Solar Water Heater Still

I decided to make a solar hot water heater, but I also wanted it capable of ethanol distillation. There are a lot of ways to do this easily if you have access to a laser cutter or cnc router. However, I don't have access to those so I tried to think of a cheap, "simple" way to make a solar still.

To make "parabolic" reflectors I decided I would try using corrugated plastic (the stuff that looks like cardboard made out of plastic), and gluing on mylar foil to create a reflector. I bought a box of 10 solar vacuum tubes off ebay to use as the solar collectors. My original design called for 7, so i sold 3 to a coworker for a project he was interested in. Further on you can see i only have 6 in the design because i ended dropping and breaking one...

To mount the solar vacuum tubes I used oversized conduit/tube clamps drilled into one side of a open box wood frame made with four 18" pieces of scrap wood i had laying around. This gave me the relative height of the solar tube with reference to the box frame.

I figured the easiest way to make (approximately) a parabola would be to constrain the edges of a plastic sheet between two sides of a wooden box frame and let it bend down through the frame. The plastic acts a spring and holds itself securely against the edges of the wood. To get a good starting point for what length to make the piece of plastic I approximated the deflection of the plastic as a constant arc between the 18" span, and made ray tracing lines in an Autodesk Inventor sketch.

To make the reflectors rotate I used long (~6in) 3/8" cap screws secured to the box frame on each side by a locknut and washers. This does leave a bit of wiggle room which is undesirable, but it seems to be working well.

To sync the mirrors I designed a double round belt pulley to be 3d printed using a Formlabs printer I have access to at work. The pulleys use the hex pattern of the bolt head as a keyway of sorts, and is very secure once epoxied into place. The end pulley has a mxl timing belt pulley on the outer section to connect with a stepper motor, and a round belt pulley to connect to the other round belt pulleys.

Starting with the length of plastic I estimated using Inventor in the previous step, I pushed the plastic pieces into the open wooden box frame and found a laser pointer. I placed a piece of mylar foil on the bent plastic and adjusted it until the laser dot would hit the tube at all points while moving the laser from the outside of the box towards the vaccum tube while making sure to kee the laser pointer perpendicular to the flat top portion of the wooden frame. Once I felt confident in the reflector shape I use a spray adhesive to glue a mylar foil onto the plastic sheet.

While a truly parabolic reflector creates a focal line, this improvised reflector is non-imaging meaning there is no true focal point/line. for the most part this is irrelevant because the target is a 2 1/2" diameter tube, which allows for some migration of the reflection lines from a true focus.

To seal the vacuum tubes I used high-temp silicone stoppers which are rated to 500*. I originally was going to use two straight brass fittings per tube (in/out), but changed it to a T fitting. The straight fittings would have worked fine, but the T fittings save a little more space when I had to actually connect up all the tubing. I used 3/8" EPDM tubing which is resistant to UV,safe for drinking water, and also fairly cheap. The idea is to use natural convection with the water tank at a higher elevation than the reflectors hooked up in series.

I opted for round belt because its cheap, and i could 3d print pulleys for free at work (within reason). I bought 1/4" belt in endless form. To splice the belt I made a little jig consisting of two plates with shaft holders screwed into them. Those plates had two bolts going through them, and the nuts are tightened to apply pressure to the two cut ends of the belt. once the belt ends are clamped together with a little bit of pressure i melted them together slowly with a small torch, and used my fingers to sort of press the belt together around the diameter to keep it from deforming.

I was a little worried the round belt might not work because its so flexible, and isnt meant for motion control like chain or timing belts. Turns out after all was said and done the round belts didn't really work for moving one reflector and having the others follow. This problem could be made better by using larger pulleys but that would still not soldve the problem. There is a rubberband effect, due to the inherent properties of urethane. Moving forward I'm going to have to use timing chain, timing belt, or could use a bunch of individual stepper motors with a timing belt for each reflector. so for now its manually adjusted.

To mount the individual reflectors to a pivot, i used 2x4s with plastic sleeve bearings pressed into them in 24in intervals. Because these measurements were done by hand they weren't perfect rotating joints by any means but they have worked quite well so far. Tubing was run in an over/under corkscrew of sorts around the frame. This came from messing with the tubing and seeing what orientation was least likely to kink the tube as the reflectors rotate. I added on support legs to tilt the frame at about a 33* degree angle . Living in Oregon, 45* is ideal for spring and 22* is ideal for summer, so i decided to split the difference. Little support legs were added on the bottom to protect the pulleys.

I haven't managed to get ahold of a spare water heater or insulated tank yet, so my testing has taken place using just the volume of the collectors and tubing (a little under 2 gallons). On a sunny day I was able to bring water to a boil by manually adjusting it here and there. It should work great for ethanol distillation, especially once I have it automatically tracking.