of 2-3 l per minute?

    Rob said … 12:48, 11 April 2008 (PDT)
    Power output is proportional to flow rate times temperature change (T_exit – T_inlet). Increasing flow rate also reduces the temperature change but transfers more heat overall. The optimum flow rate really depends on how much it costs you personally to increase pump size vs increase collector area. Both can capture more power, but usually one is cheaper per Watt of additional power captured. A pump and collector are considered well matched when the cost per additional Watt of adding panel area is the same as the cost per additional Watt of increasing the pump size. Flow rate and panel size both limit the maximum power you can hope to capture. The maximum temperature change (T_exit – T_inlet) that I achieved was about 25°C. At a flow rate of 2-3 litres per minute, a temperature change of 25°C is equivalent to a power output of 3500-5000W (remember, that’s only when it’s sunny… average power output would be much less). That is about the maximum you could capture with a flow rate of 2-3 litres per minute. Assuming 40% efficiency, the panel area to capture this much power in direct sunlight would be 9-13 m^2. As described above though, it might be cheaper to use a larger pump and smaller panels to achieve the same power output.

    Mike Hughes said … 08:23, 21 April 2008 (PDT)
    Rob Awesome idea just the kind of thing that I have been looking for I am anxious to start building and testing one myself. Have you any insight into the longevity of the system leaks etc. I also wonder if epoxy for the joints might work better than silocone caulk any thoughts?

    Rob said … 11:19, 21 April 2008 (PDT)
    Mike, unfortunately I have not experimented with this panel over long periods. I would be reluctant to install it anywhere that a leak could damage property. The silicone adhesive was a poor choice. It is more of a sealant than an adhesive, and is too flexible. There is a product called Goop that I’ve had better results with. It remains flexible after curing, but has a much stronger bond. Epoxy may work well too, but it is important to use something that will remain flexible at cold temperatures.

    Mike Hughes said … 12:16, 21 April 2008 (PDT)
    Rob, I understand the leak problem and am thinking that if installed off of the building at ground level then leaking may not be a problem. also I live in Texas so cold here is not the same as where you live. I think the low the last couple of years was somewhere around 20 deg F. Also I work for The Trane co. HVAC controls are my business and know a little bit about systems.. Anyway I am going to put your idea to work and will update you with feedback. Also I have been doing a bit of thinking and the corrugated plastic may also make an excellent plate and frame heat exchanger.. Just some thoughts.

    AndrewChristian said … 10:37, 15 June 2008 (PDT)
    This is great. I have built 2 panels w/ 1/2 inch poly and lots of “tees” to make a grid. Hooked them up to an old h2o tank. This one looks the most efficient. Did the pipe tend to roll at all when sawing the notch or did you have some type of jig to hold it steady? Thanks for a great idea-MAybe we can straighten out the world after all!

    Rob said … 15:19, 17 June 2008 (PDT)
    Mike, I have used corrugated plastic to make air-to-air heat exchangers before. Just cut squares and glue them together with the channels of every second layer turned 90 degrees. This makes a cube with 4 open faces and two solid faces. Then you just need to build an enclosure to channel air into/out of the appropriate face.
    AndrewCristian, the pipe isn’t very difficult to feed through straight freehand using the table saw fence to keep the blade centered. I thought about making a jig but found it wasn’t necessary. The writing on the side of the tubing helps visually to keep from rolling the tube during cutting.

    Bernhard said … 23:39, 20 June 2008 (PDT)
    I wonder how long this plastic sheet would stand up under UV? I thought the sun would break it down or at least make it brittle in a year or two?

    Col said … 19:54, 22 June 2008 (PDT)
    We had this material on the roof of our chicken shed for about 10 years till it was hail damaged. It was quite brittle by this time but still intact except for the holes punched through it. Without any water pressure it should last for at least 5 years I would imagine.

    Bob from Canada said … 11:03, 30 June 2008 (PDT)
    The corrugated material is also available in different materials, I was thinking of clear polycarbonate and using plumbing antifreeze died black. Another idea was to remove some of the internal corrugations and put a tight fitting cap on the ends causing the antifreeze to zigzag through the panel. Most sign suppliers have the corrugated polycarbonate and they also have very good sealants and adhesives. Also a note to Bernhard above the sign suppliers also have corrugated sheets with aluminum bonded to the front but I don’t think they are polycarbonate though.

    Peter Hitchman said … 00:32, 2 July 2008 (PDT)
    Hello Rob, I am in South Africa. How important is the size of the panel? Are your measurements mainly to suit the design of your roof? It would suit me to make larger panels, say 8′x 4′. Would this be OK? My e-mail add. is [email protected] Kind Regards, Peter Hitchman.

    Rob said … 15:13, 2 July 2008 (PDT)
    The panel size is not that important. I made an 8′ x 4′ panel before this one and it worked equally well. I made this one smaller so I could install it between my roof rafters.

    dbutlerdid said … 04:12, 3 July 2008 (PDT)
    Hi Rob, Was wondering if you have updated/redisigned your solar water heater since last year. Also, what part of the country do you live in?

    Rob said … 16:32, 5 July 2008 (PDT)
    I live in Vancouver, BC, Canada. Unfortunately I haven’t found the time to develop this design any further since last year.

    peter said … 23:43, 6 July 2008 (PDT)
    Hi Rob,thanks very much for this simple but thorough analysis. I was half way through making a copper sheet and tube collector and this has stopped me in my tracks. I’m a bit embarrassed that I did not do the same calculations before starting. I will at least need to halve the distance between the tubes. However, I have made a very efficient collector using recovered heat exchangers from air-con units. A simple insulated box with a glass front gives a peak of 700 W / m^2 in March – April on clear day. I currently have a 150 litre tank at 75 C. See images and Clearly it is not hard to get hot water on a hot summer’s day , the challenge is to capture useful heat in autumn, winter, and spring. For this you need to be boxed. I’ve tested using similar polycarbonate sheet instead of a glass front. It insulates better but does block some energy because it is not truely transparent. I measured a 10% reduction compared to a glass panel in spring. In this case the light has to pass through three layers of plastic. The air-con units are definitely the best capture I have seen and there is an almost endless supply at non-ferrous scrap dealers if you can get there before they destroy them. Great to see panel project with some science applied. Probably saved me wasting a lot of time installing this copper sheet design before realizing it was inefficient. Thanks.

    greg said … 20:40, 7 July 2008 (PDT)
    Wow. I just went to lowe’s tonight and bought some pvc and endcaps. I spent the day on ebay looking at solar pool heaters and I was trrying to think of a thin material that I could push water through. I do signs and it just so happens that I have a sheet of BLACK coroplast. Told my wife the plan and she couldn’t believe it when I showed her the picture of the pipe and the coroplast. I did a search on black coroplast and solar heating and yours was the first to come up! And here I thought that I was the first one to come up with it! My plan is to maybe use a solar water pump. Kudos!

    Mark said … 13:36, 8 July 2008 (PDT)
    Pity you are leaving all the objectionable and “irrelevant” comments on your site! I suggest that you clean it up. Otherwise very interesting. I wonder if running water in a shallow “bath” (box) with glass top and a black corrugated sheet of metal lying over the water would prove comparably efficient, more robust and simpler to construct…

    quilin said … 05:59, 11 July 2008 (PDT)
    This is a really interesting design. I am wondering if multiwall clear polycarbonate sheets that are used for greenhouses could be altered in this way to use as the walls/solar collectors for a radiant heat floor for the greenhouse.

    SharkyTM said … 06:56, 18 July 2008 (PDT)
    A good choice for adhesive would be 3M 5200 Marine Sealant, its a room-temp curable polyurethane caulking, and is incredibly durable. Its available at WalMart, or any marine supply store.

    Chris said … 17:52, 21 July 2008 (PDT)
    Excellent idea. I’ve been toying with the idea of making my own heat pipes, vacuum sealed and everything, but in the end, you are looking for return on investment – This has a low cost, provides a good way to get into solar.In terms of leaks, if this was on my roof, it wouldn’t matter if it leaked – it’s just like rain. I’m in Halifax, so I have a similar climate to yours. I was thinking of using city water pressure in the summer to drive it to my roof, and just disconnecting in the winter time. Tho I’d have to think about the chemicals in the build before I ran drinking water through it. Maybe closed loop is better..

    Arul Thomas said … 21:58, 21 July 2008 (PDT)
    this is best way to create the solat heater. really useful to everyone. everyone should read this.

    MarkD said … 14:11, 26 July 2008 (PDT)
    This looks like a great idea. The Coroplast panels are made from Polypropylene which is also used for food containers. Polypropylene pipe and fittings are also available, e.g. and If everything is Polypropylene then bonding with a polypropylene adhesive should be preferable/longer lasting to other adhesives. e.g.: Scotch-Weld DP8005 looking forward to trying this myself, thanks.

    Frazer RM Ross said … 20:27, 1 August 2008 (PDT)
    I am wondering if instead of corrugated plastic, one could use a double pane window (possibly layers of) instead?? They are abundant and easily can be freely found.

    Graham Hammond said … 10:25, 7 August 2008 (PDT)
    I built two panels about 40 years ago, one using an old radiator and the other out of odd bits of copper pipe. They were still heating my hot water when I moved ten years later. I am about to build again. Gas has just gone up 35% here in the UK. Great site, thanks.

    Ed Davies said … 14:31, 10 August 2008 (PDT)
    Some people in the UK are using silicone tube in the solar collectors for freeze protection: if/when the water freezes the pipe just expands so there shouldn’t be any problems. Have you any idea how well coroplast would do in that respect? The close connection of the water to a large area is very attractive. However, it does mean that the mass of water in the collector is somewhat larger than that in tube type collectors so the thermal mass is slightly larger making it less efficient when there are short bursts of sunshine – say on days when half of the sky is covered with cumulus clouds floating by.

    Rob said … 12:47, 12 August 2008 (PDT)
    Hi Ed, I think coroplast will burst if the water freezes but I have not tested it. Antifreeze would likely be required. Higher thermal mass does not affect panel “efficiency” (all other things being equal). It simply changes the thermal time constant (ie it takes longer for the temperature to rise/fall) resulting in an averaging effect. For example suppose a low thermal mass panel ranges between 5 and 25 degrees change (Tout – Tin) as clouds pass over. A higher thermal mass panel (with identical conductivities and flow rate) might range between 10 and 20 degrees change instead. The average temperature increase is still 15 degrees and thus both designs will transfer the same power, assuming they have the same flow rate. The coroplast design having a higher conductivity from panel to water, will be more efficient regardless of it’s thermal mass.

    Rob said … 12:55, 12 August 2008 (PDT)
    Hi Mark, A shallow bath with glass top and black sheetmetal lying on the water would be much more robust. It’s major drawback is that it can’t be tilted, making it a poor choice for most rooftop mounting or for collecting heat in the winter (when you want it most).

    Nitack said … 12:47, 20 August 2008 (PDT)
    Hi Rob, Your design is exactly what I was looking for. I have been working on a design for a poor mans solar generator using a thermal siphon setup. The whole discussion on how it would work is listed at Hypography, I think your design for the heat collector is perfect for my needs and wanted to alert you to my little project in case you found it interested and wanted to weigh in. I would have spent a lot of time or money buying copper/aluminum piping or trying to find old refrigerator coils if not for you. Thanks!

    Dwayne said … 10:16, 26 August 2008 (PDT)
    Dont know if anyone mentioned it ( didnt see it ) but there is a much more robust product they make for greenhouses. They sell this stuff under the name solexx and its guaranteed for 8-10 years (doubt if that would include this use though. It may however be strong enough to weld the tubes to the solexx and would probably be able to handle higher temps.

    AllanJ said … 06:48, 1 September 2008 (PDT)
    I thought this was an excellent idea, and then I went to find suppliers of panel plates of polycarbonate (I live in northern europe). Plenty of suppliers, but a word of warning from one of the manufacturers – it seems the material will start to break down when subjected to water above 60 centigrade! Not water alone, and not heat alone – but hot water. Since this is more or less the entire idea, this could be bad news. Any experiences with alternative materials? Someone mentioned something call solexx, but I can not find this material in my corner of the world – any info?

    Gary said … 09:31, 1 September 2008 (PDT)
    Hello Rob A very interesting article. But are your calculations for the heat transfer of a copper tube with fins correct? The conductive area of one eight-foot fin that is 1mm thick is 2438mm^2, four tubes with one fin each side is 19500mm^2. Doing the maths equates to about 205W/C power transfer for a copper system (rather than the value mentioned of 25W/C). This value is still about 40% of the ‘plastic’ value, but means that the copper fins only needs to get about 7C hotter than the water and not 56C. What is the manufacturing tolerance of Coroplast sheets? If the wall thickness increases to 0.4mm from 0.3mm, then the power transferred reduces to 370W/C. For my money there is not a lot of difference between ‘copper’ and ‘plastic’ collectors when all other things (construction, durability, power transfer etc) are considered.

    Dwayne said … 13:41, 1 September 2008 (PDT)
    AllanJ you can find info about solexx via google. Here a place that sells rolls of the stuff:

    Drone said … 21:33, 8 September 2008 (PDT)
    Great site, thank you for taking the time and trouble to do so & maintain it. My comment mostly relates to a suggestion of free flowing water within a closed box collector, is there not the likelyhood of algae and bacteria growth, particularly when the temperature is in the mid 30-40s.

    Rob said … 00:43, 14 September 2008 (PDT)
    Gary… Thanks. You’re right. I’ve made the correction. It’s also worth noting that 1mm is on the thick side for a tube and fin design (I’ve seen DIY tube and fin plans that recommend cutting up aluminum cans to make the fins). But I agree 100% that Coroplast is not robust enough for permanent installation. I recommend it only as a cheap way to experiment with solar heating and I wanted to demonstrate that its performance (while it lasts) is as good or better than commercial panels. I’m not suggesting that plastic is better than copper but rather that the widely adopted tube and fin geometry is not necessarily the best choice.

    Rob said … 00:49, 14 September 2008 (PDT)
    Drone… Yes, algae and bacteria growth would be an issue. Chemicals can be added to prevent that (as well as to prevent freezing).

    Paul Smith said … 03:35, 15 September 2008 (PDT)
    Great article! I have often looked at these channels and thought that they ought to have water flowing through them. My question is, “what do you do with the hot water?” It may seem obvious but I want to know to where I can pipe the hot water. Is it stored in the hot water tank? Directly or via heat transfer coils as in my immersion heater? Or do you pipe it directly to your hot taps? Can this system be used to effectively preheat the water going into an ‘on-demand’ hot water heater? How could I use this system to augment my gas-fired-central heating radiators? I am happy to fit as many motorised three-port valves as would be needed and I can handle controling these with appropriate temperature sensors and a bit of logic but what would be a sensible, minimum impact way to plumb this in to existing systems?

    Rob said … 21:45, 20 September 2008 (PDT)
    Hi Paul, for domestic hot water heating, or for using a preheating tank for an on demand system, there is a heat exchanger called a Solar Wand (try a Google search) that can be installed in an ordinary hot water tank allowing an external fluid to be pumped through it. This is a “minimum impact” way of heating water. For home heating, solar thermal panels go great with radiant floor heating. A typical system would involve a large insulated water tank and a series of tubes running under the floor. Water is circulated from the tank to the panels to heat it up, and from the tank through the tubes to heat the home. As you can imagine, it is not nearly as “minimum impact” as the installation for domestic hot water heating.

    PRALHAD BHIDE said … 00:10, 22 September 2008 (PDT)
    Respected Sir,I am quite impressed with your innovative idea of using corrogated plastic for making solar water heater and planning to build one. Do you have any data about the life of Corrogated plastic material. As it will go constantly under temperature cycles , how long it will last?

    Rob said … 22:14, 22 September 2008 (PDT)
    Sorry. I have no lifetime data. I only built this as a proof of concept and did not install it. I’m currently pondering more robust designs for installation in my attic.

    Frank said … 21:19, 23 September 2008 (PDT)
    Very neat application of an idea I have been toying with for over ten years. You simply worked it all out.I have done some experiments and have maybe some valuable additions: If the sheets are polycarbonate (PC), they will bond very well to glassfibre reinforced polyester (but AllanJ wrote that PC could have issues with hot water). Polypropylene is a better choice: it is (within limits) much more compatible with hot water as well as with freezing temperatures. But the thin walls will permit less and less pressure as the water heats up, so it is imperative to have a pressureless system (open to air) as water expands quit a bit while heating. You must give the water ample room to expand and compress. As PP is a real “plastic” it is far more flexible compared to PVC and PC and even ABS, there is a fair chance that the panel will withstand freezing. I have never heard of a PP pipe bursting by frost. But there stays the weak point of the collectors and their bonding to the panel. PP can more or less easily be melted, thus welding might be a solution. But the thin walls are a challenge. One way to possibly solve this would be to use square tubes as collectors, cutting a slit in one of the sides of the square as well as in the horizontal surface of the panel, aligning both slits, then hot wire welding the collector and the panel together with two parallel wires (possibly leaving the hot wires in the weld, which is a commonly used practice for welding PP fittings on pipes). then all you have to do is to seal off the ends of the panels (possibly by welding them shut too) and you have a monolithic panel. I have read somewhere that if the PP is black, it will have good UV resistance as the carbon used to blacken it forms a UV protection by itself. If not, the right paint will ensure UV protection. One last thing: the same panels can be used as wall or ceiling radiators inside the house. Many thanks.

    Deniz from Turkey said … 10:56, 14 October 2008 (PDT)
    thanks for theexplain for build a solar water heater,I will tray to built in soon,again thanks a loot

    fromhull said … 20:46, 15 October 2008 (PDT)
    For all the canadians out there…if you’re interesting in trying this, there are thousands of election posters hanging on posts and poles all over the country. Guess what they’re made of. oh yeah. i just grabbed myself a few.

    Arthur HasHagen said … 10:23, 30 October 2008 (PDT)
    Thank you Rob, and thanks to all of the contributors for your thoughtfull input

    Dean – Vancouver Island said … 13:59, 6 November 2008 (PST)
    Great info, Rob, thanks.For a more permanent installation, could one use 2, 3mm copper sheets, separated by 3mm vertical strips. It’s almost mimicking your poly unit but the greater conductivity of the copper could be beneficial. I’ve tried to calculate the thermal conductance of this by swapping out the plastic conductivity with the copper and changing the thickness of the material. My result is not believable (too large). Would you provide some insight to this idea? Thanks again.

    Rob said … 14:54, 6 November 2008 (PST)
    Hi Dean, conductance is proportional to conductivity and inversely proportional to the distance heat must travel through the material. Switching from a plastic to copper panel of the same area increases conductivity by 4000 times, and going from 0.3mm to 3mm thickness increases the distance by 10 times. Therefore, the new conductance would be 400 times greater than for plastic panels. So the conductance of such a copper panel would be around 200000 W/degreeC. That doesn’t mean you can transfer 200000 Watts of heat. It just means that to transfer say 1000W of heat, there would be only a 0.005 degreeC temperature difference between the panel surface and the water inside the panel. That’s the good news. The bad news is your panel would weigh over 70kg empty and contain about $270 of copper. 3mm thick is overkill. Going thinner would provide better results.

    Dean – Vancouver Island said … 18:06, 16 November 2008 (PST)
    Thanks for the explanation, Rob.Yes, $270 is a bit expensive. Maybe aluminum would be a better option. Further on the idea, is there any reason why the internal strips couldn’t be removed and just leave the spacer around the perimeter of the panel to create the volume between the 2 sheets?

    Rob said … 00:56, 18 November 2008 (PST)
    Hi Dean,Without internal ribs, the pressure difference between the inside and outside of the panel will cause it to either expand or contract. With that much area to act on, it takes very little pressure to generate a huge force (ex a 0.5 PSI difference over 16 ft^2 of panel results in over 1000 lb of force). It takes about 27″ of head to generate 1 PSI. So if the top of you panel is 27″ higher than the bottom of your panel, the pressure at the bottom will be 1PSI more than the pressure at the top. To withstand the resulting forces you either need lots of internal ribs or some kind of external support.
    If the panels are the highest point in your system and you pump the water into the top of the panels rather than the bottom, and you pump the water slowly enough, you can create a scenario where the pressure inside the panel is lower than that outside (effectively the water draining out of the bottom of the panel results in a partial vacuum inside the panel). In this case you can replace the internal ribs with any appropriate material to hold the two panels apart by a few millimeters while still allowing water to flow between them.

    I once built a small panel like this out of 18″ wide aluminum flashing (from the roofing supplies section of Home Depot). Rather than put something between the panels to hold them apart, I used a center punch to put divots in the front panel on 1″ centers. When the two panels are squished together by outside air pressure the divots held them apart, leaving enough gap between the two panels that water could still flow. It worked reasonably well, but was much more labor intensive to construct than the coroplast panel.

    Chris – England said … 01:30, 18 November 2008 (PST)
    Fascinating read, Rob. Did you ever measure its performance in the absence of sunshine eg cloudy day? How well do you think it would work in England in winter, latitude 52N with much lower insolation? I can’t find info about relative summer/winter Watts per sq metre from sunshine. Any pointers would be appreciated. PS I’m thinking about heating a barrel of water in a greenhouse during the day to provide a source of heat for the cold winter nights.

    Rob said … 12:39, 18 November 2008 (PST)
    Hi Chris, I didn’t measure its performance on a cloudy day. I’ve no idea how well it would work in England. I’m also not sure about summer/winter variation in W/m^2 of solar power. Sorry.

    acwhite farms said … 10:35, 20 November 2008 (PST)
    What about the solexx panels? Will they work as will as the coroplast? They are havey duty enought to stand up to tones of abuse on the farm hot or cold weather. They would hold more h2o. I would like to build one about 4×8 inclose it in a insluated frame and plumb it to my electric h2o heater. Can I hook it to my city water line that feeds my hot h2o tank. Would the coroplast work better if you painted it black?

    SdM said … 14:23, 5 January 2009 (PST)
    Good answer, wrong science! In your get-up the solar energy is transmitted to the water through radiation not by conduction (polycarbonate, specially this thin, is essentially transparent). The fraction of energy absorbed by a transparent material is Qa = Qi e^[-1/(t x gamma)] where Qi is the incident radiation, e is 2.72, ^ denote exponentiation, t is the thickness of material and gamma is the absorption coefficient. For pure water gamma varies between 0.0001/cm (blue) to 0.01/cm(red) to 1/cm (near IR) <>. Gamma is 1/cm to 3/cm for polycarbonate and is pretty flat with wavelength, but in your material it is so thin that it will not absorb much. Your sunlight (mostly visible and IR <>) traverses at best 0.8cm of water (through the sheet once and then reflected back from the white styrofoam) and can only absorb about 25% or the solar radiation and mostly in the IR. You would get better results by adding dye to your water to turn it practically opaque or by painting the back of your panel (the part against the styrofoam) black. This will cause all or practically all the radiation to be absorbed directly by the water or transferred to it by conduction from the back panel. I’d be interested in seeing a comparison test.

    Rob said … 22:32, 6 January 2009 (PST)
    Hi SdM, thanks for the comment. However, I think you missed the part where I painted the front surface of the panel with a flat black spray paint (after testing for leaks)… and the part where I measured a 38% efficiency, which would be impossible to achieve if only 25% of the solar radiation was being absorbed. So the solar energy is transmitted to the painted surface of the panel by radiation and to the water by conduction. I agree it would be interesting to compare painting the front of the panel vs painting painting the back of the panel black, or dyeing the water. My impression, looking at my particular panel was that the bare panel was quite reflective (to visible light anyway). Thus I chose to paint the front surface black instead of the back.

    SdM said … 05:57, 7 January 2009 (PST)
    Sorry I missed step 17. 😉

    Fer said … 08:43, 12 January 2009 (PST)
    I was wonder if using Bayer Makrolon sheet will work better? They have white and transparent.

    Daddyo said … 13:47, 23 January 2009 (PST)
    Really nice work. I’m aiming to use this to heat a pool. One thing about your selection of plastics that you probably considered is that the coefficient of expansion of Coroplast (polypropylene) and ABS (unglassed) is very similiar – a good thing for structural longevity. A good paint to use should be Corogloss I’m testing it right now. I had to add a bit of flattener since it comes glossy, and thinner. It is very expensive ($40/quart) and has lead in it, so you probably don’t want this in your drinking water. But lead always made paint good right?? There’s information on the web regarding roof temperature vs. roof color, and I’ve found that with black there’s a 50 C (90 F) rise above ambient air at low wind velocity ( Since I don’t get higher that 110 F locally I think I should be below 200 F w/o water cooling. The melting point of polypropylene is 329 F (165 C) which gives 129 F margin at best, seems like plenty for low pressure systems like a pool. (

    I’ve also found two manufacturers of similar design, but you’ll pay 3x easy for their premade products. They’re joints are very robust. Here’s one of them, their Vortex line

    I’m thinking with the UV protection of paint on here I may get 7 years life out of this setup. I do wish there was a better seal from the headers to the collectors though, I’m thinking a fillet using a 90 degree right angle piece of plastic, or rectangular plastic rod on the joint may do the job. I’m still not sure I need an insulator behind the panel, most pool heaters don’t have such a thing. I’m looking for someone to talk me out of covering 1/2 my roof in the next couple months with this, otherwise I’m going for it.

    Julian. said … 14:30, 2 February 2009 (PST)
    Very good article. A friend of mine in France made a similar system to feed his swimming pool. He has quite a large swimming pool surrounded by trees, so he built around 8 or so panels about 40′ up a slope above the pool level, fed by a large pump. It’s been running for about 10 years or so, and the only (minor?) problem he’s had is that over time the wall between adjacent cells break down. It isn’t too much of a problem away from the edge, but he’s had to fibre-glass reinforce the edges on a couple of panels. I’m planning on building a downsized version for my 12′ paddling pool this summer 🙂

    M.D. Naughton said … 04:28, 3 February 2009 (PST)
    Excellent project. Well considered and argued. However i have a suggestion. Would you consider bonding a layer of TiNox/Blue Tec etc. coated copper to the front of the plastic sheeting? One could use high temp bonding (heat paste) to make the connection. I know you will have to increase the distance that the heat energy must travel from 0.3mm to 1.3mm but coated copper is a much better absorber of energy than an insulating plastic.Have you also considered using insulated glass units (Double Glazed) as the primary glazing cover? Once again, thanks for the article.

    Daniel said … 13:52, 7 February 2009 (PST)
    What if, instead of using styrofoam as insulation in the fourth “building the frame” section, I used housing insulation? Would that keep more heat in and make the system more efficient, or is the styrofoam better?

    Tony said … 12:50, 15 February 2009 (PST)
    Hi 8foot (2438.4mm) * 8 fins (76.2mm) = 1486448.6 mm2 Not 19504 mm2 …..Am i reading the sums wrong?

    Rob said … 20:57, 15 February 2009 (PST)
    Hi Tony, You’re calculating the collecting surface area. In a conventional tube and fin design that is NOT the same as the conducting cross-sectional area. The cross-sectional area through which the heat must be conducted to reach the tube is equal to the fin thickness (1mm) times the fin length (8ft = 2438.4mm) times the number of fins (8) = 19507.2 mm^2 (I had rounded 8ft off to 2438mm to get the value of 19504 in the article but it’s close enough). This is the disadvantage of a tube and fin design. The conducting cross-sectional area is usually much less than the collector surface area, impeding heat transfer. In my design the two areas are the same.

    Rob said … 21:01, 15 February 2009 (PST)
    Hi Daniel, It makes no difference what kind of insulation you use. All that matters is R-Value (the higher the better). I find styrofoam is easy to work with since it is rigid and doesn’t compress under the weight of the panel. Also, if the panel leaks, styrofoam won’t soak up the water like a sponge.

    Adam said … 13:36, 17 February 2009 (PST)
    Hi Rob. I think you are doing a commendable job of enlightening those who are in need. Thankyou.Adam.

    Jeff said … 20:59, 17 February 2009 (PST)
    Would yo allow me to use your procedure for my science fair project. My topic is discovering if a solar powered water heater can sufice for the daily needs (of hot water) of a family. I have done other research and it seems as though it actually can. Please let me know. Thanks.

    jacob said … 14:42, 19 February 2009 (PST)
    hi i would like to do this to heat my poollike the store bought solar pool heaters. but the only local supply i found for the sheet.has it in 10mm do you think this will be ok

    Ron said … 10:50, 27 February 2009 (PST)
    Hello Rob – really a great idea – any update on how well the material and the glue joint has held up? I’m planning on trying cpvc end pipes (abs is hard to find here)and believe the glue joint would be a lot stronger if the pipe slit could be pryed open wide enough to pass the sheet through with a bead of the glue/caulk on either side of the sheet contact area, then released to squeeze a bead of it inside – that would likely seal better because the pressure would be trying to drive it into rather than out of the joint. Thanks

    Rob said … 19:30, 6 March 2009 (PST)
    Hi Ron, Sorry. I have no data on how well this would hold up over time. I only built it as a proof of concept and did not install it.

    Rob said … 23:55, 6 March 2009 (PST)
    Jeff, yes, you are welcome to use the procedure for you science project. Before fossil fuels and electrical power were widely available in North America, solar water heaters DID suffice for the daily needs of families. It’s ironic that most people think of this as a NEW and progressive technology when it has been much more widely used in the past than it is today. Good luck with your project.

    Rob said … 23:56, 6 March 2009 (PST)
    jacob, yes, 10mm would work just as well. It will just be a little heavier when filled so you will need to support it well. You may also have a bit more trouble cutting wider slots in the header tubes but you can probably manage OK.

  2. I believe that another reason why copper is used in collectors is sanitation. Copper is naturally bacteriacidal. So, no bacteria growth in the collector, lines, etc etc. Plus, copper is a proven container for the safe storage of potable water, even when repeatedly heated and cooled. The plastic, glue or solvents you use in your homemade collector MAY not be designed or intended to be used with potable water, or are not designed for the heat/cool cycles that you intend to put them through. Plus, biofilms (bacterial growth on the surface of the plastics) will occur and be maintained unless proper sanitary processes are conducted on a regular basis. Just some thoughts. Cool to make the hot water collector so inexpenisvely and with as high an efficiency as you observed. Not so cool to potentially cause harm to your family. Get the facts, check out your plastics and go for it!!!

  3. Hi Eric. This collector design is not intended to be placed in series with a potable water source. Not only would nasty chemicals leach into the water, but the panel will not survive city water pressure. However, in Canada at least, even copper pipe collectors are not designed for circulating potable water. The water would freeze in winter and burst the pipes. Therefore, antifreeze must be added to the water circulating through the collector and a heat exchanger (such as a Solar Wand) must be used to transfer the heat to a water tank containing potable water.

  4. Hi, This is great I have been looking at using solar power to heat water for a child’s paddling pool. Basically, heat water one day and use the following. I was thinking of using a solar powered pump, but thermo-siphoning costs nothing !

    If I used a double glazed unit (I have a couple lying around) instead of the plastic sheeting would this create too much heat and indeed melt the corrugated plastic sheeting ?

    Also, if the system was plumbed into a supply and as water was drained out, and fresh water replaced by the use of a ball valve, if the whole system was at an angle as you suggested when filling, wouldn’t it enable the air to be removed constantly ?

    If the storage container was below the height of the solar collector, I assume a pump could be used to pump it back up to the top and let gravity take it’s cause ?

    Do you think a drill pump fitted to a rechargable hand drill running off a solar panel work ?

  5. I really hope that if you install a water heater of any type in your attic that you will make a proper emergency drainage system.

    There is no way that any type of drainage system could be more expensive than replacing your walls, ceilings, and carpets.

    By the way, these folks seem to have a good deal on propylene glycol- free shipping and 95% concentration (so you don’t have to pay for water).

  6. Hi Westy,
    If you use double glazing you probably will get too hot for the plastic parts inside.

    It is possible to design a system that will expel air automatically. You simply need to ensure that at any point in the system there is a path that rises continuously from there to the highest point (highest water level) in the system, and that the highest point is the surface of the water in the storage container which is exposed to the atmosphere.

    For me that would have meant drilling holes in the side of my cooler (one at the bottom for the exit and one at the top for the inlet). Then air bubbles would simply rise through the system and be released to the atmosphere. I did not want to drill holes in my cooler so I opted to bring the exit and inlet tubes in from the top which introduces the problem of trapping air and makes it necessary to take steps to remove the air in order to get thermo-syphoning to work.

    When the solar collector is mounted higher than the storage container a pump must be used to circulate the water. A drill pump may work, but I don’t know if they are designed for continuous duty, and I don’t think they are very efficient (ie you may require a bigger solar panel than you think). If you have access to AC power, I would recommend using that instead of adding photovoltaics. You might try an AC aquarium pump or a sump pump. But then you either have to turn it on and off manually, or design a thermostat to turn it on when the collector is hotter than the storage container.

  7. Hi Rob,
    Interesting. and food for thought.` I have the good fortune for solar by having a winter place in Puerto Angel, Oaxaca, Mexico, on the west coast, about 200 miles south of Acapulco. The air temperature is relatively stable, around 80 degrees F average close to the beach, with breezes, never less than 60 degrees F at night. But are there ANY solar water heaters or electric panels in this area? Nope! How do they heat their water? By using large cylindrical ABS tanks on the roof which serve for both “hot” and “cold” water, that is, its “tepid” around 80 degrees during the day and about 75 degrees at night.
    I was planning to make a solar hot water heater from around ten black ABS pipes, around 6″ diameter, about 4′ long, all in a row, connected at top and bottom in parallel, with ABS inlet and outlet fittings at top and bottom, with the water supply coming from the large ABS tanks located about 6 feet above the solar water heater. ABS pipe is strong, cheap, and available locally. I was also thinking of putting the assembly in an insulated box, glass front, so that it would not lose much heat overnight. I think ABS can withstand some heat, but I am not sure how hot.
    When you mentioned that a black metal panel in a “bath” of water in a box would be very sturdy and efficient, I then thought of trying that method, since the tropical sun is so high in the sky, nearly vertical, and the rays so hot that even a horizontal position would not be giving up much efficiency. I was thinking that the metal plate should be copper, blackened with sulfur, as it would be very efficient as a conductor, be sturdy, and anti-bacterial, anti-fungal. Again, I think I would insulate the sides and back of the box.
    It amazes me that there are NO solar water heater of any kind or solar collectors anywhere in this area. Just those black cylindrical ABS tanks, which are about 5 feet tall and about 3 feet in diameter. No separate hot water tanks except in some hotels that cater to Americans. This area is has not yet been discovered by tourists, though Puerto Escondido doe have tourists, and a new resort, Huatulco, is being built near here. But there are no solar systems in those towns either!! Anyone I talk to around here about solar systems of any sort results in a blank stare. I was thinking of selling some, since they would be cheap. Very backward around here! If you Google Puerto Angel, to to the tomzap site, click on “aerial view”, you will see the area (beautiful beaches), see black tanks, but NO solar water heaters anywhere!
    What do you think of my two ideas??

  8. i want make soler system for heating water by my own,so SIR please anform me the basics of make the stuctur
    i will thankful to yuo for this kind faver.
    regards.MUHAMMAD RAFI

  9. Hi Dave,
    ABS should withstand hotter temperatures than you want your water, so as long as it always has water in it, you should be fine. If the water leaks out, then the ABS might overheat depending on how well insulated and glazed your panel is.

    Regarding the idea of the shallow tank of water with black metal sheet “floating” on it… An even simpler design is to paint the bottom of the tank black, or use a black liner of some sort. It does not need to be a good conductor since the absorbing surface is in direct contact with the water being heated. The only issue with this design is that water may condense on the underside of the glazing reducing the amount of radiation transmitted.

  10. Rob,after reading a lot of s/h, cost and time consuming, i am ready to make one of coroplast etc following your instructions .in benefit of the World and my pocket,because electricity is very expensive and and will become more.i have a electric heater planning to help in get the usage lower with this system.
    the place is in Mexical Mexico Baja California where the sun in summer is at 120F and average 110fand winter some times goes to freeze. and some pretty good wind
    what kind of suggestions do you have to this quetions
    1.-do you advice to frame it with all the insulation,plastic top,black paint etc.etc.
    2.-i will have to disconnect the system in summer?
    3.-in winter add some antifreeze to the water tank?
    i dont know what other kind of situations i will get into, but what i know is i am going to build it, a friend of mine is a plummer and electrician and together we are planning to build it.
    wish me luck,because if everything goes well i will do one to my mother and soo on,shes always compleining of the gas cost
    i want to thank you for all the support and help you are doing to STOP the World Heat

  11. In your drawing of the typical tube and fin design, did you mean to write 1.5in instead of 3in?

  12. Hi sameb,
    No. 3″ is correct. In the text is says the average distance heat must travel is 1.5″. That is because the fin is 3″ long. Heat must travel a full 3″ from the end of the fin to the pipe, but it travels 0″ from the base of the fin to the pipe. On average therefore, heat needs to travels 1.5″ through the fin to reach the tube.

  13. Dear Rob,

    I am interested to add a solar water heater on top of our camper. Since I also live in vancouver, BC would it be possible to meet with you and see your prototype? Where I grew up in Israel it is very popular (mandatory actually…) but those systems are too heavy and not mobile. I liked your use of plastic however if you check carefully, the acid rain might corrode your plastic so a glass cover is recommended. I also suggest you warn people not to drink those water!

    Best regards

    Alon Newton – Vancouver, BC

  14. Hi Alon,
    I experimented with this a couple years ago and I no longer have the prototype so I can’t help you there. It was never my intent that this design be used for drinking water. I’ll add a warning about that. Thanks for the comment.

  15. Hi rob, this is aurelio
    i did start putting together the solar panel and i can not find the pl700 in california and arizona is restricted what other adhesive will you recomend, i did try different other bonding like abs bond,flex,windshield sealant,and even the weather roof sealent, white silicone etc and none of this have worked, can you please advice what other kind of adhesive can use.
    thank you very much

  16. Hi Aurelio,
    There are few recommendations for different adhesives from readers in the comments above if you search through them. Even the PL700 I used is not a great choice. One reader recommended this: I have not experimented with other adhesives so I can’t make any personal recommendations. Sorry.

  17. Has anyone tried using air to transfer heat to a car heater core or aircon type of radiator? Meaning make a enclosed box with black insides that heat the air, then use a small slow muffin fan to circulate the air past the radiator (at the top of the box), and then move water through it etc..

    The idea is less materials, lower cost, less leaks, with the air taking up all the space for heating but you move that heat with the fan so it’s like collecting with long tubes over the same area. It might be as simple as painting the wood inside black or putting a black trash bag in there. Maybe convection would move the air, but I think a fan would raise the efficiency and you could run that with a small solar panel.

  18. Hi Rob,
    This is a great idea, and the method of construction is well explained. I am from India. I have seen people here embrace solar energy as long as the device is affordable. A commercial version of your solution will definitely make this solar water heater affordable to a large number of people in developing countries who otherwise could not afford one (commercial ones start from 300$). I will spread this idea to all those interested. I try constructing one myself to demonstrate this idea. Thank you for your post Rob.

    Well, just one concern. I am worried about what Eric said about sanitation. Anybody with any ideas on how to make this system “anti-bacterial”, “anti-fungal”…just safer.

  19. Hey Rob! Great little DIY Solar Collector. I came across your plans in the website while researching some stuff that I want to incorporate into a 12’x12′”Hoop-house” greenhouse. I hope to take the edge off of our Calgary winters with this. I hope you can fill in the blanks for me on a couple of questions:

    1) I plan to use some form of thermal mass in the floor to hold and maintain heat (maybe 6″ sandbed with finned-pipe throughout). Should I run my pipe direct from the collector to the sand-bed and then to the storage container, or from the collector to the storage container THEN my sandbed?

    2) Since I’m cheap and dont want to run a pump all the time (allowing syphon to do the work), will I be able to install this system to heat a sand bed?

    Any and all help is welcome. You seem to really know your stuff 🙂

  20. Hi Tweac-It,
    To answer your questions:
    1) The only purpose of the storage container is to store heat (not water… the water is just an effective medium for storing heat, much like sand). If you are using a sand bed to store the heat, there may be little benefit to having a large storage container for the water. In any case, the sequence of connection probably isn’t too important. Both the storage tank and sand bed will be in the greenhouse and both will store heat. Changing the order of connection shouldn’t have any effect on the amount of heat captured. One thing to note though is that a sand bed will dissipate heat into the ground. You might consider a couple inches of polystyrene underneath the sand bed to insulate it from the ground.

    2) Unless you can locate your collectors at least a couple feet lower in elevation than the sand bed, you’ll need to use a pump to circulate the water.

  21. Thanks for the info Rob.

    I plan on making an insulated sand-box w/ paving blocks as the top surface. This way I will always know the walking surface is in contact with the sand. I will insulate 4 sides and bottom.
    As for the location of the collecter vs sandbox….I guess I will need to use a pump 🙁

    Any idea if your design (as posted and unmodified) will be able to fend-off freezing inside a 12 x 12 hoophouse? I’m not sure of all the math aspects of air volume/heat dissipation…etc.

  22. Hi Tweac-It,
    I don’t know if you will be able to fend off freezing temperatures. It will depend on the outside temperature, the wind chill, how well the building is sealed, how the panels are oriented, and how much sunlight you typically get. One way to test how much energy you’ll require is to heat the hoophouse with an electric heater for a couple days during the coldest part of winter and see how many Watts on average are required (use a kill-a-watt meter). For example, if you find 500 Watts on average is enough, then you might get away with 4 collectors as described, assuming the collectors see sufficient sunlight for about 6 hours per day. The 4 collectors would produce about 2000 Watts in total, but they would only do so 1/4 of the time, thus they would produce 500W on average.

  23. Rob,

    I came across your site a few months back while I was prototyping my own plastic thermal panel. I think plastic solar panels have the potential to completely revolutionize the industry by lowering the cost. However, you hit on the problem I was working on:

    “because the whole collector is made of plastic, it is important that the temperature doesn’t get too high or it will soften and possibly spring a leak. 80 degrees C (176 degrees F) is about the limit. Don’t think it can get that hot? Think again. …Therefore this may not be a practical design for residential installation”

    Well, I put some thought into it and I think I have a solution that will prevent the panel from overheating (the panel turns itself off when no water is flowing), which means that it can be insulated and hopefully compete with commercial systems, only at 20% the cost. I think the efficiency may also end up being higher then commercial systems. I built a prototype:

    …and posted a video of it to youtube.

    I just wanted to give you a big thanks for your website and the work you are doing. Its people like you that are going to save this planet.

    Blue Skies,


  24. Hi,

    I’d like to use this design to heat a hot tub (in addition to a bio-gas powered water heater: read more about biogas here:

    I’m a bit worried about chemicals leaching into the water. I’m not planning on drinking the hot tub water, but are there any issues with exposure to the plastic chemicals?

  25. Hi Mike,
    Your guess is probably as good as mine, but my guess is that exposure to the plastic is not an issue. I doubt the chemicals that leach from the plastic are any more harmful than the chemicals people typically add to their hot tubs intentionally.

  26. I have read your blog and another one you commented on about using Coroplast for thermal collector panels with great interest. I think Coroplast is an excellent material for a collector mainly because of it’s higher temperature tolerance. My wife tested baking a sample at 250 degrees F and it came out feeling just about as rigid as normal. We also filled it with water and put it in the freezer and it did not deform from the ice.

    I am however concerned about freezing problems, more due to glued joints at the ABS pipe at the top and bottom of an assembled panel bursting. I plan to have a large holding tank of water used directly with the panels without a heat exchanger to the holding tank so I’m not too interested in using a bunch of anti-freeze. I am planning to build a drainback system using these panels, and just painting them black.

    I also saw that polypropylene does not tolerate UV light well and will become brittle and break after long exposure. Coroplast can however be made special order with a UV absorber mixed in with the polypropylene. I am getting some regular Coroplast from a local sign shop that does not have the UV protection, so I’m looking into paints that absorb the UV. If the prototype works well I’ll look into getting UV protected Coroplast for additional panels.

  27. Alex, your particle panel idea is interesting, but I wouldn’t go throwing a lot of patent lawyer money at it until you have it actually working in a real situation. I mean protect it with minimal costs and move forward with making it actually do something at home.
    From a guy who does this stuff a lot, I see problems. First, you are going to have to control the flow, so there’s more cost. The restriction of flow will use more pumping power, calculate that in.
    You are going to have to insulate it in a box of some kind during the winter if you want heat enough to cover the costs of building it.
    When you do insulate it, the box is going to get hot, maybe too hot for your plastic. Any curve or angled surface at the right angle to the sun will look “black” to the sun and cause heating, even with no water.

    Now the question is, did anyone read my above post about the hot box with a simple car heater core, or car oil cooler in it and a fan? Using the air in the box to transfer the heat?
    Just a black box with fiberglass insulation and glass front. Simple, cheap.
    I’m going to build it, I did some small scale tests and things look good heat wise, but if someone already did this and has bad data I don’t want to waste my time.
    As for the heat box I’m talking about, there would be no problem with it getting hot when turned “off”, the wood, fiberglass insulation and glass front would handle it.
    You can “reverse” it in summer (nights) when it’s cooler outside than in the house, open a “summer” vent hole (with screen) and let cool outside air circulate inside.
    And, don’t worry, you can’t patent a black solar box.

    So please, comment, or did I not explain it properly? (feel free to ask questions)

  28. Repeating these comments I left on the instructables site for this topic. My wife and I finally got one of these built and tested. We built it as a drainback system and used plain water dyed black using pond dye. We did not paint the panel. We got frosted tempered glass panes from Craigslist to build this and the next ones. We used an old hot water circulator pump also from Craigslist.

    We did a 4 hour test on a clear day, readjusting the panel angle a few times during the test. The full spreadsheet is available but I was not sure if it could be posted here. It is a 1.814 square meter panel with 37.85 liters (10 gal) of water in the system. We used a 55 gallon plastic drum for the tank. The tank and hoses were not insulated.

    Starting temp was 53.8 F. At the 1 hour mark the temp was 92.7 F, average power was 952 watts, 52% efficient. At 2 hours, 117 F, 768 watts, 42%. At 3 hours, 127 F, 593 watts, 33%. At 4 hours, 130 F, 464 watts, 26%. We also did a stagnation test with no water in it, and it got up to 152 degrees F on a 45 degree day. We are looking forward to mounting it permanently and testing reliability/longevity. One thing we still need to do is get UV clear paint to help protect the panels from UV breakdown, and see if that affects the efficiency much.

  29. Hi Jim,
    I think your black box is a workable idea. Cost would be low and it would be quite reliable, but efficiency is lower if you are using convection instead of radiation to transfer the heat. The air will give up it’s heat to the enclosure and glazing as readily as it will give it up to the heater core. While the cost per unit area would probably be lower, the cost per kWh might be high. But perhaps not. Hard to tell without trying.

    Good luck.

  30. Hi John,
    Sounds like a great project. Keep us posted. If you like, you can send me some pictures and your excel sheet and I will post them on this page. -Rob

  31. Hi Attila,
    Sorry for the slow response. You are welcome to republish the content on this site as long as you attribute and link to the original source. Thanks, and good luck.

  32. I know this may sound ignorant, but i have a few questions. 1. I understand that thermal siphoning occurs with the “heated” hose higher to fill the “tank” (whatever storage container youre using for heated water), and the lower hose (cooler water) goes back into the heated source (the homemade panel)? correct? 2. How do you use the water in the “tank”? Do I make an exit hole from the tank to use the heated water? My idea is as follows: Build the system as you describe – hot hose ABOVE the cool hose (as a result of slightly tipping the heat source – panel), you now have a tank to collect the hot water………from this tank you drill a hole to insert and seal a fitting and hose that has a water pump inline. The water pump is connected to say a shower head. You turn on the water pump to expel the water from the “tank” to give a hot shower? Will this work as described if you put the “outlet” of the water pump at the low end of the tank to pump the hot water out? I understand that the system will run dry when all water is expelled. This would be used in a remote situation.

  33. The method you described would work fine. If the tank can be opened, you could use a submersible pump right inside the tank and have the exit hose come out the top. Submersible is usually cheaper than inline.

  34. Hi, Rob. This is an amazing amount of usable information. I have a couple questions. I have long wondered about the feasibility of a dewer flask on a large scale for thermal storage, even a DIY tank. Could a polished metal or plastic tank wrapped with MLI insulation (space suit material) be enclosed within another tank and evacuate the void? Would that be a viable way to store heat long-term? A 1,000 gal. tank should store 62,500 btus per degree. Any idea on how to test the idea? The other thing I noticed from this article was the shallow-water bath. That seems like an easier system for an insulated system like your solar attic. How much more energy is collected by a bath setup versus the coroplast collector? Thanks for sharing knowledge, Ian.

  35. I wonder if there is a even lower cost method with even greater efficiency.

    Has anybody tried corex? – its plastic corigated cardboard. Its only £7 for an 8’x4′ sheet. Its light and easy to cut and handle. weight per mm surface area in contact with the water and surface area per m2 would be good as well.

    Might break down with UV but could be coated and its available in black.

    The top and bottom tubes might need some work.

  36. Your collector is doing a lot better better than 38%. After the system reached about 37 C you started losing a higher percentage out the front window and maybe evaporation from your open cooler. Your real efficiency below 37 C was closer to 4.18*20L*30C/3,000s = 836 W ~ 60%. Judging from the shadows, the sun was vertical enough to get close to the 1,000 W/m^2 as you stated.

    A polystyrene or polycarbonate cover reflects 5% from it’s front surface and 5% from the back for a 10% loss. Glass, polypropylene, acrylic, and plexiglass reflects about 8% total from front and back. Thick plastics like plexiglass are also going to absorb a bit of the infrared. A 2-pane window is only needed for fairly cold places because it also causes a penalty of 8% or 10% from reflection. The front surface of the corrugated plastic would lose another 4% from reflection if it is clear and the liquid is black.

    A decently insulated system like yours will lose about 10% from heat escaping. So 75% is easy, and it’s hard to do much better no matter how professional the design. The objective is to transfer heat to the fluid without raising the air temp between the corrugated plastic and the front window (your polystyrene). Higher temps in the air gap mean higher energy loss out the front. Your plastic method doesn’t lose the additional 4% from reflection that occurs in a black water design, but the front side of the black-painted plastic is heating up the air gap a bit more than ideal which is why you’re getting 60% instead of 75%. You could probably get up to 70% with your method.

    If you make your collector 4×8 and lay it long along the ground tilted back 30 degrees from the vertical and then lay one 4×8 insulation board on the ground in front of it with the reflective side up, you’ll get 30% more light for 3 or 4 hours/day. Add another insulation board at the top (tilted 30 degrees forward from the vertical) and you’ll get another 30%, bringing your “efficiency” to 1.3*1.3*0.60 = 101%. It’s basically a trough design. These 30 degree angles are sufficient for all of the U.S. (the reflection boards reduce losses from not being at an ideal angle). The longer the better to get the early morning and late afternoon. If you add a car radiator on the inside of the house with one 4×8 collector and two reflectors, you should easily transfer 6 m^2 of sunlight to the house (4 kW average for 5 hours in the winter, equal to nearly three 1,500 electrical heaters, enough for 1,000 sq ft).

    You do not want an air gap between the surface of the water in a black water design and the corrugated plastic (or whatever you use to contain the water) because then you’ll lose 8% more reflection: 4% from the inside of the plastic and 4% from the water surface, black or not. When the water touches the plastic it’s like a “coupling”. It’s the transition from air to plastic/water/glass AND vice versa that causes the reflection loss, based on a difference in the index of refraction (1.0 for air and ~1.5 for plastic, water, and glass).

  37. My efficiency calculation for his system uses some complex reasoning due to his inlet temperature measurement not being a direct way of calculating efficiency. But he reports the initial outlet temp to be 27.7 degrees hotter than the inlet and and that flow rate was 0.8 L/min. This gives a direct efficiency calculation:

    4.18 * 27.7 C *0.8 L/min * 1,000 mL/L * 1/60 min/sec = 1,544 W per 1.4 m^2. This is 110% which shows his estimated flow rate is too high. It appears he was just watching the temp change to estimate the turn-over in tank volume which would give too high a flow rate because the water is mixing. The inlet temp would get hotter before tank volume had actually turned over. By my estimate of efficiency, his flow rate at the beginning was actually about 0.44 L/min.

    I saw in a different forum that there is some concern about the plastic aging way too fast when it is touching water. This idea could use a solar pool heater plastic piping ($160 for 80 ft^2 on amazon) and this type of insulation backing, frame, and polystyrene front cover.

  38. Congratulations on a very novel way of building a solar heated water panel.
    I want to try something similar to yours, using a similar framing technique
    from perhaps upvc material and some sheets of glass I have saved from some old sliding doors.
    I want the circuit to be as simple as possible so that the heated water feeds into our existing hot water cylinder,
    currently heated by immersion element.
    I aim to hang the panel on an outside west facing wall because that is where the sun is strongest from about noon onwards.
    The problem will be circulation of the water -unless I use an indirect cylinder which I could, so that the solar heated water simply feeds the coil in the cylinder and heats water fed from a tank.
    In the winter time I could drain the system down to avoid freezing and cracking of the glass.
    What do you see as the immediate problems?

  39. Hi John,
    I’m not sure of the exact setup you’re proposing but it sounds like you may be considering a heat exchanger external to your existing tank. If that is the case you should consider something like a Solar Wand (do a google search). This is a heat exchanger that can be inserted into your existing hot water tank and would likely be more effective and probably cheaper than an external heat exchanger which would require two pumps… one to circulate water from the tank and one to circulate water from the solar collectors. Good luck.

  40. Thank you for a great webpage, idea and explanation. You’ve done a good deed for society!

  41. How about using a truck radiator as a solar heater, with as much of the fins shaved off the core? It would be painted flat black of course. Thanks.

  42. Brilliant idea! I am planning to build a green house with aquaponic system. I have been trying to find a way to heat the water for the fish. I could use your method on the roof of the greenhouse as a multipurpose.

  43. Hi,
    First I would like to thank you for a great idea. I made water heating solar panel by using polycarbonate and ABS tubing (version 2.0).
    The main problem I had was how to make something that I can use every day and that does not leak. In first version I used some kind of cement but after few days it started to leaking. Then after few modifications I realized that I should build it from scratch once again. I bought good cement that can connect ABS tube and polycarbonate with good flexural. And after two months of using it (in previous summer) solar panel was working really good. Water in my swimming pool (7m3) get around 30C after three days of sunny weather.
    Now I’m planing to build something that is bit biggger and can warm up watter bit faster.

  44. Hello Rob
    I was thrilled to find your solar water heater instructions, at my current budget! Some Q’s:
    1.Can I tilt the panel sideways and get the same results?
    2.Do the cloroplast channels have to be vertical?
    3.For esthetic reasons I would prefer to have a 3’x8’ panel on the ground, with the 8’length in contact with the ground. If I cut the coroplast so the channels are vertical, would a 3′ high path of water flow be enough for the thermosyphoning to take full effect?
    4. I am heating the water to channel it into my radiant flooring for a small room. Where the cold comes out I will add a small bucket buried into the ground for the eater to spill into and to relieve pressure from the systems, then I will add an out hose followed by an el sid solar pump (i happen to have one with a 15Watt solar PV panel), to pump the water up the panel. (I have a diagram but I don’t see where I could attach it)
    5. I live where we have 4 months of below freezing weather at night, and 30 F warmer during day, in very sunny Santa Fe, NM. Would it be better if I added some prpilene glycol to the water mix? At what proportion? Will glycol eat up the panel?
    6. Alternatively, I could add to the bucket one of those electric sticks that heats the water in the bucket, specially at night, what do you think? Thanks a million and keep up your ingenuity!

  45. Has anybody tried using double-glazed window instead of Coroplast?
    The basic idea is to drill holes in top and bottom of a double-glazed window, add two nippels and connect it to the tank to let it thermo-siphoning. Ill add black dye to the water instead of painting the glass.
    Ill just leave it in the original window frame and use a similar double-glazed window as front cover.

  46. The coroplast design is clearly the best, because it is low cost and massively parallel (not like spiral black tube). For us, the difficulty is assembling tubes with coroplast and using high cost glue.
    After a lot of try (18 month of work), we finally find a design easy to make without glue or complicated tools. We will publish it soon. You can stay tuned at
    We developped a tester based on raspberry Pi and we measured more than 650W/sqm on sunny days!

Leave a Reply

Your email address will not be published. Required fields are marked *