DIY solar water heater, my own combination design

[spiral_72]

Hello! I've been looking for a place to ask opinions on an idea. This forum come recommended.

I live in SC. We have three hot humid seasons typically, the winters are generally mild with 1-2 light snow falls each year but the temp usually drops to low to mid 20's in winter. What else? I'm on well water pumped by a 3/4Hp pump through a water softener. All the pipe is PVC / CPVC except the bladder tank, switch and gauge which are all on about a 3ft glavanized section. Not sure if you'll care about all that.

I am new to this, but have wanted to try building a solar water heater for some time. I did quite a lot of research, some drawings and a schematic for what you might call a flat panel / evacuated tube combination design.


The (my) idea started with a video from a young man in Brazil? that built his own flat panel:

Which led to a DIY evacuated tube (cheap - proof of concept) video:

I'd like to use 1/2" copper pipe in 4ft section, painted black, inserted in a glass evacuated tube. All DIY stuff. The glass will be cut and cleaned 4ft T8 bulbs which have gone bad and been replaced. I plan to use a resin compound to seal each end with about 2" of the copper out the top and bottom. Assembly will be the typical parallel tube flat panel thermosiphon design. the bottom plug on each tube will have a 1/8" pneumatic line cast into the plug, all leading to a manifold to draw vacuum and maintain vacuum if needed.

I'd like opinions please. The thought is the vacuum is an excellent insulator (obviously) which will contain the heat and boost efficiency a LOT?. The other option is a typical flat panel with glass face and evacuate the air in the box, but I don't think that will work very well.

From there, I'd like to put an electric shorty 48gal water heater in the attic over a wall with power to supplement the solar. The collector will be roof mounted about the same level as the electric water heater, which SHOULD eliminate the need for a pump. The plan is, the collector heats the water all day in the tank, then we use the water in the evening.

The prototype will likely be 2-3 rows of CPVC inside evacuated glass tube heating a small sealed 1gal tank. If all goes well, I'll boil water and melt CPVC then build a small 2x2ft panel for the hot tub, work the bugs out, then one for the house.

 

[spiral_72]

If a picture is needed, the cyan color is T12 glass tube, green is 1/2" CPVC, yellow is the resin plug.


it's not annotated but the first drawing of the system. The top valve is an air purge at the highest point in the system. It will probably be a solenoid or something with 1-2ft of vertical collection pipe.


Here's a link describing one idea for cutting Fl tubes (NOT with a hacksaw as the link suggests): http://www.instructables.com/answers/Can-you-cut-glass-with-a-hacksaw-blade-/
They already had the argument about phosphorus, mercury, death and dismemberment, etc.

 

[Sponsor]

 

[Dana]

The relative ruggedness of T8 or T12 tubing is pretty sketchy. A hailstorm or minor earthquake could take it out.

Stagnation temps @ 110F outdoor temps in early & mid summer is likely to exceed the operating temp of CPVC: Schedule 40 CPVC need to be derated to 20psi @ 140F, 10psi @ 180F. This means you can't really run it at typical household water pressures in this application. Even schedule-80 would have issues at 150F+.

Reliability of the vacuum seal is critical, and it's not clear how that is going to be accomplished over temp with the very different coefficients of expansion of copper vs. soda-glass. Commercial variants that are not heat-pipe based use a U-pipe of copper in a sealed evacuated un-coated dewar, without a copper-> glass penetration. The U-pipe is in air at atmospheric pressure, but that air is not allowed to circulate freely at the open end of the dewar, but the dewar itself is a fully sealed glass construction.

See if you can't come up with a DIY methodology on that approach, where the vacuum seal won't fail quickly. You'll have to also figure out a configuration for making a U-pipe collector that can thermosiphon too, but that's not impossible. Still, the entire collector to tank loop needs to be hard-copper, and use a thermostatic mixing valve at the output to bring it under 125F at the connection to any CPVC, unless you like the notion of taking scalding-hot showers dripping through the ceiling on a midsummer day.

 

[spiral_72]

Wow, thank you for the information! It'll take me a bit to digest what you said as I'm not familiar with the terms.


The CPVC was a proof of concept "Prototype" as I called it, but I fully expect (at least I hope) you are right and the CPVC will melt. I'll have to watch the prototype closely. The final, if the idea works would be copper.


Interesting. I have yet to see this u-shaped copper in glass idea you have pictures of. What is this system called? Obviously my idea is not a new one huh?

"very different coefficients of expansion of copper vs. soda-glass"

'll have to look up the data, but the copper should not come in contact with the glass. I hope to have a resin plug sealing the pieces together. The thermal expansion is a very good point however and needs to be looked at.​

"without a copper-> glass penetration"

OH! so the open end of the glass seals to the black manifold in your picture as does the copper. What a good idea! I'll have to consider this.​

 

[spiral_72]

"a Dewar flask is a glass or metal bottle, with a double-layer construction; in fact, it can be considered to be two thin-walled bottles nested one inside the other, and sealed together at the neck. The narrow space between is evacuated almost entirely of air; the near vacuum minimizes convection of heat."

Hmmm, specialized parts eh? That makes sense then about the copper being at 1atm. The Dewar flask does not seal a vacuum inside the glass with the copper, it contains the vacuum within it's walls. In this design the aluminum fin absorbs heat and bakes the copper / water, all insulated from thermal loss.

Not a very DIY system with those components

 

[Dana]

The U-pipe approach is definitely not super-DIY friendly, but maybe not impossible. I expect it's an easier problem to solve than the vacuum seal problem.

With different coefficients of expansion the length of the copper and glass would change relative to one another, which means the resin seal has to be quite flexible over some flex distance and thousands of cycles, yet still hold a vacuum, which is not the simplest of mechanical engineering problems to solve. (That's probably why the commercial units use the U-pipe-in-dewar approach rather than your initial approach.)

Heat-pipe type evacuated tubes are a different beast, but also use a dewar with plugged-end approach:

The phase-change material is fully contained within it's own isolated (usually copper) tube, the dewar is at very low Torr, but the space between the heat pipe and the inner surface of the dewar isn't evacuated, and the plug through which the heat pipe runs is somewhat mechanically-complaint, absorbing most of the temperature related dimensional shifts without inducing a large mechanical strain on the glass.

 

[spiral_72]

Alright, thank you Dana. Let me do some figuring. Obviously the only benefit of the system I propose is the insulating vacuum. If I lose vacuum, I might as well build a flat plate.

If it works, it'll work well. You've given me some real things to think about. Thank you for not blasting me about mercury poisoning inside the T12 tubes.

 

[spiral_72]

I've been toying around with the thermosiphon. Honestly it took a lot of figuring to get it working, but work it does

56F water temp at 10am up to 122F at about 3pm heating 5gal of water. It's pretty cool watching this thing pump heat through the system without electricity.
Here's my budget collector built for less than $5 with recycled / leftover parts:


and a very poor cell phone video of the hot water output:


Just thought I'd share since y'all were nice enough to help out!