Building a Parabolic Solar Hot-Water Heater Using 123D




Introduction: Building a Parabolic Solar Hot-Water Heater Using 123D

About: I'm passionate about combining engineering and sustainability through DIY technology. I'm a mechanical engineering student at Olin College and currently building things at TechShop in San Francisco as an int...
For this project I set out to create the solar equivalent of the hot-water tap on a coffee machine: a solar on-demand hot-water heater.  I was inspired by the functionality of a new software called 123D-Make that makes it easy to build large, geometrically precise forms.  My goal was to build a device using 123D-Make that illustrates the power of the sun and has many practical uses.  I decided to make my solar water heater by creating a mirrored parabolic dish that focuses sunlight to a point, then add copper tubing that runs a thin stream of water through the very hot focal point, creating near-boiling water on-demand.  By creating a way to boil water without electricity or gas, this project will hopefully lead to less Carbon Dioxide emissions and be a net-positive for the environment

Calculator (or spreadsheet)
123D (free CAD software by Autodesk)
123D Make (free slicing software by Autodesk)
CorelDraw or Inkscape
Laser Cutter (a shopbot or waterjet would also work)  I used a 120W Epilog Legend 36 EXT with a 36"x24" cutting bed
Utility Knife
Hot Glue Gun
Spackle Spreader
Pipe cutter
Wide Masking Tape
Sandpaper (Coarse & Fine)
Paint (to use as a sealant) like Acrylic base or Polyurethane
Stick-On Mirrored film
Thin copper tubing
1/2 cup of table salt
High-temperature black stove paint
1/2 inch metal conduit pipe
Rubber tubing. Use high-temp stuff if you plan to use it for a long time.

Step 1: Is It Possible?

My first step was to do a calculation to figure out if my idea was crazy.  I started with my goal:  Make the solar equivalent of the hot-water tap on a coffee machine.  
How much power do I need?
Lets say I want to fill up my mug with hot water in about 15 seconds.  That means I'll need to do about 1 liter per minute starting at room temperature (~20°C) and ending at tea temperature (~80°C).  That's a 60°C temperature change, and water has a heat capacity of 4.18 kilojoules per kilogram °C.  So, for 1 liter (one kg.) we need 250.8 kilojoules.  To do that in 60 seconds, we need about 4.2 kilowatts of power.  That's a lot!

How big of a mirror is that?
In most places, on a clear sunny day, there's about .8 kilowatts of sunlight per square meter.  That means that we would need 5.25 square-meters of mirror, which would be a dish 2.6 meters (8.5 feet) across!

That was a lot bigger than I could make, but it gave me an idea of scale.  For example, if I could wait 1 minute instead of 15 seconds for my cup of water, I only needed a 4-foot dish.  In the end, I decided to make a 36" dish because that was the biggest I could make on my laser cutter without slicing up my support rings into smaller pieces (which is certainly an option.)


Step 2: Designing the Shape

My second step was to define the shape of my dish.  The general shape is called a parabola (or technically a parabaloid because it is a 3D surface created by rotating a parabola about its axis of symmetry).  Parabolas are shaped such that they reflect rays (in our case, sunlight) to a central focal point.  They can be described (and graphed) using a mathematical equation:  
y = x2/(4f)  Where f is the height of the focal point.  For example, I could make a parabola with a radius of 2 feet and a focal point height of 1 foot by using the equation y= x2/4 and calculating from x=-2 to 2.  

Another way of thinking about your parabola is to use the equation FD = R2/4 where D is the depth of the dish from the center to the rim and R is the radius of the rim.  

Using my equation (I used y= x2/4 and calculating from x=-2 to 2),  I calculated about 20 (x,y) points between 0 and 2.  (I didn't worry about units, because I was planning on changing the scale later.)

Step 3: Creating the Shape in 123D

Once I had the set of points, I used 123D to make the 3 dimensional shape.  123D is a free, easy to learn CAD software available from Autodesk.

In 123D, I created a sketch and entered the coordinates of my points one at a time.  Then I connected them with a spline.  This is the curve that defines the inner surface of the dish.  I added some more lines to complete the cross-section of the dish, then revolved it 360 degrees about the center line.  

Step 4: Slicing the Shape in 123D-Make

From 123D, I exported my dish as an STL file and imported that into 123D Make.  This is another free software from Autodesk that you can download at or the Apple Mac-App-Store.  

123D Make lets you turn your 3d shape into a lattice of 2d profiles that you can cut out and fit together.  After importing my shape, scaled it to the biggest size I could make on my laser cutter and I chose a radial slicing pattern.  I chose the slice direction from the dish's center outwards (like spokes) and upwards like rings.  123D Make automatically calculates the profiles I needed for the lattice and adds slots so that every piece slides into the correct place.  

I started by making an 11-inch diameter dish out of cardboard as a prototype, then did a 36" diameter dish out of .25" plywood.  123D Make's nesting algorithm wastes material by putting too few segments on a given sheet. I wanted to lay out the pattern to use the most of my material, so after generating the laser sheets, I consolidated them in CorelDraw.  For the 36" dish, I had to cut the biggest rings in separate pieces, so I split them in CorelDraw before cutting.  (Looking back, I would have made the pieces of the ring interlock like a puzzle, adding more rigidity.)

Step 5: Laser-Cut and Assemble

I cut the parts on a 120W 24x36 Epilog laser.  For the 36" dish parts, it took about 3 hours to get everything cut out of .25" plywood.  It's very important that your wood is perfectly flat on the cutting bed, otherwise the laser won't be focused correctly and won't cut well.  I used tape to help hold mine flat, but keep flatness in mind when buying your material.  

Once everything was cut, it was time for assembly.  123D-Make does a neat animation of how your dish can be assembled.  For this job, I used a bit of hot glue to hold the first three supports in place, but after that, no glue was needed.  There was enough friction that the dish held itself together perfectly.

Step 6: Create the Surface

Now that I had the wooden lattice, I needed a smooth, continuous surface to serve as the inside of the dish.  I did this using a layer of masking tape, then several layers of spackle.  The first layer of spackle filled in the depressions in the tape.  The later layers I used a pre-cut wooden form, rotating it around to spread the spackle evenly to the correct curvature.   I used a laser-cut piece of wood glued to a piece of sand paper to sand the surface to the correct curvature. 

Once the spackle had the smooth, curved surface that I was looking for, I sealed the surface with three coats of clear acrylic paint. I considered epoxy and polyurethane, but decided on acrylic since it was the least toxic and had the lowest environmental impact.

Step 7: Make It Shiny!

My next step was to add an adhesive-backed mylar film to the surface of the dish.  I got mine online from Green Power Science (who also have some great youtube videos about making solar parabolas).  You can probably get the same stuff elsewhere for less money, but I knew that this stuff would work.  I was able to do my large dish with strips cut from one 24"x36" roll. 

I used a knife and a straight edge to cut strips about 1" wide.  Then laid them down, starting with a strip across the center and working up to the left and the right.  At the end, I trimmed off the extra around the rim and it was ready for the first test.

Step 8: Test 1: Light Things on Fire!

To test the parabola's performance, I brought it outside, pointed it at the sun, and lit things on fire.  I wanted to figure out how small of an area the dish was able to focus the light.  Wearing dark welding goggles, I held a piece of wood at the focal point (about 10 inches above the center) and could get the light to concentrate on a circle about 2 inches in diameter!  That means I'm getting over 250 times the power of the ambient sunlight!

Next, we tried roasting marshmallows.  Unfortunately, they hardly cooked at all.  They were so white that they reflected all the light that hit them.  When we rubbed some black charcoal on one and it started charring instantly when we put it in the beam. 

On a Safety note, always store your mirror covered or out of the sun so it will not burn your house down.  Try not to leave your mirror unattended in a sunny spot. 

Step 9: Add the Boiler

The next step is to add the boiler.  I decided to use a coil of copper pipe for my boiler, so it would function like an on-demand hot-water tap. I used 5 feet of 1/4 inch tubing.  I coiled it myself without any proper pipe-bending tools.  I used salt to fill the tube so it wouldn't crush and I coiled it around a series of smaller and smaller cylindrical objects until I got to the diameter I wanted.  Check out my instructable on bending copper tubing.

After completing the coil, I mounted it to a piece of scrap pipe going through the center.  The diameter pipe is about 1/2 an inch smaller than the inner diameter of the coil.  I filled the empty space with spackle, which will function as insulation for the inside of the coil. Finally, I painted the coil with a coating of black stove-paint.  It's paint that's designed to handle high temperatures and it's very important.  Without it, the coil would just reflect most of the energy we are beaming at it.

I drilled a whole through the center of the parabola and mounted a piece of pipe that will hold the boiler using some laser-cut disks of wood.  Two plastic hoses come up through the center to run water to and from the boiler. There's a needle valve to regulate the water flow.

Step 10: Make Some Tea!

I brought it outside and pointed it at the sun, using the shadow of the boiler as an aiming device. It gets really hot.  A black piece of metal gets up to 600 degrees F (however a white marshmallow barely gets warm).  It is able to make steam, and very slowly produce a stream of boiling water.  It's about a drop every second.  It can also cook Bacon!



    • Tiny Home Contest

      Tiny Home Contest
    • Water Contest

      Water Contest
    • Creative Misuse Contest

      Creative Misuse Contest

    82 Discussions

    Many thanks for sharing your project results, I have a question about the mirrors. which is better to use mirror pieces or applying a reflective coating on the parabolic dish?


    2 years ago

    I think if you winded the pipe between 2 metal plates (so that the pipe looks like a plate) or around a cone you would have had larger area of the pipe exposed to the sun rays and better heating coefficient. Well done!

    Well done !

    Love your work . I am currently working on a Satellite tv parabolic hot water heater myself. I will post my results here with a how-to blog....

    Very good. I have a satellite dish plate, old and unused.
    Can I use this your idea?
    I would like to see the video that you made, this time as this video is private. You can show me working?
    I could use the mirror in place of foil, what do you think of the idea?

    2 replies

    Indeed you could! Using satellite dishes as a parabolic reflector is common practice in creating high powered home built lights! The benefit is you have the backing of a million dollar fabrication method to ensure a more perfect surface - with his masking tape technique the efficiency of his design is unfortunately a lot less than the conceptual model on his computer.

    covering the coil with glass would help it retain heat,block wind.
    Adding some thermal mass around the coil would allow it to store a bit of heat, for passing clouds and the like. If you were to put the coil into a block of concrete say inside a mason jar,using a black pigment to color the concrete I think you would get a more even heating of your water. Good luck with the application,it was a nice instructable.

    2 replies

    A thermal mass was my first thought - leave a larger black mass in the focal area and let that heat the water rather than the copper pipe as the mass of the copper is tiny.
    Good project, though.

    A thermal mass was my first thought - leave a larger black mass in the focal area and let that heat the water rather than the copper pipe as the mass of the copper is tiny.
    Good project, though.

    re tracking: I remember an article fom Mother Earth News from back in the 70's tsome one had a dish he wanted to track the sun.. they had two ram actuators, plumbed to two flat black reservoirs. the reservoirs were mounted to rotate with the dish. the one on right had baffle shading it from sun rays from left, but could be warmerd by sun rays from right. similarly, the one on left was baffled to be shaded from right and warmed by sun rays from left. the reservoir on right was plumbed to actautor that extended to turn dish and reservoirs to right, and reservoir on left plumbed to actuator to extend to turn dish to left. thus when sun is to tp right of dish, iit warms reservoir on right and not the left, causing dish to rotate toward sun .dish stops when both reservoirs are warmed equally, and dish is pointing at sun. as sun moves, heating is unbalanced again, anf reultant difference in reservoir temps changes volume of fluids in each system, to cause actuators to work to rotate dish back into alifhnment with sun.... no electricity required.

    1 reply

    Nice work!!! Whot do you think about my heater?

    Using Copper pipe polish it then Just heat it up with a gas flame and surface will oxidise to copper oxide on air contact in seconds which is black saves having to paint!

    How come you did not simply use a discarded satellite antenna for this project? Would have saved a ton of time and calculations. The receivers are already mounted at the focus point and the bracket is already made to the correct length.

    1 reply

    This was going to be my question. Be really cool to get one of the big old ones that had a diameter closer to what your data showed you needed, then swap out the copper coil for a magnifying glass and make a death ray!

    I had an experience like you bt the biggest problem was the sun movement as you know after minuts you have to change the dish location .do you have any solution for this problem? Id like to know your idea .thanks