Device for Water Heating Using Solar PV Energy - Feedback Wanted

[SolarMan]

Hi All 

I've written a proposal for a device which will allow solar PV power to be used directly for water heating (see attached PDF).

Have you any suggestions on how to develop the design?  Unfortunately, my electronics knowledge is far from enough to do the circuit design etc. myself  .

I'm not looking to make any commercial gain out of this.  If someone to put the product and on the market themselves, that would be great, too  .  It's mostly about getting the proposal out there and offering any help I can to anyone who would like to make it a reality.

Let me know what you think! 

[I wanted a rude username]

The elephant in the room is why not just use direct solar water heating instead? It's far more efficient, probably more durable (consider the complete lifecycle cost), and arguably simpler to install (plumbing versus electrics).

[SolarMan]

Thanks for your interest   

Absolutely - the system would compete with solar thermal technology .  You're right that solar thermal panels win on efficiency.

But don't you think that solar PV wins on reliability?  Think burst pipes, overheating, circulation pump failures, leaks, etc. versus a solid state PV panel with no moving parts.

Re. installation, the advantage is that this can work with the existing hot water tank in many homes.

[NiHaoMike]

So you're trying to design a device to use excess solar power for heating water (or air or whatever), for places where exported power is paid at far less than retail rate or is not allowed at all? What you need to do that is some way to detect that solar power production is in excess of consumption and then switch a SSR to try to bring up consumption to match production.

More reasonable would be to build some way to allow loads to vary their power use to match production. I can suggest either PLC or Wifi for the communications to allow that to work.

[tautech]

If one analyses the available surplus PV power budget, hot water usage and whether grid tie is needed I think the new ranges of domestic hot water heat pumps show lots of promise apart from their initial cost.
Of course the efficiencies of the complete system of PV inverter and heat pump vs resistive heating need be considered too.
Climatic conditions will also have a bearing on efficiencies but luckily where PV is in widespread use ambient temps are also generally higher and hot water heat pumps will do well in these locations.

[SolarMan]

Hi NiHaoMike 

So you're trying to design a device to use excess solar power for heating water (or air or whatever), for places where exported power is paid at far less than retail rate or is not allowed at all?

No, not exactly.  That's a great idea and there's a device available for that (https://myenergi.com/product/eddi/).  My idea is different in that the system will be completely unconnected from the AC mains.  This avoids the cost of a DC-AC inverter and means that the system can continue to work even when there's no mains power available.

Another member kindly pointed out that you can already buy a device similar to what I am proposing (http://www.techluck.com/).  I really should have mentioned this in my initial post  .  Unfortunately, the product from http://www.techluck.com/ does not have a high enough power rating for families in cold places like Northern Europe.  The PDF with my initial post contains the calculations behind this.  Another issue is that the product from http://www.techluck.com/ does not have any system to prevent Legionella growth.

Using a SSR to vary the load sounds right.  Communication will not be needed as such (see Figure 2 in the PDF  d-Lumen: A System for Cost-Effective Water Heating From Solar PV.pdf (202.45 kB - downloaded 215 times.))

[SolarMan]

If one analyses the available surplus PV power budget, hot water usage and whether grid tie is needed I think the new ranges of domestic hot water heat pumps show lots of promise apart from their initial cost.
Of course the efficiencies of the complete system of PV inverter and heat pump vs resistive heating need be considered too.

I think you're totally right about the promise of heat pumps and analysing the performance of the alternative options 

The system I am proposing is not for everybody, but I think it could find a significant niche.  As you mentioned, the initial cost of heat pumps is a big problem for a lot of people and even 1-2% of the market for hot water systems could equate to millions of people on the global scale.

Economics wise, the current low price of solar PV panels (http://pvinsights.com/) helps  , and any future improvements in PV technology could make the financial returns even more favourable.

Climatic conditions will also have a bearing on efficiencies but luckily where PV is in widespread use ambient temps are also generally higher and hot water heat pumps will do well in these locations.

Funny thing is I'm actually thinking about this for parts of the world that are relatively less favourable for solar PV (but also with high energy prices, eg. many countries in Northern Europe).  Given the recent low prices of solar PV panels one could use a "brute force" method to get enough energy from the sun by using an oversized PV array in a cloudy country.  This is where not needing a correspondingly huge DC-AC inverter becomes especially helpful.

Thank you for taking the time to give your feedback 

[tautech]

Climatic conditions will also have a bearing on efficiencies but luckily where PV is in widespread use ambient temps are also generally higher and hot water heat pumps will do well in these locations.

Funny thing is I'm actually thinking about this for parts of the world that are relatively less favourable for solar PV (but also with high energy prices, eg. many countries in Northern Europe).  Given the recent low prices of solar PV panels one could use a "brute force" method to get enough energy from the sun by using an oversized PV array in a cloudy country.  This is where not needing a correspondingly huge DC-AC inverter becomes especially helpful.

Thank you for taking the time to give your feedback 

Yes well in less PV favourable climates the PV 'available power' budget will be tighter still to the point where heating water might not be feasible with resistive elements and also there is space heating to consider.

In an off grid house not far from me both potable hot water heating and concrete slab heating was done with a wood pellet furnace that ran 24/365 to supplement their 20+KWh of solar storage for a not small house in our relatively temperate NZ climate.
Most fittings were LV DC however selection of very efficient mains appliances ensured this family didn't suffer from any lack of the creature comforts most of us require.
While the purists might shudder at the thought of a fossil fueled pellet furnace while we still build with timber it's a perfectly valid use of the timber industry's byproducts of wood waste.

Still, if water heating is to be electrically powered with PV and without grid tie options IMO there's a opportunity to develop water heat pumps that are DC powered such are their efficiencies of heating water /KW used.

[NiHaoMike]

My idea is different in that the system will be completely unconnected from the AC mains.  This avoids the cost of a DC-AC inverter and means that the system can continue to work even when there's no mains power available.

The better inverters already support daytime backup without batteries.


The smaller grid tie inverters are not that expensive nowadays, what's missing is that none of the low cost units I'm aware of support "zero export" for areas where exporting power is not worth the hassle or is not allowed.

A DC side controller would pair well with a smaller grid tie inverter. It could direct power to DC loads like heating water when production exceeds what the inverter is able to accept.

Another member kindly pointed out that you can already buy a device similar to what I am proposing (http://www.techluck.com/).  I really should have mentioned this in my initial post  .  Unfortunately, the product from http://www.techluck.com/ does not have a high enough power rating for families in cold places like Northern Europe.  The PDF with my initial post contains the calculations behind this.  Another issue is that the product from http://www.techluck.com/ does not have any system to prevent Legionella growth.

The bacteria growth issue really is mostly for well water nowadays, thanks to the use of chlorine or chloramine in city water specifically to prevent that problem.

[jonroger]

Using upper and lower elements, one can use a mix of power sources, increasing capacity and even addressing any legionella issues.

There is an engineer (search for electrodacus) who has run the numbers and for his case, he finds resistance heating of his house from PV solar cost effective.

[NiHaoMike]

The problem with solar water heating is the poor utilization factor during the summer.



To fix that, add some way to use the solar power for other purposes.

[fourtytwo42]

I have had a PV water heating system for several years, the whole system design was arrived at on a cost/payback analysis.  In my case I had an existing hot water store heated by both an oil boiler and backup 3Kw immersion heater (that I never used). The advantage of PV to me was the ability to use the existing hot water store and immersion heater, installing direct solar/water heating with a new special purpose tank would have been very expensive and difficult in comparison.

My PV system started at just 500W as a trial but is now 1Kw, this is more than adequate for us two people in summer and means the oil boiler is only used in winter, originally surplus energy was wasted but that now feeds a grid-tie with an export inhibitor so it supplies whatever house load there is (mostly fridge/freezers & cooking).

Of course being an engineer I was able to design and build all the electronics without having to rely upon commercial products that may or may not exist, the panels I bought ready made as I don't think i could get the same durability with DIY and frankly I think hard to beat on cost if you spend time looking for a good deal.

Payback, already and some just on saved oil costs 

[Seekonk]

techluk, GAG. Think their little capacitors will be dead in about 4 years.  I heat water with PV off grid in summer very successful from panel voltage at power point (no batteries) and I only use excess energy not used for charging. Store energy in capacitors and PWM it to heater to maintain power point voltage. Helped one guy with a microprocessor, six heating elements switched in and out with one always being PWM. That keeps capacitor bank low power. In Europe you can get tanks with two 208/380V 6000W elements for 6 elements in total.  These can be switched in and out simply by monitoring the current for pseudo MPPT which is close enough. PV is quite efficient when you consider a  direct thermal system is worthless in the summer and PV could be powering something.

[NiHaoMike]

My PV system started at just 500W as a trial but is now 1Kw, this is more than adequate for us two people in summer and means the oil boiler is only used in winter, originally surplus energy was wasted but that now feeds a grid-tie with an export inhibitor so it supplies whatever house load there is (mostly fridge/freezers & cooking).

Of course being an engineer I was able to design and build all the electronics without having to rely upon commercial products that may or may not exist, the panels I bought ready made as I don't think i could get the same durability with DIY and frankly I think hard to beat on cost if you spend time looking for a good deal.

If you have any dedicated freezers, install a relay to run them more when the PV is producing, good way to add thermal storage with little added cost.

And yes, building your own PV solar panels is only worth it for really small ones nowadays, much less than 100W.

techluk, GAG. Think their little capacitors will be dead in about 4 years.  I heat water with PV off grid in summer very successful from panel voltage at power point (no batteries) and I only use excess energy not used for charging. Store energy in capacitors and PWM it to heater to maintain power point voltage. Helped one guy with a microprocessor, six heating elements switched in and out with one always being PWM. That keeps capacitor bank low power. In Europe you can get tanks with two 208/380V 6000W elements for 6 elements in total.  These can be switched in and out simply by monitoring the current for pseudo MPPT which is close enough. PV is quite efficient when you consider a  direct thermal system is worthless in the summer and PV could be powering something.

Probably would be better to run all elements in PWM at different phases. That would multiply the effective switching frequency and make it pretty easy to only need a film capacitor for decoupling.

[SolarMan]

There is an engineer (search for electrodacus) who has run the numbers and for his case, he finds resistance heating of his house from PV solar cost effective.

Pure gold!  The Electrodacus website has a wealth of information, thanks very much 

[SolarMan]

six heating elements switched in and out with one always being PWM. That keeps capacitor bank low power. In Europe you can get tanks with two 208/380V 6000W elements for 6 elements in total.  These can be switched in and out simply by monitoring the current for pseudo MPPT which is close enough.

Pure genius !  Minimizing (or even eliminating) the need for electrolytic caps (and instructors) is a major plus 

[SolarMan]

Payback, already and some just on saved oil costs 

Great to hear that the system has performed well for you! 

Of course being an engineer I was able to design and build all the electronics without having to rely upon commercial products that may or may not exist

Would you consider sharing a schematic (assuming of course that you have no plans to develop this commercially)?    I assume you had to use some large electrolytic caps seeing as you used only a single 3kW heating element.

[SolarMan]

Probably would be better to run all elements in PWM at different phases. That would multiply the effective switching frequency and make it pretty easy to only need a film capacitor for decoupling.

I really like your suggestion here  .  Anything to get the big electrolytic caps out of there is a major plus  and this sound like a very elegant, efficient, and cost effective solution

[Seekonk]

You can simply have one heater element connected all the time and monitor total current from the solar array.  When current approaches power point voltage  (IxR) of array, connect next resistor in parallel.  Panel is current source so current will stay the same or lightly higher because closer to ISC.

If you work out the numbers it will drive you nuts. At the time you switch, there is an instant significant power drop for the near current increases.  Overall it works better for daily total and gets smaller the more resistors.   If you just PWM one heater a couple hundred watts it then becomes seamless.


Multiphase.....Yea, but you can't even get the solar muggles to do simple things. Simple PWM is cheaper and probably more reliable.

Electrodacus (the real MATRIX man) is an engineers dream. Fine if you design your whole house around his system. For someone trying to use standard heating elements it isn't very practical. Odd that it is the only time he cares about MPPT. But, I admire his purity of thought and he takes a real economic analysis of wasting electricity. Sometimes it is just cheaper to waste it. I just find it difficult to believe his purchasers have any real understanding of what they are buying.

[richard.cs]

I am thinking of doing something similar, but am considering skipping the MPPT in favour of oversizing the panels and optimising the load impedance for low-sun winter days - under those conditions MPPT would allow me to heat the water more quickly in the summer but so what if the tank gets hot a couple of hours earlier. This kind of system is best designed (in terms of power points and panel angles) to be useful for as much of the year as possible rather than to produce the highest annual total. I am not fully settled on this, but the increased reliability appeals to me. If I do go the MPPT route I'll build a suitable buck without electrolytics.

Something to be careful of - the standard thermostat in most immersion heaters will be extremely unhappy on DC, best case it has a short and brutal life before welding closed, worst case it catches fire the first time it opens under load. Re-wiring it to control the current indirectly is strongly recommended, though note that it is an important safety device* that prevents boiling in the tank so I'd be uncomfortable with it just becoming a control input to the MPPT. Mine will probably end up with the thermostat controlling a hefty DC rated contactor.

* https://www.hse.gov.uk/services/localgovernment/hotwatersystems.htm "The alert follows two strikingly similar fatal incidents in the last four years where a large volume of near boiling water poured through bedroom ceilings onto occupants sleeping below."

[fourtytwo42]

Payback, already and some just on saved oil costs 

Great to hear that the system has performed well for you! 

Of course being an engineer I was able to design and build all the electronics without having to rely upon commercial products that may or may not exist

Would you consider sharing a schematic (assuming of course that you have no plans to develop this commercially)?    I assume you had to use some large electrolytic caps seeing as you used only a single 3kW heating element.

The design would be very particular to the panel configuration and power even assuming a single 3Kw/240V heater element.
There is certainly no need for electrolytic capacitors, I use a 125Khz switching frequency with film capacitors.
I use both buck and boost converters, the buck is used when insolation is low and the heater would otherwise overload the panel and the boost when insolation is high and the heater would not present enough load at the panel voltage (in my case around 120V).
I would recommend avoiding low voltage small or parallel connected panels, with a high voltage heater the higher the panel voltage the better, if you go over 240V then you only need a buck converter making life a little simpler.
You need a bit of experience to make high power converters and there's some high frequency high power inductor construction too

Depending upon how much hot water you use it's easy to end up with surplus solar power in the summer so you might want to consider integrating a GTI too and that just makes everything more complicated..........

There are many ways to crack an egg, mine is only one, definitely crank up your spreadsheets before spending money

[tszaboo]

Just so you are all aware: Solar panels will generate high voltage DC and high current. So you need special fuses, DC relays (sometimes called contractors) to safely handle it. DC rating of a fuse is different than AC rating. Shorted cable will weld itself together, and then it sets stuff on fire, just like the recent Tesla Solarcity and Walmart (maybe) case.
And installers apparently dont take any responsibility, if you touch their system, or just plainly dont install the system.

[Red Squirrel]

This is something I've given some thought myself as I want to eventually live off grid.  Electrical storage (batteries) is expensive, but thermal is cheaper.  So it does make sense to try to use extra PV to simply store as heat once your batteries are topped up.  The trick is, you don't really need the heat as much in the summer months as you do in winter.  Though you always need hot water so simply doing hot water is simple enough so if the rest of this is not feasible then hot water it is.

But what would be really cool is if there is a way you could store the excess power generated throughout the summer and use it in winter.  I'm not sure how feasible that would really be but I'm thinking something the size of a basement under ground with crazy insulation like R100, and you use extra PV in summer to get a liquid or material (sand? Salt?) to 100's of degrees.  You then have a heat exchanger with glycol or other liquid that can be used to extract that heat as needed.   What I'm thinking, and I would be curious to experiment with this at smaller scale, is inductive heating.  You have a pipe coil that acts as the induction heater with a big chunk of metal in it, maybe a long rod so the heat can rise up - maybe several of these spread out but could maybe have just one right in the middle, and the entire thing is filled with heat retaining material.  The pipe coil would have liquid circulating in it so it does not get too hot and melt and you just run the induction heater at the highest rate that can be sustained while not using battery power.  Reason I'm thinking induction and not just an element is I think an induction heater would be able to reach much higher temp, to the point that it can get metal to be red hot.    So imagine this at a very large scale, and with proper insulation, I wonder how long that heat could be retained for.    The challenge would be to not melt the actual heat exchanger and/or coil so you don't want to exceed that temp.  You would need some kind of temp sensor in there, but any kind of wiring or sensor would melt at these temps.   I suppose you could measure the temp of the liquid in the exchanger as you would need to circulate it at least a little bit even in summer, so you could stop the process if it exceeds a certain temp.   We'd be dealing with super heated steam at this point though, so it's something that could get dangerous very fast. You could do molten salt but trying to find pumps etc that can work would that would be hard.

No idea how well this idea would really work though compared to simply heating water with an element.  Just something I randomly was thinking about the other day.  A smaller scale version could be a simple metal tank and it could perhaps store enough heat so that in the morning you can have heat when showering.  Ideally it would be stored within the living quarters like in the basement or something, so heat losses would not necessarily be wasted.


Perhaps a simpler thing to do, since excess PV is always in summer, is to use it towards A/C.  I wonder if they make A/C units that use soft start DC motors for the compressor, so you don't have to worry so much about the startup draw.   Bonus: Use the hot side of the AC unit to warm water up for domestic use. 
 

[jh15]

Wat is the recent Tesla fire? I have a gas powered snowblower in the garage, and was more worried about it leaking than my S. 

[oPossum]

He is referring to the Solar City (Tesla) files at Walmart stores. There have been a number of Model S fires as well... https://www.thedrive.com/news/28420/parked-teslas-keep-catching-on-fire-randomly-and-theres-no-recall-in-sight