MAILBOX: PV for Hot Water


PV has gotten cheap and is easy to install, providing opportunity to defect hot water heating from the electrical grid or from propane or heating oil. Electric resistance water heaters can be powered by direct current, straight from a PV array, with no batteries, no inverter, and no grid-tie. The PV is connected to the bottom element, with the top element still connected to the electric grid as a backup.

At one time, solar thermal was a tenth of the cost of PV. At about 50 cents per watt, solar thermal equipment is still cheaper than PV, but often not by enough to offset its higher installation and storage costs. A PV-based domestic hot water (DHW) system using a large electric water heater for storage, and a resistance element as the “heat exchanger” can be cheaper for homes with minimal hot water needs, and very much a DIY project. You can even use the water heater you already have. I use the 50-gallon unit that has been here since 2007.

Another benefit of using PV for DHW is that the distance from the array to storage isn’t nearly as critical as it is with solar thermal. That’s because electrical line loss can be lower than thermal loss and the cost of wire is so much less than plumbing plus insulation. With my PV array, the distance was 240 feet, but it cost just over $250 for the #8 cables to keep line loss under 2%.

For homes with more hot water needs, solar thermal may still be a less expensive option. A solar thermal system will collect more energy per collector area than a PV array. For modest hot water needs, such as for one or two people in a single-family home, using PV for hot water makes sense. Using PV for DHW also presents an opportunity for backup power, since the PV electricity could in a pinch be redirected to DC loads or to a battery-inverter system.

Pete Gruendeman • La Crosse, Wisconsin

Comments (12)

Graig Pearen's picture

Hi Peter
Yes, I read the whole thing including all the comments and I agree with the original article. In fact, if I were doing it today, I would just install more PV modules and run a standard 240v water heater off my inverter.

I was commenting on the comments where people seem to think that flat plate collectors are appropriate for northern climates. They are not. I'm at 53.5 degrees north latitude where during the winter solstice, the sun is only 12 degrees above the horizon. My Apricus system dramatically out performs the flat plate collectors in this area. During that cold period mentioned in my previous post, the circulator on a friend's flat plate system only ran for about 10 minutes during the 3-week period while mine provided 100% of our hot water.

The snow/frost problem is mostly a myth. If the collector is mounted at an appropriate winter angle for the latitude (68 degrees in my case) the snow slides or blows off. With the 68 degree vertical tilt and the curvature of the tubes, not much sticks to them! Installing them at the optimum winter angle maximizes winter production and prevents over heating in the summer.

As for frost build-up, it is extremely rare because at this latitude, the relative humidity is very low in the winter. It is rare to loose a day's production due to frost or snow. One also has to look at the cloud cover. Yes, I've had snow on the collector but since the cloud cover was thick, it didn't matter.

By the way, that book that everyone thinks is the SDHW 'bible' (I have a copy) was written by a guy who has spent his career in the southern US so his comments about evacuated tubes are off the mark for high latitudes.

PV rocks!

Peter Gruendeman_2's picture

Graig: You did not mention your latitude. It's true that evacuated tubes lose less heat to ambient than flat plate at colder temperatures and poor sky conditions. Rick's point was that they are so far north at Fairbanks, with the sun being 8 degrees above the horizon and its light being attenuated by effectively 7 atmospheres, that there isn't enough irradiance or insolation to heat up the flat plate panels even at -18C, which is not cold for flat plate collectors farther south.
There still is the issue of snow and frost accumulation on evac tubes. Since they are better insulated they are not as effective in melting their way out, compared to flat plate solar thermal and PV.
And there is the higher installed cost. The point of this thread was that on small systems, the high installed cost of solar thermal and required specialized equipment such as coil-in-tank storage tanks leaves solar thermal at a cost disadvantage on small/ single family home sized systems. I need to emphasize that-- single family home sized systems. Solar thermal can not provide backup power and in some cases, because of heat damaged glycol, actually presents a power failure liability so my choice to go with PV-->DHW was pretty easy.

Graig Pearen's picture

IF you use the correct collectors, solar thermal works well at any temperature, including -30. Flat plate collectors are not the correct technology for cold climates. You need evacuated tube collectors such as the Apricus brand. A few years ago, in February we had 3-weeks of sunny days with highs of -20C (zero on the antique scale) (I don't use the F-word). During that time, our Apricus collectors provided us with 100% of our DHW needs.

Peter Gruendeman_2's picture

Hi Rick:
Wow!, you are way up there in latitude. Now I get your comments on solar thermal and PV being so far off spec in performance.
Towards your goal of achieving C/10 charging under any sky conditions: Here (44N), with overcast skies, ~50-100 W/m2, I have found that my 1.2kW array produces about 10 watts when tipped up 70 degrees from horizontal and 25+ when tilted to nearly horizontal under the same sky conditions. It's tough getting power from PV under heavy overcast, but you might do better if you can easily tilt your proposed array so they can see the overcast sky and not dead grass on the ground. Do you have a pyranometer to test for optimum panel orientation under varying sky conditions? Can you find a way to easily tilt your array, day to day, week to week and not just seasonally? There is a big difference between trivial output and 2.5 times trivial.
My own rule on seasonal tilting of my array is to not tilt to the winter position until after there is a persistent blanket of snow on the ground. We get lots of overcast skies between Nov 10th and New Years Day, and with no snow on the ground, I am better off with the array pointed up to the gray skies. The first significant snowfall is forecast for today- Dec 28th. After that I'll tilt to the winter position, because sunny or cloudy, the sunlight reflecting off the snow will be helpful.

Rick Zuber's picture

Peter, I suspect with a large array tilting based on cloud cover would be a lot of work. I have thought about getting a manual hydraulic pump and several hydraulic cylinders to tilt the array though. That way tilt could be changed throughout the day with a single pump. Maybe some day.

Dan White's picture

Merry Christmas. Thanks for the info on the A-Frame. If you have some pictures of the construction before you installed the panels I would appreciate seeing them. Please send them to me direct at My panels are much larger 305 watt but the principal should be the same just scaled up. Thanks, Dan

Peter Gruendeman_2's picture

Define northern climates and define useless. I live at 44 degrees north latitude and here 0F degree temperatures are not a problem. -15F and windy is a little tougher. The key to solar thermal is to buy quality panels with low-iron glass and selective absorption coated collector elements. And use the harvested heat at as cool a temperature as is practical. Don't try to push that heat into storage at 160F when you could be using it to warm your space with it at 100F. Homemade panels, with twin-wall polycarbonate glazing, flat black paint on a paper thin aluminum collector element and no snow on the ground to reflect extra light onto the collector definitely struggle at 0F, especially when trying to push heat into storage. Use the heat as you produce it.

> Where a solar array will still produce 60 ~ 70% of it's rated output even on the winter solstice.<
Very wrong. PV produces MORE power in the cold weather than it does in the summer. Add in reflected sunlight from snow on the ground and you'll find that PV will produce 100-110% of nameplate power on those cold but sunny days.

The advantage PV has when it comes to domestic hot water is mostly due to its lower installed cost. This does Not scale up to be a cost or space effective solution for heating a whole house or for providing large scale DHW-- hotels for example. Depending on ambient temperature, PV takes 3-4 times as much area for the same power production. In terms of dollars it's a good deal. In terms of space, PV requires a lot, which some sites might not have.

Rick Zuber's picture

The Fairbanks branch of the University of Alaska has done a lot of research on solar thermal. They found that when temperatures drop below zero they do not produce in these northern climates. In Fairbanks at ~ 65 N. latitude when the temperatures drop to -20 ~ -50F the sun is so low in the sky that there is no usable heat from it. Same with here at 60.5 N. latitude on the Kenai Peninsula.
In December the sun is at around 8 degrees above the horizon and produces very little warmth.
Solar PV at this latitude produces about 60 ~ 70% of rated output. That is not a guess that is an empirical measurement that I have made myself. With the sun so low and traveling through so much atmosphere to get here the only reason they produce that much power is because of the subzero temperatures that typically accompany a clear day and the snow on the ground.
In late winter / early spring I have seen a 1KW Solarworld array produce 150% of rated power because it still remains quite cold and there is still snow and the sun is getting higher in the sky.
As far as it's usefulness for space heating, I have yet to determine that. I am adding 3.75 KW to my off grid 1.5 KW array right now. My goal is 20 KW. With that much power I should have enough solar to charge our battery bank at a C/10 rate most of the year even when there is significant cloud cover.

Rick Zuber's picture

In northern climates solar thermal is useless when temperatures drop to or below zero. Where a solar array will still produce 60 ~ 70% of it's rated output even on the winter solstice. I am planning on heating water and circulating it throughout my log / straw bale home for space heating. I suspect it will cut down on our wood consumption considerably on sunny days. Those are the days in the dead of winter that are the coldest.

Peter Gruendeman_2's picture

I built this A-frame seasonally tiltable rack from stock. The "A"s were made from 3" x 3/16" wall square steel tube, and the tiltable frame made from 3" x 1/4" wall square aluminum tube. A better plan is to make the horizontal parts of the tiltable from from aluminum I-beams. Their strength to cost ratio is better and it would be easier to attach the panels to them. And it will be easier to construct with overlapping corners instead of the butted corners + reinforecments that I used. If one has access to a metals dealer who will sell cut to length stock, the really adventurous DIY-er could site build this with a Sawz-All and an electric drill.

In regards to efficient or not efficient use of 112 square feet of sunshine, point conceded. This is not an efficient use of sunshine. Let's look at dollars. That's how cost is assessed in a capitalist economy. My array is 1.2kW nameplate, which costs $1,200 in late 2015 for made in USA single crystal silicon based panels; less if one bought imports. What $1,200 will Not buy is a heat pump water heater + grid-tie inverter to run it + monthly connect fees if you even live in a state that offers net metering. 13 of the United States do not have net metering. Wisconsin has net metering and also has cold winters. To reduce my water heating bill by pumping heat from my conditioned space to my water heater tank is not helpful, at least not in the winter. $1,200 will also not buy a coil-in-tank solar DHW tank + circulating pump + controller to work with the solar thermal space heating system that I installed six years ago. $1,200 will certainly not purchase a wood fired boiler + storage tank.

The article that HP has not yet published about this system points out that it has the dual purpose of DHW + backup power. Providing DHW pays for the system. Backup power is not a feature of solar thermal based DHW, nor is being up and running for $1,200 of panels + rack + cheapo electric water heater.
Pete Gruendeman's picture

All well and good if you are willing to waste energy generated by using the most inefficient method of creating hot water. All things considered it is better to put in a wood fired boiler if cost is your only concern. For efficient use of solar it is much better to use a heat pump style water heater. The system produces multiple times more hot water per KWh compared to electric resistance heaters. The reduction in the cost of the solar system needed to produce the same amount of hot water from resistance to heat pump can likely come close to paying for the increased cost of the heat pump water heater. You need to consider the whole system efficiency, not just how cheap a resistance water heater is.

Dan White's picture

This is the second article showing this A frame tilt mount. Where did it come from? And what is the price?

Dan White
Cat Spring, TX

Show or Hide All Comments