A Smarter Solar Water Heater

Installing solar power at a home is a great way to reduce electricity bills, especially as the cost of solar panels and their associated electronics continue to plummet. Not every utility allows selling solar back to the grid, though, so if you’re like [Rogan] who lives in South Africa you’ll need to come up with some clever tricks to use the solar energy each day while it’s available to keep from wasting any. He’s devised this system for his water heater that takes care of some of this excess incoming energy.

A normal water heater, at least one based on electric resistive heaters, attempts to maintain a small range of temperatures within the insulated tank. If the temperature drops due to use or loss to the environment, the heaters turn on to bring the temperature back up. This automation system does essentially the same thing, but allows a much wider range of temperatures depending on the time of day. Essentially, it allows the water heater to get much hotter during times when solar energy is available, and lets it drop to lower values before running the heater on utility electricity during times when it isn’t. Using a combination ESP32 and ATtiny to both control the heater and report its temperature, all that’s left is to program Home Assistant to get the new system to interact with the solar system’s battery charge state and available incoming solar energy.

While it’s an elegantly simple system that also affords ample hot water for morning showers, large efficiency gains like this can be low-hanging fruit to even more home energy savings than solar alone provides on paper. Effectively the water heater becomes another type of battery in [Rogan]’s home, capable of storing energy at least for the day in the form of hot water. There are a few other ways of storing excess renewable energy as well, although they might require more resources than are typically available at home.

79 thoughts on “A Smarter Solar Water Heater

      1. The downside is that there are a lot more moving parts, where something could go wrong. Not something you want Joe Public to have to deal with, unfortunately.

        As an example, I am currently fighting with the modbus interface to my inverter. It is managed via a USB-RS485 dongle attached to my OpenWrt router, and the last week, mbusd (modbus TCP to modbus RTU gateway) has had to be restarted twice, because it stopped being able to communicate with the inverter. So Home Assistant has been blind to what is really going on with the inverter, and probably making bad decisions during that period. I’m putting in measures to restart mbusd automatically should the readings get stale (and notify me), but ultimately, I probably need a new USB-RS485 dongle.

          1. Happens to be a Sunsynk 8kW single phase, but I don’t think it is the inverter that is at fault. I see error messages in the OpenWrt kernel log, which suggests that it might be a driver problem, or else hardware problem. As soon as I restart mbusd (which provides the interface between the kellerza/sunsynk addon and the serial port), everything comes back immediately. Well, after restarting the Sunsynk addon, too, although I haven’t tried to see whether it actually needs to be restarted or not.

        1. I tried that with a SMA sunny Island but there was some problem where it would stop allowing any non SMA products to view the values. Tried their support for this and other issues I had and long story short, I won’t buy them again.

        2. I agree that John Q Public shouldn’t be dealing with this. But what is frustrating is that this is exactly the kind of low-hanging fruit that should have been standard in all hot water heaters 50 yrs ago (there were microcontrollers and ss thermometers capable of doing exactly what was described in the article waaaay back in the 70s)

          So it’s somewhat depressing that even after half a century, it takes some enterprising Hacker on a DIY to save a few kWh

    1. Thermostatic mixing valves are required by law in many places: the water tank is required to be hotter than 50 C to prevent bacterial growth, and water at the outlet must be less than 50 C to prevent scalds. In practice here the tank is set to 60 C (which increases its reserve capacity too), and the mixing valve is set to 50 C. Appliances that want hotter water (e.g., dishwasher) can tap the hot water line ahead of the mixing valve.

      1. Temperatures below 60 are “conducive to Legionnaires growth”, apparently. My geyser goes as high as 70C at least once per week, which is considered more than adequate to prevent such colonies from forming.

        I was quite surprised that my shower doesn’t seem to need adjusting of the mixer to account for the hotter water when available. I wonder if there is a thermostatic mixing valve built in? The rest of the outlets do get a bit hot if you run only hot water, but pretty much the only place that that happens is the kitchen sink, and I’m the one doing the washing, so ….

      2. This is the situation in NZ, mixing valves mandatory and domestic HW has a min temp setpoint on the thermostat to prevent bacteria growth.
        We used our 300L hot water tank as a dump for excess solar, thermostat turned up to almost max. Meant in summer we had 300L of 85-90degC water by lunch time in the tank.

    2. Since you are adding and relying on smartness for your water heating, you can have it in your water usage as well: use a smart tempering valve in the outflow of the hot water heater to bring the discharge temperature down to 104 (F).

      I suppose you could add anti-scald valving in sinks, showers and tubs, but it costs more. And since you are working with plumbing and safety, it has to be surpassingly reliable.

  1. My power utility company does the same: They are building a “thermos flask” and “electric kettle” to save surplus (green) electricity as heat, which is used in the district heating system. The same system is also heated by excess data centre heat.
    However, this boiler is about 100MW (or about 44000 standard electric kettles).

    https://group.vattenfall.com/nl/newsroom/persbericht/2020/naast-thermosfles-ook-grote-waterkoker-in-diemen-voor-duurzame-warmte (🇳🇱)

    1. There’s also the sand battery, which can return some of the energy back as electricity as well as heat, because it’s hot enough to run a steam turbine.

      I wonder if it would be possible to make a small scale sand battery for home use. One cubic meter of sand holds 1.34 MJ/K so if you get it up to 600 C it goes for 204 kilowatt-hours all the way down to 50 C, which is the lowest useful temperature for hot water use. You get 10-20 kWh more if you calculate down to room temperature. One cubic meter of water only stores about 70 kWh between 40-70 C, so for the same size of a “boiler”, you’d get three times more storage capacity.

      The heat exchange is done using air and a fan into a stack of channels inside an insulated barrel of sand, so when you don’t need the energy you cut off the airflow and the heat can’t escape through any bridge of solid material. A radiator then picks up heat from the hot air and pipes it into the home central heating system, or into the hot water line etc.

      You could technically get it much hotter than 600 C without any problems, but the cost would go up because you’d need to make the container out of something more heat resistant than steel.

      1. For modest heating needs, like in France or UK, storing 200 kWh is actually quite a lot. A well insulated house could keep warm for a week with that. Usually with cold spells come dead calm and darkness, so wind nor solar power won’t be available for several days, but if you have a pile of almost glowing hot sand in your basement, who cares?

        The only issue is that the system cost should be extremely cheap, because gas is normally only about 5-6 cents a kWh and recently it’s gone up to about 10 cents thanks to a certain country. Still, it’s tough to beat that price with direct electric heating, which is typically more than double, and becomes highly volatile in the heating season when wind power comes and goes.

        People are already talking about switching back to fuel oil for heating, since it can be stockpiled for the entire winter, so they get to pick the price when it’s cheapest.

        1. Local energy store that holds enough to last you through the times the local weather isn’t good for your solar/wind etc and keeps your comfortable even if it costs a heap to install now isn’t a big problem – peace of mind you will always have energy, and known upfront costs you can budget for. Effectively prepaid free/heavily discounted energy for the rest of your life, and if the system is as durable as it aught to be quite possibly your kid’s lives too. It can be expensive, really expensive and still worth it, at least for those that can afford it. Which in the UK at least is probably the folks too old to actually see a return on the investment in what is left of their life if its expensive…

          With any fossil fuel you have to ask if it will remain available and cheap long term – gas might be excessively insanely cheap now, even with the war going on disrupting the flow a bit it is still by far the cheapest way to get energy here. But eventually the supply is going to dwindle, or worse economic if not full on warfare conflicts will limit the export/import markets for a while, plus the subsidies and economic inertia of old subsidies will fade as nations go more and more towards local green grids. All of which means the usually available and cheap fuels of old may cease to be either of those things.

          1. >energy for the rest of your life,

            Or the next 15-20 years which is the expected lifespan of these kind of systems. It’s not that you could “break” hot sand in a box per se, but everything around it, especially the plastic parts like gaskets and seals, or motor wire insulations etc. degrade. As always, the cost and rate of maintenance goes up over time (see: “bathtub curve”) until it’s cheaper to just replace the entire thing.

            >eventually the supply is going to dwindle

            When gas supply dwindles to the point that people can’t afford it, grid electricity prices will go up higher because the cost of stabilizing the grid will shoot through the roof. IF someone comes up with a solution for that, then we’re talking about an entirely different situation where storing heat in hot boxes of sand is a moot point, because there’s this other technology that can store or produce equally massive amounts of energy at will and output electricity rather than heat.

            And, there’s the push for turning excess renewable power into… guess what? Hydrogen and synthetic natural gas, for one because the industry needs both to run processes, and because there’s not many other places you can sink such massive amounts and still recover the economic value. The industries consume and need multiple times more of these than domestic consumers, so there’s bound to be reasonably priced gas available, or else we need to give up making things like steel, gas, and fertilizers.

          2. correction: steel, GLASS, and fertilizers. Point being that turning electricity into high grade heat is very difficult, since reaching temperatures where things like glass melt also melt the cheaper kinds of resistive wires, and processing the kind of materials that DO withstand the temperatures requires even greater temperatures. Also, the chemical reduction of iron, silicon, etc. from oxide ores needs hydrogen and/or carbon.

          3. In the case of a hot box of sand the few bits that may fail in a human lifespan are (or at least should be) trivial and pretty cheap to replace – the structure itself should be almost bulletproof reliable and enduring as can be. While being a very large portion of the materials and labour to construct – In much the same way old houses tend to still have the original central heating pipes, many the original radiators but they certainly have gone through a few boilers…

            So many industries that have required carbon sources are finding ways to move away from it, and the few that it just isn’t practical are likely to be the only consumers when it gets scare. The everyday citizen isn’t going to get much priority there – afterall what is more important the individual who could get their energy easily from almost any source, so they really don’t need this scarer resource or the industry that actually make stuff, that enables other stuff to be made and effectively forms an essential part of the economy for a nation..

            Turning renewables into a combustible fuel is a great idea, but actually doing so at scale and efficiently enough to replace all the current consumers doesn’t seem plausible to me. By the time you have enough renewable excess to produce nearly that much all the domestic consumers are not likely to be using any combustible fuels – electricity will be so oversupplied so often using anything else will seem stupid.

          4. > but they certainly have gone through a few boilers…

            For the same reasons, the box of hot sand will have a limited lifespan. Even stainless steel corrodes when you keep it so hot, and you don’t want to use expensive materials like titanium for obvious reasons.

            Aside from that, what makes it expensive to maintain is replacing bits one at a time, which incurs extra labor cost and overhead. When parts start to break, they soon follow one another and you’re calling the maintenance guy over every couple months until it’s all “refurbished” – so it’s cheaper to pull the entire thing at once when it reaches that point.

            >So many industries that have required carbon sources are finding ways to move away from it

            Yep, and the answer seems to be “hydrogen”, or synthetic methane as hydrogen is a bit unwieldy to transport.

          5. >by the time you have enough renewable excess to produce nearly that much all the domestic consumers are not likely to be using any combustible fuels – electricity will be so oversupplied

            To get to that level of excess production, you need to solve the storage problem, because the grids just can’t handle the amount of excess without a place to sink it. Massive over-supply of electricity can’t happen, because nobody will build the capacity without anyone buying the power.

            Consider, the peak to average ratio of a thing like wind power is about 5. In order to supply some average demand, the peaks exceed the demand by a factor of 5. This will simply break the grid – it cannot function like that. Nobody will build wind turbines that have to sit idle right when they would produce the most energy.

          6. The economic term that applies to the case is “price cannibalization”.


            The solution we’re using right now is power purchasing agreements that force the utilities to buy the power, but that’s simply kicking the can down the road. Then THEY have to find buyers for power with no demand. This results in negative prices, and when even those aren’t enough, curtailment and penalties to be paid to the producers. This is an unsustainable situation, because it simply increases the cost of power and under-utilizes the resource. Adding more renewable power is cannibalizing its own market unless.

        2. The outer casing of your basement full of hot sand (the usual method suggested) can be entirely gone and not make any difference to many of these systems Dude, as the ground its burried in is still there – and it probably won’t go anywhere at all anyway as most of them are designed such that there is more than enough mass and volume of sand/ceramics to buffer and insulate the actually hot part from the exterior – and there has to be for safety and efficiency. While the direct heat exchange parts it is worth making in more expensive materials (assuming the design of this particular energy system gets hot enough and puts any parts that may degrade in hard to get locations, which not all of them do) – You are not going to want to dig up and reconstruct the basement ever – its just too inconvenient and expensive to effectively put a lifespan on quite likely the whole building that isn’t in the decades.

          I agree the current economic models have to change, but they are changing to be less fossil fuel friendly as time goes on already. For instance the ECT is being abandoned by many nations. And to actually get to the stage that you have enough renewable to produce and ship all that fake fossil fuel the whole model has to change – you can’t get that quantity of synthetic fuels from renewable without a huge excess of renewable, at which point electricity in general becomes massively cheaper and most users will switch to directly using that electric, being cheaper and more efficient – you don’t even need the grid scale storage nearly as much as you suggest as the individuals and companies that get cheap rate energy during the boom times will be load balancing for their own benefit while the baseline loads that can’t shift have enough. Some synth fuel is certainly going to be part of the future but wholesale replacement of natural gas with renewable created synthetic stuff just requires way way way too much renewable capacity, and serious investment to be able to take those boom times into synthetic fuel production at huge scale at which point anybody that can use the electric directly will, as its cheaper.

          1. >The outer casing of your basement full of hot sand (the usual method suggested)

            I was talking about sand heated to 600 C. It has to be isolated or it would lose heat too quickly. Point being that you don’t need to build a house on top of your heat storage, because it can be the size of a regular water boiler and still hold significant amounts of energy. That way it can be retrofitted to existing houses without major modifications.

            >they are changing to be less fossil fuel friendly as time goes on

            That says nothing about renewable synthetic fuels.

            > the whole model has to change – you can’t get that quantity of synthetic fuels from renewable without a huge excess of renewable

            And you can’t get huge excesses of renewable power without having processes that can absorb the excess – these things go hand-in-hand – plus you absolutely need these chemicals anyways. You will never have cheap excess electricity first and then synthetic fuels later. The production can only expand as the consumption goes up, because the price has to stay up to make it worth the investment. The consumption goes up by industry using the cheap electricity to make chemicals that they need, which they can no longer source from fossil fuels.

            >will be load balancing for their own benefit

            Whether you build the big batteries on the grid, or have the consumers shoulder the cost and buy batteries themselves, it makes no difference on the amount of storage required to deal with the case. You use “load balancing” as a magic word without considering the implications of how one can actually pull it off.

      2. I saved a pdf of an example of this 20 years ago, but would need to dig it out. A woman in New England (I believe) dug a sub-basement maybe 20 feet square and 4 feet deep. Four inches of foam all six sides. Filled with wet sand from hot air or water, she hoped to heat it all summer and heat the house most of the winter.
        The article was written near the end of construction, not sure how it worked out?
        Wasn’t horribly expensive to do , while building a house with a basement.

        1. A similar thing has been done on a communal scale in Drake Landing, Canada, but the system is a little bit different since it still involves water, which limits the temperature. It works just as well, but it takes more space and infrastructure, so it’s probably more costly per unit of capacity.

    2. Anyone knows if it’s possible to “dim” a water heater and make it effectively use only excess solar power, without requiring additional power from the grid? (Assuming I have that data available)

      1. I suppose you want to dynamically adjust the power draw to match the solar panels, rather than just switch it on and off?

        There’s water heater accessories like remote control sockets available, if all you want to do is switch it on remotely/automatically when it’s sunny.

          1. They might be more commonly sold as motor speed controllers than as high power dimmers, but the cheap devices with a dial that work by chopping the AC waveform are basically the same as a conventional lightswitch dimmer, and are by far the easiest way to quickly and smoothly vary a multi-kilowatt resistive load, even if they’re crude.

          2. Yep. I think I’ve already seen solar power controllers that can run loads like water boilers on a chopper switch. If my memory serves right, there was an Australian company that sells just the thing, but I can’t remember the name.

      2. There are absolutely commercially available devices for this exact task – they generally use pulse width modulation to meter out the exact amount of excess power that is available from solar and no more- so a 3kW element can be delivered only 1kW for example, if that is what is available from the PV array. A decent system like this has top and bottom client temp sensors and ability to switch to “boost” the tank if the temperature hasn’t reached a set point by a certain time of day due to cloudy weather for example (so that the home owner has hot water when they need it). See catchpower.com.au for an example.

      3. Yes. There are commercial products that do it, eg Marlec Solar iBoost+

        It uses a clamp meter on the wire into the property to spot exported power, and drives the water heater to use that power instead.

  2. I’ve been thinking about doing this for a long time. But when I do the math, it doesn’t seem worth doing. Please comment.
    For example, my utility is charging roughly 16 cents per kilowatt hour. Even if I had 400 watts of solar power 4 hours a day that would be about 25 cents per day. Around $7.50 per month. With a $1,000 invested into the whole system, it would take 10 – 11 years to break even. Thoughts?

    1. 400 Watts of panels on a yearly average does somewhere around 400 kWh depending on where in the world you’re at. It’s probably worth $7 per month at maximum for those electricity prices. If you’re not careful about what and where you’re buying, and you’re not getting subsidized, the payback can easily be “never”.

      The average is about 7% of the nameplate power if you’re in Sweden or Canada and 17% if you’re in Spain or California, then just multiply by the number of hours in a year – but you have to remember that this is a whole year average. You can’t expect to get any power December, and vice versa you get more than you need in June.

    2. I’m not sure about your local market but for mine a 400W solar panel is about 70 GBP, payback time about 160 days. It sounds like you’re either overpaying for labour, mounting hardware, wiring, inverters etc or paying for equipment that can handle more than one panel. Say I paid 500 GBP for a Growatt MIN 4200 TL-X inverter, that’s got 5880W rated input so would handle 14 panels. That’s a much better ratio.

      1. Average cost of installed watt is between 3-4 dollars. The smaller the system, the more it costs per watt, because there’s licenses and inspection and permitting costs as well, and those don’t scale down.

          1. Dude is about right. I’m having a system installed (8800 kWh/yr) for a bit over $35k (all-in, permitting labour, battery system) which, to my untrained eye looks like $4/W.

            Payback time is not an issue as the local utility is now charging $1/kWh during peak ToD at tier 3 rates…

          2. On a side note, I found that the output of solar panels in Amsterdam is estimated at 0.9 kWh/Wp per year. That’s a capacity factor of 10%. At €1.20/Wp and a payback of 15 years that corresponds to a grid value of 7.4 cents/kWh. That’s not bad at all, considering you don’t pay for transmission.

        1. All this is if professionals have to be involved. If people are allowed to use solar panels to power something that’s not grid connected without any license/permit/inspection, then systems with much lower cost per watt can make sense, even if they aren’t using every available watt optimally.

          1. Any time you run significant power down wires, there should be at least a fire safety inspection. There’s already enough trouble as it is with people running illegal “extensions” on their properties and causing electrical fires.

    3. If you’re at $2.50/watt when the panels themselves can be found for $0.25 to $0.75 per watt, your stated system price is mostly made of unknowns. I’m going to guess that you are either looking at a kit with a surcharge, professional labor, an inverter that ties into the grid, or batteries, all of which aren’t strictly necessary. The same strategies that work in a place with expensive power and favorable utility rules (net metering, for example) aren’t what you should emulate in other places, even though solar can still be beneficial especially in terms of a backup supply of power in a grid outage.

      It can be cheaper and easier to do clever tricks like the solar heater in this post, especially if you don’t need a professional when you aren’t touching the grid. As an extreme example, imagine if you propped up some panels at an angle against the south wall of your house, without an actual “installation” to cost money. Then you pick whatever’s the cheapest way to get that power into your hot water while cutting off if it’s at maximum temperature. I believe some electric water heaters have two separate heating elements with separate thermostats and everything, so maybe you use that. Or you add a circulation pump and heat the water externally, maybe. Whatever’s not too unsafe for wherever you live. There’s plenty of clever ways to get it done for cheap, so taking that as a given, the same budget allows much greater capacity of panels, and even if you waste a lot of their capacity, you’re bound to save more power than 25 cents per day.

      A different example that is easier to do if not always as cheap is running specific devices off of solar with grid as a backup. A small scale example is if you had a small solar panel charging a cheap battery that powered your home wifi, and a battery charger that only turned on if the battery got low. That won’t save you much power, although it could help for reliability. A larger example would be if you did the same thing with a refrigerator, or some lights, or an air conditioner or heat pump. In the last case, it’s good to note that there’s hybrid mini-split units that operate at variable speeds, so they can be set to continuously use however much solar power is available without needing any batteries or huge inverters or anything like you’d need to power a typical central hvac system. The idea of all these being, it’s cheaper to get extra panels and use them less optimally, if you have enough room and won’t need to pay too much more for installation. If that’s not true, that’s when you’re looking more at typical suburban systems where you want to see a net metering or other easy home generation agreement where although you need an inverter that talks to the grid, at least you don’t need to buy batteries too.

      1. There’s cheap, and there’s effective.

        For example, propping a panel on the ground by the south wall might be cheap, but it gets shadowed by buildings and trees, so you lose power and whoops, cost per watt actually goes up. Likewise, forgoing a charge controller to save money means you don’t have MPPT, which reduces efficiency and again, cost per watt goes up.

        1. I wouldn’t go that far either, to be fair. It’s generally easy enough to do a bit better than just propping up a panel somewhere that isn’t shaded. I guess I wasn’t clear enough that it was an extreme example.

          But would you really rather have one panel on a tracking mount than (for instance) four of them on a seasonally adjusted south-facing mount or twenty of them at the wrong fixed angle, if all options are the same price? Because while you won’t get full capacity out of the suboptimal setups, that doesn’t mean you can’t ever sacrifice efficiency.

          1. I would pick the middle option, because having more panels means having more hardware to maintain, which costs more $$$ over time. That’s also why I wouldn’t pick the tracking mount – it’s higher maintenance – since it’s these sort of “secondary” costs that make up most of the total price with solar power.

            Having the “same price” up front doesn’t necessarily mean they’re all equally good deals.

          2. Sure, seems like a sound thing to consider.

            Of course, in the third case maybe if the angle is steeper than what would generate the most theoretical power, it could reduce problems with snow accumulation or reduce the likelihood of hail damage. Or maybe it sometimes makes more sense to arrange panels to smooth out the peak, instead of buying more energy storage multiple times over the lifetime of the panels. In addition to just aiming somewhat east and west, I also saw some bifacial panels oriented vertically so they are edge-on at noon, which seemed odd but apparently it worked out – when snow was involved at least.

  3. I don’t know what current practice is in South Africa, but in the 70s & 80s the geysers were open to the atmosphere, with level regulated by a float: they gravity-fed to the outlets. This was to “prevent boiler explosions”. Commonly placed in the attic of a house, they had an overflow pipe projecting above the roof tiles.

    With the hot water being gravity fed, and the cold water being pressure-fed from the water main, it made for interesting showers when someone else used water in the house.

    If that arrangement is still used, I wonder if a thermostatic mixing valve would work properly on such disparate pressures in the hot and cold lines.

    1. Our geyser is a commonly available brand, and is sealed. We do have a pressure relief valve that vents to a tray under the geyser (required by code, both the valve and the tray), and the tray is piped to a drain. Ideally, one that is visible, so you can see your steaming water being pissed away, and do something about it! :-)

    2. The sort of diaphragm pump used for RV’s seems as if it could help, although you’d need a somewhat larger one for a house’s flowrate. Or, I think there’s pressure regulating valves also originally made for RV’s that could slow the cold side to a more reasonable pressure.

  4. I made a domestic solar water heater about 35 years ago. Two panels, a 120 gallon tank with electric elements at the top and my own control. Its been running all that time with minimal maintenance. During the summer, the tank will be 160F at the top and a little less than that at the bottom. The control has a max temperature setting that is set at 180F. During all that time, my hot water costs me about $30 to $40 a year to run the pump. Our winters being mostly iffy for solar here in New Hampshire, having those backup heaters in the tank was a wise choice for us. We don’t have to do anything when the cloudy weeks occur.

    I find it interesting that people don’t really look at the long term advantage of solar energy for heat.

    When I was first starting in solar energy in the mid 1970s, storage was probably the field where experimenters were most focused. Sand, small rocks, water, sand and other mass storage systems were all considered. The size of the storage medium was a real roadblock for everyone. I tried a few things from Mother Earth News and Alternative Sources of Energy magazines but always came back to a regular hot water tank as being easiest.

    1. If you’re confined to temperatures below boiling, then water can’t be beaten. Even phase changing materials like some waxes and salts don’t store that much more energy per mass or volume, and the cost compared to plain old water means there’s absolutely no sense in using them.

      >I find it interesting that people don’t really look at the long term advantage of solar energy for heat.

      Direct solar thermal maybe, but wasting solar PV power to heat is just senseless. It’s too expensive for that, unless you live in California or Hawaii where the power system prices are ridiculously high to begin with.

  5. I have always wondered about backing solar panels with, essentially, liquid cooling heatsinks tied to a solar hot water preheater, since solar panels are reduced in efficiency as they get hotter; transferring some of that heat to a water tank that feeds the HW heater should result in a gain in efficency for the entire system.

      1. I have seen mention of water-cooled solar panels, but not really seen any indications of cost. The added complexity/weight/materials also sounds problematic.

        That said, if the tubes were built in at the factory, and provision made for water connections, it might help raise the efficiency of both the panels and the water heater. But, it also means that you have changed a solid state solution to one with moving parts. And you’d have to drain it in cold weather in a good part of the world, etc.

        1. I doubt you would have to drain it in most places. Just don’t run straight water, add a little antifreeze and have a big enough tank to hold substantial heat from the daytime that can be pumped around over night as needed to keep the temperature in the panels up enough to prevent freezing. Yes you’d lose a little of the heat captured for maintaining a working system rather than your benefit, but unless you are talking above the arctic circle type we get really cold with no daylight for a very very very long time the loss wouldn’t be much compared to the heat captured over the day.

          Also as the Solar PV side is still solid state even if you added liquid cooling, will work without the cooling connected I’d not call that a major problem for a system like that. As the panels are effectively still simple solid state with some tubes on the back the one moving part likely to fail is the pump that will be put somewhere easy to work on and the joints between the tubes on the panels are simple mechanical parts that should be easy to replace/repair (once you have the scaffolding (etc) to get up there).

          1. Thing is, for efficient solar thermal collection you want the water to run at relatively high temperatures up to the boiling point, which hurts the solar panels. Direct solar thermal systems run hot oil, so they get more efficient heat transfer with reasonably sized and priced components. It will be a compromize, because you want the panels to remain relatively cool, but that means you get luke-warm water out of them and the thermal storage needs to be bigger to store the energy.

          2. Indeed Dude, but that is an entirely different issue to water cooling your PV and making use of the heat you are removing. Moving into whole different categories of system. Though I’d disagree on ‘hurts’ the solar panel, it doesn’t do their efficency any good, but doesn’t (generally) cause them any damage to be in the water boils ballpark – good thing too as that is the sort of temperature they can easily get up to without cooling…

          3. Dude the degradation of a modern PV is really really minimal and not made massively worse by the temperatures we are talking about here in any wear test I’ve seen. Older panels absolutely couldn’t take it (or the UV) but modern ones are vastly better at dealing with these temperatures – which is probably the largest part of why they just don’t degrade in use the way the old ones used to.

        2. The solar panels should not be getting much hotter than 30-40 C with sufficient ventilation under the panels. The efficiency will start to suffer above 50 C, but that’s too low for hot water tanks.

          One way around this is to run the heat into a pile of gravel or a large pool of water that never gets too hot for the panels. This is not hot enough for household water, but it’s hot enough for space heating, The reason why you can’t use it to pre-heat water going into your hot water boiler is that no water that comes out of your tap should ever dwell in the 20-45 C range that is conductive for bacterial growth – it should be either cold or hot, not in between.

          Another variation is to have a heat pump operated hot water tank, that draws its source heat from your solar thermal tank, so you use minimal electricity to boost the temperature up to your boiler temperature.

    1. What makes a lot of sense is to use a concentrating collector, to concentrate 3-4x solar power onto a solar panel, and place a 10cm-thick water bath in front of the solar panel. The water contains a red dye, allowing only red light to hit the panel. The dye absorbs the green-blue light, and water absorbs the infrared. Most of the sunlight is turned into heat in the water, and the solar panels remain cool, since they receive only red light, where they are 30%+ efficient at converting light into electricity. It will double the amount of electricity from a given panel, while reducing its temperature.

      10 cm of water is a large weight load, but you need 10 cm of water to get enough infrared absorption to make sense. Specific IR absorber dyes (in addition to the red colour green&blue-absorbing ones) might help reduce the thickness needed.

      1. If all but red-orange-yellow light is blocked from the panel, I’d estimate that the panel would produce considerably less than 50% of the electricity it would if it got the full spectrum. That makes 4x concentration a minimum to get double the electricity per panel.

        If the cells could be make absolutely waterproof, they could be submerged in the water without concern for structural integrity. Contact with water is the heat sink, no need to filter out the IR. In fact, silicon has good sensitivity down to 1 micron, so filtering out near IR is counterproductive.

        1. Silicon cells produce one electron per photon, producing about 0.6 volts. Any photon energy in excess of that gets turned into heat.

          A red photon, with its 1.2 electron-volts of energy, can produce electricity at around 50% efficiency. A blue photon, with 2+ eV of energy, can only manage 25%. In practice, other losses reduce those numbers.

          And you absolutely want to absorb the IR before hitting the panel, and also thermally decouple the panel from the hot water: Let the water get hot, and keep the panel cool for better efficiency.

          Maybe immerse waterproof panels for even better cooling, but you want to keep that part of the system cool.

          1. Silicon PV cells work well into the infrared spectrum, at least down to 1100 nm (1.1 eV). Below that, there sun’s emission spectrum tapers off, and the atmosphere itself blocks enough IR, that you don’t actually have all that much radiation to remove.


            Just eyeballing the graph, probably 4/5ths of the energy from sunlight that reaches the ground is above 1100 nm.

  6. The charge controller I got from Midnight solar
    has a number of programable set points to enable things like hot water heaters,so that as soon as there is excess power ie:droping out of mppt
    it will start managing another device,and can do PWM on a resistance coil HW heater
    or even better a domestic HW heat pump

  7. I got similar setup, except I added a few additional components. I use a 200l Solar geyser, then the geyser is connected inline with my gas geyser (60degrees mixing valve before the gas geyser). I only heat my geyser with solar power (if it is available) from 11am to 1pm to a max of 65 degrees. As its a solar geyser, the sun also keeps it warm and temperatures can raise to in the 70’s.

    Once you open the tap, the water will flow, the gas geyser will ignite but once the warm water reaches the gas geyser, it will determine if the water is already at 60 degrees, if it is less than, it will boost the temperature to 60 (for example if the water is 50, it will just boost it to 60). If it is 60, it will shut off and not use any gas.

    Week temperature image

    So, I always have warm water, if there wast enough solar to heat the water, the gas will make sure its warm enough.

    1. Except I already HAVE these PV panels that are generating more power than I can otherwise use, and I don’t have “solar thermal panels”. So, should I spend $80 on retrofitting my existing heater (including the new 2kW element), and make use of that excess PV, or should I spend $1500 on buying a new solar water heater, find space for those panels on my roof next to the existing PV, and pay someone to install all of that? My $80 outlay seems like a fairly smart thing to have done, IMO.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.