Solar Water heater

I am interested in getting some general guidance in making a solar water heater for my domestic residence using mostly all home-made components (i.e. Solar collector, piping but probably will purchase the heat-exchanger tank and other components beyond my ability and patience).

So I envision making a solar collector using simple, sweat-soldered copper piping, with this feeding a heat-exchanger tank, to pre-heat the potable water going to my conventional, (tank) heater. So the heat exchange fluid (distilled water or maybe a non-toxic fluid) is separate from the water being consumed.

Beyond having two temperature sensors (one at the collector, to measure when the collectors become warm enough and one at the exchange tank, to measure when the water in the exchanger needs to be heated, i can think of a few other circumstances that I don’t know how to address. 1) How do I handle the circumstance when there is no usage (i.e. As when on vacation) and the hot water accumulates and the tank gets too hot. (I envision a collector that is ~eight ft by eight ft with one-half inch copper pipe that is sweat-soldered “serpentine fashion” over the entire 8x8 area). The heat exchanger would take the form of an old, jury-rigged 50 gallon tank with a home-made, coiled copper tubing exchanger inside. I suppose I could rig a simple overheat pump that would dump the overheated water outside, or maybe some kind of cover that could be moved in front of the solar collector if it got too hot. Any ideas are appreciated…I want this to be relatively simple but not unsafe.

Of course an Arduino Uno would be my controller and I intend to communicate with the sensors with a digital I/O because the wire runs to the sensors might get pretty long.

Any suggestions and constructive criticism is appreciated.

Try googling "Drainback" systems. They use a larger pump and second smaller storage tank at atmospheric pressure. There is a heat exchanger between this drainback tank and your domestic water tank. Its a bit more complicated but it will solve your overheating issues.
Most solar water heaters don't employ a serpentine design, but a manifold instead. This keeps your output temperatures a little more reasonable, and you'll pick up more btu's in the long run.

Hope this helps,
dan

tentoes1962:

  1. How do I handle the circumstance when there is no usage (i.e. As when on vacation) and the hot water accumulates and the tank gets too hot.

First question: can it overheat? Do you even have enough solar radiation available in a day to overheat the thing? What is overheating for you? How much solar power can you get? And more to the extreme: what would happen if the water in the pipes starts boiling? Pumps may fail, sensors may fail, your power may be out and your Arduino goes haywire. It shouldn't be too hard to make it fail safe.

If you want to stop it from heating up when you're on vacation, pull a light coloured tarp over it and maybe even drain the water. Do look up health issues with solar heaters: many bacteria love the intermediate temperatures of 30-50° - on a not so sunny day you often end up in this range. You'll need a backup heater to get to at least 60° or even 70° to kill them. The Legionella bacteria is an infamous one that flourishes in such just-not-quite-hot water and can make you really sick. It's also the main reason solar water heaters are not catching on as much as PV solar panels, in some jurisdictions solar water heaters are not even allowed.

Of course an Arduino Uno would be my controller and I intend to communicate with the sensors with a digital I/O because the wire runs to the sensors might get pretty long.

Long wires are bad for sensors. Anything on an I2C bus can't do more than a few meters (I forgot the specs, maybe 5-10 meters or so, tops). So you better get your Arduino close to the solar heater, and communicate with you over WiFi or so (for that, look at the ESP8266 boards as alternative, such as WeMOS and NodeMCU - got WiFi built in).

Can't I use coax cable on the long sensor runs?

I live in Texas so freezing isn't a big concern and we have a lot of sunshine (and humidity)!$&@!

We use thermostat wire with 10K thermistors

If you cant maintain 120+ in your DHW tank you can set up your system to cycle on the water heater once a week or so for bacteria kill.

My system at home uses the top electric element only to maintain the top half of the tank at 120. The bottom half of the tank gets heated only by my wood stove or solar. Someday I'll add another tank for more efficiency and storage. Taller tanks work better for this heat stratification effect.

I really want to use a separate, pre-heat tank for modularity (so I can change the size easily with shutoff/bypass valves and for maintenance).

Maybe I should just start with a "demo" system that's not connected just to see how much heat I can collect using the 8x8 collector. I am not thinking about smelting any iron any time soon.....:slight_smile:

tentoes1962:
Can't I use coax cable on the long sensor runs?

I live in Texas so freezing isn't a big concern and we have a lot of sunshine (and humidity)!$&@!

Maybe so, but I have experienced ice everywhere in Houston, all Texas airports closed due to ice and 32 degrees in Brownsville. An ounce of prevention is worth a pound of repairs.

Paul

You could find some insight here if you're patient enough to sit through the video
Take care tentoes

I don’t see any need for a heat exchanger. If you’re going to heat-up some other liquid you can just heat the water.

And, I don’t understand the concern with overheating. If you build the thing with metal and solder the joints it can take whatever heat you directly collect from the sun (it’s not going to get any hotter than your roof). That is, unless you are going to use mirrors to concentrate the heat… In that case you could overheat something, but I have a feeling too-much heat isn’t going to be your problem…

If you’re worried about scalding, I suppose that could be an issue on a very-hot day, but I’d deal with that by mixing-in cold water rather than by limiting the water temperature. (You’ll probably want to collect & hold as much heat as possible.)

Coax is for analog signals. You can get a bit more distance out of your I2C by using twisted pairs and lowering the bit rate but it's still not recommended. These protocols are simply not designed for long distances.

Would ds18b20 temp sensors be suitable? They use the 1-wire bus system which can work over much longer distances than i2c.

8x8 feet gives you a maximum heat input of about 6kW - that's quite a lot.

I built a solar heating system for a swimming pool a while back - the pool contained about 200 tons of water , and on a long summer day in England , 15 square meters of heating area would raise the pool by about 4C.

Why not dig a swimming pool as a heat dump?

Allan.

Came back to forum to check on some other posts I made earlier...
An 8X8' collector space is an appropriate size collector for a 100-120 gallon storage. Ballpark is 1.5-2 gallons storage / sq.ft. of collector space. This is for manufactured collectors with suitable glazing and heat transfer fins brazed to the manifold piping... DIY collectors no doubt will not be as efficient, so you can go with a smaller water tank. If your collectors have a glazing they can absolutely get hotter than the surface temp of your roof.
Not to the be the alarmist here... but be sure to allow for expansion when your water heats up. Don't let it boil in a closed loop or it'll burst with steam and scalding water... somewhere in the piping...
Most residential closed loop systems use a 75 psi safety valve and expansion tank. This kind of thing is unnecessary with an open pressure drainback tank.

Thank you to fieldsofdan for the recommendation of the drainback system. That design DOES solve a lot of problems that would otherwise add a lot of expense and complexity....