Crazy idea of the week : photovoltaic DC hot water

So, photovoltaic panels are now soooo cheap. The raw costs are down to 50 cents a watt. The inverters to convert to AC run another 50 cents a watt.

So, I was looking into solar hot water. The most economical - and pretty effective - way to do it is you buy a new water heater to use as a storage tank. The dedicated solar storage tanks are sold at ripoff prices, and it is actually cheaper to just buy a new hot water heater from the hardware store than to buy a tank meant for solar storage. Like this.

So 900 bucks and you have a tank. Then, you need a way to heat the water. There’s a buncha ways, but the non-janky, high performance way is to use evacuated tubes. You also want that Federal government 30% rebate, the one that expires at the end of 2016.

Such as this. Advantage of evacuated tubes is that since there is a vacuum inside the tubes, the water inside can get extremely hot - boiling, even - without conducting heat back to the environment.

Plumbing is simple. You just remove the drain valve from the bottom of the hot water heater and replace it with a threaded T. One side of the T you put the drain valve back to, and the other side you connect to a pump. Water gets pumped out the bottom up a line to the input port on the rooftop array. I’d probably use half inch thick walled copper with an insulated jacket for this as it is easy to work with and solder (and the copper can tolerate hot water temperatures).

Water comes back down from the array, and you replace the pressure relief valve with another threaded T, so the relief valve is on one part of the T and the water input goes into the top of the tank. Hot water tends to stay in the top of the water tank, so this should work well with low pumping effort.

You need a pump for this, I’d use a low voltage pump because it removes the need for AC safety. They sell pumps that can run off a solar panel, I’d probably run it off of house power for reliability. You need a controller, it’s just a couple thermistors and a relay to control the pump. I know how to put this together with maybe $20 worth of parts so I’d just build my own controller on a tiny scrap of perfboard. Control logic is simply - if it is daytime, and the water from the array is more than 10-20 degrees hotter than the water in the tank, and the water in tank is not more than 170 degrees fahrenheit, leave pump on. If it’s daytime and the water in the tank needs to be heated, turn on the pump for about 30 seconds to find out if the input water is hotter than that threshold.

It’s about 10 lines of code anybody could bang out, and a couple inputs from digital 1-wire bus thermistors and a single output to a mosfet which drives the relay to turn the pump on and off.

Or I could pay for a ready made controller. Either way.

The hot water, which might be anywhere from the temperature of the input water to 170 fahrenheit, has to go to a mixing valve set to about 110 fahrenheit, which is a safe but warm temperature for showers and sinks. After the mixing valve it flows into an electric on demand heater set to 105 fahrenheit, so on days when there is inadequate solar, it heats the water up.

So, a roughed out design that would probably work. One little “glitch” that would need to be solved is what happens if the tank temperature is too high, and the pump is off, and there is still light shining on the solar panels. Maybe a rooftop pressure relief valve would fix this problem, or there are these tanks they make for this.

Anyways, the other way is I just buy some cheap 50 cent a watt photvoltaics, and I just run a wire to the AC heating element on the hot water heater I already have to buy to make this plan work. I install panels that peak at about 240 volts DC under load. PV wire between the ends of the panel circuit and the hot water heater heating elements. Same setup in that a tankless heater tops up the water temperature, same mixing valve.

This saves a lot of plumbing, and solar panels have greater lifespan than evacuated tubes, and are easier to fix than plumbing, etc.

The problem is impedance - with a fixed resistance as the load on the circuit, I won’t get maximum power from the panels. What do I have to do to solve this?

Use passive propylene-glycol solar collectors, like this. It’s vastly more efficient than trying to construct some Frankenstein system out of photovoltaic panels and microcontrollers.

Follow the money.

Heating water with PV is a fool’s game.

BTW, there is one case where a system like this makes sense, and that’s with Wind generators.
Those need a load to prevent overspin in high winds, so diverting to resistance heaters is SOP.

You would think. Roughing it out, it may not be actually true because the manufacturing cost of PV has dropped more than the cost for water heating tubes.

Are they 20% of the cost? Because, that’s about the difference in efficiency.

I don’t need propylene glycol because of evacuated tubes and I’m in a hot climate zone. Do you have any info on costs? I’m estimate about $70 + $80 + $900 + $400 + $1230 + 250 = $2930. That’s the complete system including an on demand tankless that means I shouldn’t run out of hot water, and including money budgeted for misc costs. Just a passive setup I looked at was about $2000 without the other stuff.