Solar water heater

[charliefleet]

A 12V panel will put out 17-18V at power point. You would want a lower wattage 12V element than the panel wattage.

element 300W /12V = 25A 12V/ 25A = .48 ohms
panel 300W /17V = 17.6A 17V/ 17.6A = .96 ohms 8.44v x 17.6A = 148W with 300W 12V element
element 300W /24V = 12.5A 24V/ 12.5A = 1.92 ohms
element 600W /24V = 12.5A 24V/ 25A = .96ohms 300W

For the ACTII AC7391 PWM MPPT Heater Controller... it mentions matching heater to panels. Knowledge base (actii.pl/knowledge/?p=14). Bearing in mind this is a Polish translation, it says "Remember that if you use a stronger heater or a weaker one, it will not be corrected and may overload the PWM controller, and after it leads to its damage after several months of such work". So the inference is seeking to roughly match resistance of panels and heater load.

In my case, the panels are 41.3v (Vmp) @ 11 Amps (Imp) = 3.75 Ohms each. Connected in series = 3 x 3.75 = 11.3 Ohms.

For the heater. I have an existing 3,000w 230v AC element. This draws a current of 13 Amps. Hence resistance is 17.6 Ohms.

I'm not sure whether this represents a significant mis-match?

And note that on first impression you might look to fit a 2Kw element instead, but the resistance of such an element would be 26.5 Ohms - an even greater mis-match.

I'm going to stick with zero change to the 3kw element and monitor to see how the system copes.

Obviously i could simply buy a lower voltage element that simply gives the best resistance match... something like a 1.2Kw at 120v would be perfect as this gives 12 Ohms.

 

[efficientPV]

I actually test my board with 40A. This is a RC circuit and higher resistances which create longer discharge times are much better. This is the reason these circuits are designed to operate with higher voltages. If the capacitor bank is large enough for current, it matters less. In your calculation, to get the maximum transfer the resistance should be lower. This is because with higher panel temperatures the voltage will be lower, yet the current will remain the same. I assume you are in Europe as 3,000W elements are uncommon here. In the end, you got what you got and it will work fine. It only becomes an issue at very peak power times which will not be often. At those times the PV voltage will rise over power point and it will produce more power. Heating is the power of squares and that small voltage rise over power point will produce more power.

Everyone wants simple and they only look for one or two items in a specification of a heater element. The actual ohms matching has to be calculated. In PV direct connect, the so called experts like POZ are just wrong. Ideal resistance is not what everyone thinks. That is true for absolute peak times which happen less often than everyone thinks. Studies have shown that "Ideal Resistance" can be almost doubled for much higher average daily & monthly production.

 

[charliefleet]

I actually test my board with 40A. This is a RC circuit and higher resistances which create longer discharge times are much better. This is the reason these circuits are designed to operate with higher voltages. If the capacitor bank is large enough for current, it matters less. In your calculation, to get the maximum transfer the resistance should be lower. This is because with higher panel temperatures the voltage will be lower, yet the current will remain the same. I assume you are in Europe as 3,000W elements are uncommon here. In the end, you got what you got and it will work fine. It only becomes an issue at very peak power times which will not be often. At those times the PV voltage will rise over power point and it will produce more power. Heating is the power of squares and that small voltage rise over power point will produce more power.

Everyone wants simple and they only look for one or two items in a specification of a heater element. The actual ohms matching has to be calculated. In PV direct connect, the so called experts like POZ are just wrong. Ideal resistance is not what everyone thinks. That is true for absolute peak times which happen less often than everyone thinks. Studies have shown that "Ideal Resistance" can be almost doubled for much higher average daily & monthly production.

Ok thanks for the comments - interesting that the real-world resistance is likely to be lower (at least on warmer days)... and hence to match without wasting Watts the heater resistance should be lower.

And that makes sense with RC circuits; using higher voltage. I'll hunt for a reasonably priced lower Ohm element...

Ebay 4.5kw 220v element

 

[Vigo]

A 12V panel will put out 17-18V at power point. You would want a lower wattage 12V element than the panel wattage.

element 300W /12V = 25A 12V/ 25A = .48 ohms
panel 300W /17V = 17.6A 17V/ 17.6A = .96 ohms 8.44v x 17.6A = 148W with 300W 12V element
element 300W /24V = 12.5A 24V/ 12.5A = 1.92 ohms
element 600W /24V = 12.5A 24V/ 25A = .96ohms 300W

So i think i have some idea what you are getting at, but the math you posted is not helping me get closer to it..

So here's my background knowledge which may need to be corrected to 'get this'. A water heater element is rated in Watts at a certain Voltage. From that we can calculate amps and ohms if we choose to. Normally speaking feeding an element a higher voltage than what it is rated at, would give you more Amps and thus more Watts of heat delivered. Although you cannot exactly 'overheat' the actual element itself because it is immersed in water and it would take a tremendous amount of energy to raise the heating element past whatever the boiling temp of water is at the system pressure (spitball maybe 280f in a tank @40psi? could be calculated), it MAY be possible to overheat your connection to it because the terminals you're connected to cannot handle much more current without themselves overheating, even though the actual immersed heating element is still at an acceptable temp.

But that's with an unlimited current source. With a current-limited source such as a solar panel, it will never deliver more than whatever the short-circuit amps or Isc is, no matter what. So if you took a 70 volts open circuit panel with an 8 amp ISC and hooked it to a 12v element, you would get.. 8 amps. And from what i've seen, when you actually short circuit a solar panel its voltage drops close to nothing. So as long as your PV's isc is lower than the calculated current of your 12v element (W / 12v = A) you are in a very safe zone regardless of how much higher the voltage open circuit of the panel is from the rated voltage of the heating element. It will simply drop the panel voltage to whatever it can make at it's ISC, and your true power delivery will be that voltage, times ISC = watts of heating. Which is probably way way under the element's rating because the panels can't make very much voltage at ISC.

So to me it seems like having your solar VOC be over the rated voltage of your element is completely fine as far as safety, until your solar is WAY higher in wattage than the 12v element is. The problem is that it will be inefficient because you will only be making a fraction of the solar panels' wattage when you flow so much current out of it that its voltage drops to a very low level.

So the way that a 24v element might be better for a 12v panel than a 12v element, is because since a 24v element at the same wattage will have a higher resistance, you will not be getting the current as close to the panel's ISC and it will be able to make a higher voltage at that current level, meaning your total Volts X Amps = Watts will be a higher number. You'd be operating the solar panel far more efficiently at that 'load' and be able to get closer to its rated power.

Is that a correct way of thinking of it?

 

[efficientPV]

The ultimate goal is to heat water and as soon as possible. Many battery voltage elements are considered high heat density which means they already get hotter. In calcium rich areas this leads to scale build up on the elements and possible failure. Bubbles on element slow heat transfer and make the element run hotter. Maybe that will never cause an element to fail. It does lead to more sediment buildup in the tank. You may have heard popping at some time when boiling water. They sell longer low heat density elements just for that purpose. I only include this for background.

Think about low resistance elements. At a half ohm, the resistance of wire switches connectors can be significant. That energy is lost. It actually doesn't change how much heat the element produces because the heating element is already operating at Isc. At least not till the panel reaches the voltage where Isc approaches Impp. You are correct that the panel will be operating at Isc for the given sun intensity almost the entire day. Volts and amps are just not interchangeable. With high currents you face higher costs and losses.

It is all about getting the maximum power transfer for the lowest cost. With a lower resistance heating element adding more panels will produce more heat but it will never catch up to the potential. 12V elements are designed for batteries, not panels. There are too few options for 12V panels. Anyone wanting to heat with panels should look for dual elements with shorting bars which can give you three options for resistance. That allows for later expansion and power options. In the example given the 600W 24V element would give more power for most of the day if only half the element was connected due to lower current. A timer or some other control could switch in the second element for two hours mid day when the higher currents are expected. Studies have shown that even a single element with higher resistance gives more total daily power.

Just find a real world 12V element which makes sense with any solar panel.

 

[charliefleet]

Ok so I've completed the install, and it's been a few weeks' operation now... very happy with the results from the MPPT/PWM device.

As a reminder, I'm using 3 X 455w panels on my South facing detached garage roof. This feeds, via the appropriate DC fuses and breakers, into a consumer unit housing the MPPT/PWM unit.

I've not yet changed the standard 3kw 240v immersion element. And it seems to work fine, although I suspect it isn't as efficient as it could be if I changed for one that matches the panels' resistance better.

I'm getting about 900w in bright sunlight dumped straight into the hot water cylinder.

On bright cloudy days this drops to about 300w. And gloomy days it's around 100w.

It's basically now a fit and forget system. The gas boiler kicks in if the water temperature drops too low for any reason.

I can see that on long sunny summers days I will be making use of the relay output that triggers when the temperature sensor reaches a set value. And I will then rig up a suitable small battery system to power something else useful!

Thanks to the help and advice of folks above.

 

[willem wikkel spies]

Whats up all, newbie here.
1st post.
more questions and more info needed.
We are also looking into the water heating directly from solar panels.
Yes i know about the difference between DC and AC and contactors and arching and ac voltage via thermos stat and that it cannot handle DC switching etc.
That we can sort out via a heavy duty 12 volt activated dc contactor which we will just need to monitor the contact points for now.
using external temp sensors to activate the contactor when temperature of water is reached.

In sunny south africa we commonly use 150 liter "geysers" or hot water tanks and they are driven by 230 vac / 3000 watt elements.
we only have 1x element per geyser. our geysers are normally placed horizontally.
it must be added, our electricity is never smooth in that the lowest i have measured has been 198 volts ac and the highest was around the 247 volts ac. our HZ does jump up and down at times.

I have done quite a lot of research into water heating directly from solar panels and i understand that one needs to match volts and amp rating to achieve results with resistance included.
Technology has jump much higher and quicker as we expected.

I have looked at Divid poz videos as well, but all is not stated clearly, so we are all left in the dark to a agree.
his calculator we can download, i did look at, and there was no mention of our 3000 wat elements. most liklely that he need to make a sale by getting people to buy other elements.
in any case, more research and his calculations and i managed to get a solar array that might work. question is, can i believe it?

we have either canadian 550 watt or JA 550 watt solar panels that we can use.
there is not much difference between the two solar panels even price is the same.
JA vmpp is 41.96 volts
impp is 13.11

so when using 4x 550 watt Ja panels in series we should get around 2200 watts of power, sunlight is not a concern, cold winters is only around minus 4 at best during winter.
The 3000 watt 230 vac element resistance is around 13.04 ohms, when calculated, but not tested. internet says resistance should be around 17.63 ohms.
Total VMPP is 167.84
MPP resistance is 12.80.
so, will a 4x panel string in series work right?
without the need to fit a mppt.
this calculation is done under the standard tesing conditions.
if the numbers of normal operation conditions are calculated, the system still says that the 4x panel in series will be good.
so please educate me, thanks.

EfficientPV, can your mppt handle the voltage of 4x panels in series?

 

[efficientPV]

All depends in how you make it. I just did a build using a board less than $3 from Ali. That board can only take Voc less than 55V. Change the FET and it can go higher. Don't listen to Davis Poz.

 

[willem wikkel spies]

All depends in how you make it. I just did a build using a board less than $3 from Ali. That board can only take Voc less than 55V. Change the FET and it can go higher. Don't listen to Davis Poz.

Cool thanks.
So will 4x550 watt in series be enough to run a 240/ 230 ac 3000 watt element directly?
We do not have any controllers for this.
thanks.

 

[efficientPV]

Divide array Vmpp voltage by Impp current. That gives you "ideal" resistance.

Heating element resistance is 3000W / 240V = 12.5 amps then 240V / 12.5A = 19.2 ohms

Ideal resistance is only good for an hour a day. Maybe only one day out of three.

Choose a realistic panel current will provide more power daily. Choosing 3/4 panel current for your calculations is a little more realistic. Even that will be an emotional stress for many who can't get off the idea of absolute maximum power. The chart below indicates the ideal resistance can be doubled and still get more average daily power.

 

[Vigo]

Ouch.. David Poz leaves info out in order to sell things, and dont listen to David Poz? That's harsh and accusatory, in my opinion he is honest and helpful and just like the rest of us he learns along the way and at any given moment his knowledge is limited to what it currently is, until he learns the NEXT thing.. just like the rest of us! Noone should be expected to be perfect just because they are in the public light. I have learned some things from his videos but even when i havent ive found his attitude motivational and think he makes the subject approachable for people that are not CURRENTLY so technically minded already.

I think the hard part about understanding using solar panels to power a heating element is that it's not obvious what the voltage will drop to once the solar is hooked to the heating element. That's what makes it unintuitive to know how many watts will result.

 

[willem wikkel spies]

Cool, thanks for your reply.
still trying to understand the gibberish but im sure i will get there.
In any case, before we go further down the ribbit hole.

just some other info that might be relevant to this.
we have 1st hand experience in heating elements " portable soldering iron" even it is small it is still relevant to understanding all of this.

so a 15 watt soldering iron runs off a 12 volt battery.
it has a resistance of 8.4 ohms.
so in essence pulling around 1,42 amps on 12 volts.
the soldering iron will only get so hot and nothing more, when soldering thicker wires, heat loss is quite drastic and one needs to allow the soldering iron to heat up again with the wire which will allow one to start soldering.
eventually one will be able to start soldering.
when the 12 volt battery is depleted and the battery voltage drops, one is not successful in soldering as the heat present is not enough.
We have in the past connected 2x 12 volt batteries in series, this would cause the soldering iron to heat up to a glowing red condition and its "too hot" to solder, lucky we have not left it on for too long as we might burn out the heating element.
resistance stays the same, but current increases as voltage increased.

so in water we have the exact same application, to a degree.
i do not want to discuss davids video here as is does not make sense to me.
what i would disagree on is that:

even if you placed the solar panels in parallel as claimed, the current will increase to a degree, but one can only feed so much current into the heating element.
all electronic components will only use the power they need in order to operate.

so, in my situation, the heating element has a power output of 3000 watt on 230 volts ac, it has a resistance of 16,2 ohms "checked that by testing one this morning", not all have the same resistance.
so current drawn by the element on 230 volts ac amounts to 13.04 amps and thus the output of 3000 watts.

im not including higher resistance due to heat, lets just leave that one out for now.

now, because the element only runs on 13 amps, my question would be.
can one force more amps through the heating element?

the reason for this question and equation is as follows:

if maximum panel voltage in 4x 550 watt solar panels in series @ 41,96 volts dc would give you 167.84 volts dc.
on this string, the maximum current would be 13.11 amps
if this was connected to a 3000 watt element what would happen?
Calculated i would only get 2200 watts at maximum operation.
the problem with this is that, in order to successfully heat the water to a certain degrees, one would need the heating element to run at optimal performance for a period of time.
because the "soldering iron is not getting all the voltage needed it does not get warm enough" and the same principal should apply with the water heating element. so your water will not get hot enough.

so would it be better to scale it all up, to 5x solar panels in series?
we would then get 209.8 volts dc @ 13.11 amps.

but should we not figure the RMS into the equation as well?

its getting complicated, but im sure we will get there.
thanks

 

[efficientPV]

Panels are current producers based upon sunlight. The voltage of a panel is the result of load resistance within the useful range. Clearly you are uncomfortable with electrical calculations. I don't like spreadsheets because they isolate people from the actual math making it all a distant concept. Just sitting down and doing a bunch of calculations can give you a much better feel. Power is a function of a square where small changes make a big difference. The chart below demonstrates the problem of using a fixed resistance. I'm not going to fix up the spacing. Look at when the panels can produce only 50% of rated current. The power doesn't drop to half as expected, but one quarter.

Power Point
% Rated Increase over Voltage Power With
Panel Amp Direct Connect 10 ohm 10A panel Direct Connect

100% 0% 100V 1,000W
90% 10% 90V 810W
80% 25% 80V 640W
70% 50% 70V 490W
60% 67% 60V 360W
50% 100% 50V 250W
40% 250% 40V 160W
30% 333% 30V 90W
20% 500% 20V 40W
10% 1000% 10V 10W

 

[efficientPV]

Ouch.. David Poz leaves info out in order to sell things, and dont listen to David Poz? That's harsh and accusatory, in my opinion he is honest and helpful and just like the rest of us he learns along the way and at any given moment his knowledge is limited to what it currently is, until he learns the NEXT thing.. just like the rest of us! Noone should be expected to be perfect just because they are in the public light. I have learned some things from his videos but even when i havent ive found his attitude motivational and think he makes the subject approachable for people that are not CURRENTLY so technically minded already.

I think the hard part about understanding using solar panels to power a heating element is that it's not obvious what the voltage will drop to once the solar is hooked to the heating element. That's what makes it unintuitive to know how many watts will result.

memento quo nunc es

Counselor Verspasian said “Remember where you are" to Emperor Trajan
who was betrayed by his own men, buried up to his neck in sand.

 

[willem wikkel spies]

Panels are current producers based upon sunlight. The voltage of a panel is the result of load resistance within the useful range. Clearly you are uncomfortable with electrical calculations. I don't like spreadsheets because they isolate people from the actual math making it all a distant concept. Just sitting down and doing a bunch of calculations can give you a much better feel. Power is a function of a square where small changes make a big difference. The chart below demonstrates the problem of using a fixed resistance. I'm not going to fix up the spacing. Look at when the panels can produce only 50% of rated current. The power doesn't drop to half as expected, but one quarter.

Power Point
% Rated Increase over Voltage Power With
Panel Amp Direct Connect 10 ohm 10A panel Direct Connect

100% 0% 100V 1,000W
90% 10% 90V 810W
80% 25% 80V 640W
70% 50% 70V 490W
60% 67% 60V 360W
50% 100% 50V 250W
40% 250% 40V 160W
30% 333% 30V 90W
20% 500% 20V 40W
10% 1000% 10V 10W

Cool.
i have been working out the numbers to a certain degree.
im getting there.
slowly the lights are coming on.
so we have a fixed resistance in the water heating element, "lets just keep it there for now as we all know resistance will rise with temp" and due to light fluctuating we have less power to supply the heating element thus output drops.
this means, you will still have power on the water heating element, but not enough to heat it at the temps you want it.
So one would aim to get the water up to temp and then keep it there, so should clouds/ bad light conditions occur, the water is already up to temp.
so one will need to factor in bad light days, and in essence, the system will have to be oversize.
with this, one will loose power generated and never used.

joh, at 50%, 1/4 power at is quite low.

so why, in this day and age, have they not made a mppt for higher dc voltages?
where is the problem?
why has no one build a mppt for the 200 to 250 volt range?

switching high dc voltage will burn contacts, i do know that.
i know its not ideal, but one could use a very high rated contactor to switch the dc. it will be consumed over time and can be replaced when needed, but for now i would keep it out of the discussion.
one can activate the contactor via temp switches mounted on the water cylinder. a 2nd form safety.

 

[willem wikkel spies]

Ouch.. David Poz leaves info out in order to sell things, and dont listen to David Poz? That's harsh and accusatory, in my opinion he is honest and helpful and just like the rest of us he learns along the way and at any given moment his knowledge is limited to what it currently is, until he learns the NEXT thing.. just like the rest of us! Noone should be expected to be perfect just because they are in the public light. I have learned some things from his videos but even when i havent ive found his attitude motivational and think he makes the subject approachable for people that are not CURRENTLY so technically minded already.

I think the hard part about understanding using solar panels to power a heating element is that it's not obvious what the voltage will drop to once the solar is hooked to the heating element. That's what makes it unintuitive to know how many watts will result.

Please understand my viewpoint.
i have looked at the video, it is incomplete and not all facts are there, for you and i to make the right informed decision.

Yes, he has to make a sale, and in order to get the sale, the system he uses is not a standard system used world wide.
in south africa we might have 10 million homes, most running on 150 geysers using 3000 water water heating elements on 230 vac power.
the listed heating elements does not include our 3000 watt elements.

ok rather join the crowd and make the sale of 10 million heating elements, see where the money is flowing.
i have no beef with David and dont think that we need to discuss it further.

 

[efficientPV]

Cool.
i have been working out the numbers to a certain degree.
im getting there.
slowly the lights are coming on.
so we have a fixed resistance in the water heating element, "lets just keep it there for now as we all know resistance will rise with temp" and due to light fluctuating we have less power to supply the heating element thus output drops.
this means, you will still have power on the water heating element, but not enough to heat it at the temps you want it.
So one would aim to get the water up to temp and then keep it there, so should clouds/ bad light conditions occur, the water is already up to temp.
so one will need to factor in bad light days, and in essence, the system will have to be oversize.
with this, one will loose power generated and never used.

joh, at 50%, 1/4 power at is quite low.

so why, in this day and age, have they not made a mppt for higher dc voltages?
where is the problem?
why has no one build a mppt for the 200 to 250 volt range?

switching high dc voltage will burn contacts, i do know that.
i know its not ideal, but one could use a very high rated contactor to switch the dc. it will be consumed over time and can be replaced when needed, but for now i would keep it out of the discussion.
one can activate the contactor via temp switches mounted on the water cylinder. a 2nd form safety.

If you go thru the calculations for ideal resistance, the resistance changes are quite dramatic. The system shouldn't be designed for the best days. You get too much hot water on those days and hot water is not easily stored. My wife doesn't want to hear excuses why she can't use the dishwasher. My water heaters turn off by 10:30 in the morning even with my minimal system. Often not as much power is needed than everyone expects. I roll my eyes when they say they need over 1,000W on the element. China doesn't build these systems because they couldn't sell them if they did. Well, at least not in America. Making hot water is just for those who are bored, and they don't want to spend any money. Here is a little board I got from AliExpress for less than US$3. It is limited to 55V because it was designed as a 12V inverter. They could put much higher voltage FET in for little money. China is selling water heaters with controllers built in. I put one of these on two 12V 100W panels in series for 40V. I used a 200W 12V heating element which was a little more than a half ohm. Ideal would be about 6 ohms. Transferred power to that insanely low resistance like a charm. And I think stand alone water heater systems are an insane idea. Why devote panels just to water heating? I operate diversion in parallel with MPPT controller. Ebergy should always be going somewhere.

 

[wattmatters]

Making hot water is just for those who are bored, and they don't want to spend any money.

This. Home water heating has many sound options one can buy already. I admire those clever enough to build their own but it's not for me, regular off-the shelf products do just fine, meet all necessary regulatory requirements and are not overly expensive in the grand scheme.

Ours is a regular electric resistive element (3.6 kW / 240 V AC) heating a 315 litre tank. Power used to be supplied via a dedicated off-peak grid power supply but nowadays I use a smart PV diverter which adjusts power delivery based on available excess grid-tied solar PV. It can boost from the grid if necessary.

In 2023 our water heater consumed 1893.1 kWh (5.2 kWh/day), of which 99.4% was supplied via the solar PV diverter. Grid boosting was just 12.3 kWh for the year (and that wasn't actually required, I was just being absolutely certain we had a full tank of hot water while the in-laws were staying over the holidays).

In our case a heat pump water heater is not an easy or cheap installation (a function of the home's design) but had we used a heat pump then water heating electrical energy consumption would be in the 450-500 kWh for the year, something like 1.3 kWh/day.

 

[Vigo]

Im still finding that my 4500w 240v element, rewired to 120vac (~1100w?) and running only ~4 hours a day, is working for my family. Once i have more excess solar ill simply increase the on-time of the timer switch and potentially install a thermostatic mixing valve and crank the thermostat, but i dont see ever doing anything else to get hot water unless I build a water-to-air heat exchanger system to heat the air inside the house from hot water which i would probably try to heat from exhaust heat of a generator. Even then, that's mostly because i find the ~20-30% efficiency of a generator conceptually offensive. Not really because i need another heat source.

 

[willem wikkel spies]

here is another string of questions, where we need to get the answers first before we can proceed with this project.

if we take a heating element of 2000watt, which is used on 230vac, we get a current draw of 8,6956 amps.

and we can calculate this heating element has a resistance of 26.48 ohms.

i have seen that guys advised, running lower voltages but higher current.
but the big question is, what limits the current on a heating element?