Compound Parabolic Evacuated Tube Heater Experiment - Part 1

I thought I'd share a couple pictures of my latest water heating experiment. These evacuated tubes can be found fairly cheap, but making a manifold for a whole bunch of them is complicated and expensive. With just two evacuated tubes a couple simple Tee joints will do. If you've never played with these heat tubes, they really are amazing. Just like a glass thermos they insulate the central part from outside, and they have special low emissivity coating as well to make sure that they trap all incident heat radiation. A full tube of cold water can come to a boil in about an hour just sitting in the sun. Even on cloudy days they can get quite warm. A well designed system can avoid a glycol heat exchanger due to the inherent freeze resistance, which makes them popular in higher latitudes.

I originally designed these compound parabolic ribs for a 6" focal plane and dual 3/4 copper tubes. The previous heat fin makes a fine support and rear reflector for the heat tubes.
Each side will get a 2'x2' polished aluminum reflector. I took the close up after I had finished priming all the bare wood, and I've already realized that I need to add a TPR valve before adding water. I'm going to use some eco friendly antifreeze in the tubes to help buffer and transfer heat to the copper tube. The black angle iron holding the dish in place in the picture will be moved and with a bracket like the one at top, attached to the middle of the axial pipe support where it will then become the mounting support for an actuator to make the dish track. I'll use a Track'n Park controller to be able to set the park position, and make sure that it's shielded from reflections off the dish. The compound parabolic design doesn't need the same tracking accuracy as an ordinary parabola. The important thing is just to maximize the light that it can gather, just like with PV panels.

Anyway I hope to get some good performance measurement data from this to see how well it actually works. It should produce close to 500W, which won't heat a tank very fast, but definitely deliver one nice hot tank per day.