MAILBOX: PV-Direct Water Heating Vs. Solar Thermal

A Willis style immersion heater is plumbed in parallel with a standard water heater to provide a small amount of hot water quickly, while also contributing to the main tank temperature.

In Mailbox HP179, Hugh Piggott cited some advantages that make PV-powered water heating preferable to solar thermal technology. Here are a few more:

  • Reliability—There’s no heat-transfer fluid that can leak or deteriorate; and no circulating pump means no moving parts that can fail.
  • Simplicity—The systems are easier to install, test, and commission. A tank water heater having one ordinary and one solar immersion element is no big deal, compared to a special “solar tank” with its necessary heat exchanger.
  • Autonomy—Many solar thermal systems need grid electricity for the pump controller and the circulation pump. PV systems can be entirely self-powered—ideal for those in developing countries, as well as “off-gridders.”
  • Layout flexibility—It’s easier and cheaper to run cables than plumbing. There are no airlock worries, either.
  • Guaranteed energy harvest—Lower solar input in early spring, late autumn, and winter, can compromise a solar thermal system—will the collectors achieve the necessary temperature to input heat to the tank? With PV, if there’s any energy available, it’ll be harvested. Plus, PV modules are more efficient in cold weather.
  • High-temperature capability—Even if the available PV power is low, it can directly heat the hot water supply tank; there’s no need for a preheat tank.
  • Instant delivery—By using a Willis-type external immersion heater arrangement (popular in Northern Ireland, but universally applicable), hot water can be drawn almost as soon as your PV array starts generating a surplus.
  • Greater end-use flexibility—With a PV-heat arrangement, you aren’t just restricted to heating water. With a little creativity and electrical know-how, you can use conventional switches and relays to readily swap between water heating, background space heating (very useful in spring and autumn), or greenhouse soil heating. You can also cook with PV power—think of the deforestation, the fuel-collecting time, and the smoke-related illnesses this could prevent. Meanwhile, even Scots, Alaskans, or Canadians can—using a conventional AC hot plate fed from a dedicated “PV surplus” socket, on sunny days—be bulk-bottling fruit or juice, jam-making, or brewing beer in large batches, and enjoying full tanks of hot water!

Christopher Jessop • Pembrokeshire, West Wales, UK

Comments (6)

Captron's picture

Peter, the idea that you can get solar based heat from a PV system is valid. If all you need is a heat dump once batteries are topped up it can work fine. If however you have a grid tied system then grid storage is out while heating the hot water tank. Bit of a tradeoff here.

Also your control system for heating, if on grid, will need to manage both PV-DC variable voltages and wattages as well as 240v if you want continuity of thermal heating.

To use PV effectively you need to match the resistance heat load, lets say its a 2500w element, then you would need no more than a 2.1kW solar PV system. You don't actually need this much as a DC system will still dump heat to the element at lower wattages, it just takes much longer to reach HWT thermal set point or saturation.

Up to this point I agree with you.

However: If you have limited available roof space, a single 2.5kW Solar thermal panel would be a great idea and you can still possibly install more PV. Also if you are thinking of heating more than a HWT, say a whole home or warehouse then you really need some real substantive heat, especially if matched with insterseaonal storage.

FYI: By the end of this year we expect to field a combination PV & ST panel. Reducing the static heat in a PV panel improves its operation as much as 25%. These are flat plate panels so not good for heat all year but can help heat things like pools that are limited in when they can be use in cold climes, possibly extending their use throughout the year. In areas that don't get seriously cold, lets say below -10c (~15f), these combo panels will function quite well for heat and hot water.


Peter Gruendeman_2's picture

Hi Solarfire:
In regards to what kind of PV-->DHW system one wants, the biggest consideration is whether one wants backup power or is living off the grid. If batteries are part of the system then a diversion controller is probably best. Once the batteries are topped off, surplus power is sent to the water heater. That could happen about 8:00 in the morning if the batteries are not used much, but they will be topped off daily and they are ready when needed. Under poor sky conditions there is little or no surplus (beyond charging batteries) so any extra power that can be collected by a MPPT controller isn't there.
For one who only wants hot water, a MPPT controller will harvest power when sky conditions are poor. All those watts are good whether one is getting 10 watts or 1,600. It all heats the water. I am not familiar with the MPPT controller you cited. The controller I designed and built is shown on page 52 of HP 177. It's a pretty simple project. Mine is rated 340 Voc, 40 Isc-- up to 10kW of power handling though 2kW ($1,400USD) of made in USA panels produces as much hot water as 2-3 people can use. When the water heater reaches its maximum temperature my controller diverts the power to electric baseboard heaters, also run on DC. For others reading this post, your 240 VAC electric resistance water heater runs perfectly fine on DC, of any voltage as long as the resistance value of the element works with your PV array. Electric baseboard heaters also run perfectly on DC.
Switching high voltage/ high power DC with mechanical switches is a bad idea; switching DC with MOSFETs or solid state relays is trivial. Ask the sysop to send me your email and I'll send you the handouts from my lecture at the MREA energy fair.

Regarding the videos (last few characters of the URL):
Arizona, '7B5s: He put the temperature probe in the top of the tank--No! it goes towards the bottom of the tank, preferably at the same elevation as the lower heating element, either left or right of it, though mine ended up at a slightly higher elevation, which is not the best but it's ok.
Arizona, 'RVe4y: This is a really informative video. He needs some kind of temperature limiting device. The T&P valve is NOT acceptable!! Solid state relays can be bought used on ebay for $75. I use the Crydom D2D40, rated for 200 Voc, 40 Isc. That's a lot of power, and if designed into an "on or off" control ( it's definitely good enough as long as the water heater has the proper resistance. It's true that early morning sunshine and cloudy days will reduce the array voltage (wasting power) but those are the conditions when little power is available. My MPPT collects 100% of nil-- which is still nil.
North Caroline, 'emteY: The MPPT controller looks a little light for 2-3kW, or maybe it's only rated for 1kW? I specifically stayed away from electrolytic capacitors as I found none that were rated for more than 20,000 hours of service. I used motor run capacitors which have service lives measured in decades, even when running 24-7. That is the point of PV for DHW-- no maintenance, no moving parts. For me, backup power is a bonus. Nobody ever got backup power from their solar thermal system. In 2018, and beyond I presume, PV provides more watts of heat per $1,000USD of panels than solar thermal so why would you spend more for a system that requires some maintenance and can not provide backup power?
My PV for DHW system provides all the hot water this household of one can use. 1,200 Watts or $850 in panels in 2018. It's been in service since August 2015 (three years and a week as of this writing).
Pete Gruendeman
La Crosse, WI USA

Captron's picture

Chris, as I intimated previously, I appreciate your consideration of solar PV-Thermal but still believe that while technically correct you still are not giving the whole story. And, as given you leave a reader with a severely slanted view of the realities of a PV heating system. Asa principle, I spent many years in the Solar PV industry, and now I am in the Solar Thermal (ST) industry also as a principle again for many years and understand both systems intimately.

Let me answer your questions and let’s move on.

PC: You said: “energy harvested per $1,000 of investment.” Let’s Use your number of $2k for the PV system. You harvest energy at a rate of roughly 66% of the size of your panels so let’s say on average your solar array produces 1kW. Over a month, on average you might generate 1x10x30 or 300kW based on an 84SqFt panel footprint .
ST: Qty 2 x 30 tube ST panel, with a smaller aperture of only 58SqFt, generates ~4.8 kWt, therefore with an 84% efficiency over a month you might expect to harvest closer to 1,200kWt. The performance envelop comparing both is closer to 4:1 at peak and yet when the variabilities of weather are factored in the numbers skew well in favour of ST.

Zero Thermal mass: Very true and it does take a few seconds for a true solar thermal system to begin full operation. Let me stress here, a few seconds. Does a DC based PV system lose any energy to transportation, absolutely. If someone wants a HW system with no moving parts then PV is the way to go, sort of. Remember when dealing with DC and heat you also need to consider corrosion, not mentioned, as well are critters.

Let’s use your PV example for comparison: You have 4 solar panels connected to a HWT (Hot water tank) and let’s say you are wired using DC. This does save some conversion losses. Thus while rated at 340W (An awesome solar PV output panel BTW), with a total potential output of 1360W on a perfect insolation day, applied it to a typical 2500W resistance element, then I agree you will get a low but decent thermal rise if the sun is out all the time providing this peak energy output. It would be like trickle charging a Battery, it just takes time.
We both know perfect insolation performance is rare over a given month, and that EvT (Evacuated Tube) solar thermal panels perform quite well in subdued visual light and with cross shadows, due in no small part to their comparative visual acuity spectrum association. Thus if you compared a solar thermal setup in a similar fashion to your PV setup, your EvT performance would get roughly >3.5X more heat energy instilled into the HWT for the same solar panel area.
Having said that, if we were talking only about a HWT, you make some fair points, assuming you are willing to live with the moderate to poor performance. If you want performance over a limited roof area, you can do much better with Solar Thermal [ST].
The way I see it is you have put a lawnmower engine into a small car. Will it work; yes. Will it get you where you want to go; perhaps. Is it the best option, because the engine is small and uses less gas; you decide, because just getting there is only part of the equation.

Yes your $2k in solar PV would cost more as a Solar Thermal [ST] system, although to be honest we would need only 1 panel to match your comparative poor performance. If we used your same solar aperture, then your hot water set-point would be hugely more reliable with ST, especially in cloudy environs where PV is poor and ST functions well, or with a high hot water demand situation such as with a large family.

ST overheating during power outages: not sure where you got this but it does not happen with a ST Drainback system.

Remember, a HWT is a fraction of an average single family dwelling’s energy footprint. We (Digital Solar Heat) are now heating whole homes, including their hot water with solar thermal, even in cold regions like yours, some are heated using solar for 100% of the year.
I did a quick calculation here for a standard fully insulated home in the 139SqM or 1500SqFt range, and there is not enough roof space to put a PV system out there to come close to the ST performance envelope. I also calculated it would cost about $27/yr in electrical energy to run the system if a PV panel was not installed. I do acknowledge you are more focused on a single HWT but the same principles apply: The right tools for the job.

Double your PV system size to more accurately and reliably heat your HWT and your costs have increased to where it is cost compatible with ST.

So in conclusion, I am not saying this PV HW heating system will not work as described, and yet I am not saying PV is better or has more going for it other than no physically moving parts, what I am saying is if you have limited space, want thermal performance and temperature reliability, while maintaining a reasonable price point, I would take ST any day of the week. It’s simply the right tool for the job.

Ron Theaker CD, Calgary, Canada.

Peter Gruendeman_2's picture

Hi Captron:
In order to have a meaningful discussion it is necessary to speak the same language. The language I speak when I teach PV for DHW is that of energy harvested per $1,000 of investment. It's true that solar thermal products harvest more energy per square meter of sunshine but square meters of sunshine are not in short supply. It's dollars that are in short supply, so making better use of them via PV for DHW is the solution of choice.
There are several aspects of PV for DHW that make it less expensive than solar thermal for hot water: It's a very simple system that more people can install themselves. There are no moving parts, no heat transfer fluid or pumps to maintain. There is exactly zero thermal mass in the system so the instant the clouds clear, energy is collected. Or the clouds don't clear and at least some Watts are still flowing into storage. No Watts are wasted heating the panels or tubes, or piping that delivers the heated fluid. No Watts are wasted to cold ambient air, no matter what the temperature and sky conditions are.
Where most people get hung up is they don't know that PV, which now costs 0.70-0.75 USD per nameplate Watt is only slightly more expensive than flat plate solar thermal and PV costs less than Apricus evacuated tube systems per Watt of heat. The other confusing area is that people are unaware that electric resistance water heaters work perfectly well on DC directly off the PV panels. The array needs to be wired to match the impedance of the heating element, or swap the heating element to match your PV array. It's pretty simple stuff.

My PV for hot water system is four SolarWorld 340 Watt panels, which has reduced my use of electricity for hot water by an estimated 97%, to a measured cost of just over $3 USD/ year. This is at La Crosse, Wisconsin, not exactly the solar capital of the USA. My PV panels cost 1,200 USD two years ago. Can you provide a complete solar hot water system, with a 25 year guarantee and no maintenance for $2,000 USD? I can.

Potential buyers of domestic hot water systems should ask themselves:
Do I want a system with zero moving parts and a maintenance-free working fluid?
Do I want a system where all the components are guaranteed by their makers for 25 years?
Do I want a system that can provide backup power (with the addition of batteries + inverter), or do a want a system where the glycol is overheated and ruined during a prolonged power failure?

Learn more by seeing my lecture at the MREA energy fair in June.
Pete Gruendeman
La Crosse, Wisconsin, USA

Solarfire's picture


I have read your comments with great interest here at HomePower.

I currently have a solar hot water using an efficient solar flat plate solar hot water collector. It produces hot water in the late spring, summer, and early autumn. In the winter, the water tank has an electric element which heats the water to the desired temperature; however, in the summer, it has been causing all sorts of problems.

Therefore, I have been researching PV-hot water, and it seems to be a simpler and cost effective solution. The costs I have without the water tank are around $1,500 to $1,600 with 3-330 watt solar panels from Panasonic and an MPPT device for hot water heating.

I have been looking for a resistive MPPT device to boost the PV energy from the solar panels. I have been able to, sadly, find only one. The rest of the MPPT devices out there treat hot water heating as a dump load resistor. In addition, I have found several useful videos of installations using PV panels linked to a hot water tank, and users appear to share your experience with them.

The link for the MPPT device:

The link to the videos:
No MPPT Arizona:
No MPPT Arizona:
No MPPT Arizona:
MPPT North Carolina:

As you can see there are different approaches to achieve the same outcome.

Would you mind discussing these different approaches?

Best regards,

Captron's picture

Chris, it's not that the comments are technically wrong, but they do give you the wrong impression. This article reads like someone trying to convince me walking is better than driving a car. Yes if you are walking locally this may be a good idea, but if more than that is involved, such as distance, carrying capacity, infirmity, and braving the -40c elements, then a car is a much better idea. You can scrounge up all the minor details to support your conclusion but in the end it’s a disservice even if some of the some of the words can be construed to be correct.

Christopher, this article while barely accurate is so far removed from the reality of solar thermal that you leave readers with the impression that this is a valid premise when in fact it is not. The only person who might garner some confidence from this is a solar tinkerer, not a valid homeowner or business owner who wants to make a real difference with respect to heating or hot water.

I live in a country where heating is life or death, nothing you have illustrated would make me think this option of PV powered thermal hot water for this use is even close to valid. At one point I thought a PV powered heat pump might fit this bill but in testing this I found it too was poor substitute for the real thing. Not only did you need twice as many panels to get the thing started but locations with ambient temps of -25c can’t efficiently operate in an air-to-water environment. On top of that you needed 4 times as many PV panels for the same thermal energy output. I am focused here on quality EvT panels not pipes or flat plate panels.

In addition some of the comments made are categorically inaccurate. Low angle solar is just as bad for PV as it is for EvT panels. There is a scientific reason for that, it’s called solar angle of inclination and if solar energy is drawn through thousands of KM of earth’s atmosphere it will be attenuated or degraded. And, an EvT solar panel will outperform a PV panel in this insolation regime over 3:1. Also the EvT panels are not susceptible to random partial shadow events, or angular orientation or even bird droppings.

This article needed an intelligent and thoughtful counterpoint otherwise it is just not good journalism even if the title forces you down one path: Change the title.

Show or Hide All Comments