Water heating with solar

[ksmithaz1]

Not exactly "dedicated", as it's a secondary use case, but I have 4x 300w panels 2s2p (about 70 vdc) that I use to run my solar direct well pump through its controller. It doesn't take much energy to keep that low producing well pumped dry, so I have excess solar available from that array...

I have a 110vac receptacle, that I can switch from the pump to the receptacle in a breaker box, wired directly to the pv panel output, so 70vdc at the receptacle. I have a 5 gal single element water heater that I sometimes run off that array. (I can't remember if I used a 1000w 48v or 110v element, but I think 110v as I can also just plug that WH into the grid.) When the sun is good, the water heats up in a few hours. The WH has a standard WH thermostat, so it shuts off at the set temp.

Got it. You have a "Fun project" for already dedicated solar panels. OTOH the OP asked:

Hello, I currently have excess solar capacity (3.2kw) and I am wondering if anyone has experience using a dedicated water heater powered directly by dedicated solar panels. My option would be to use a 24 or 48v heating element and connect directly to a single inverter. Maybe a bad idea, but I wanted to check ,.. I have the ability to use a propane on demand heater as a backup. Any thoughts are appreciated ... Thx

So, ... "Excess solar" apparently some "extra" panels laying around. IMNSHO I'd buy more AIO before I'd cobble something together to heat water. An awful lot of effort to dedicate panels for a relatively small load. Hooking the same panels to a small $500 AIO would be more expensive but vastly more useful. An even better solution would be to scale up the existing system with more inverter/mppt/panels, then use the excess wherever you want albeit with some conversion losses.

I mean is it free to retrofit heating elements and control mechs to connect a HWH to a bunch of solar panels you have laying around? Without knowing what is existing, again tossing the panels onto another parallel AIO, or MPPT, increasing the battery and scheduling HWH is more flexible.

 

[efficientPV]

I think you made my point, you already spent a lot of money. This discussion is not applicable to you. Everyone sees things as how they relate to their own situation. Somebody wants to get rid of 3KW of power instantly. That is just not easy or cheap to do. HPWH only use about 600W of power and run a long time. It takes batteries and inverters to run them. I have this guy up in Michigan with 600W of panels and a 100AH lithium battery. Now he has hot water and he heats a cabinet in the winter where it is stored. No expansion needed in his system.

 

[martinwvogt]

I heat 2 50 gallon electric water headers (in series) with my excess PV power AFTER my battery is full
I heat them to 125deg F and it usually is complete by 2-3pm
I use 3000 watt 240v coils in both, upper and lower heaters. All cascaded

here are the 2 links in my build thread

theory

implementation

works great, very little nat gas is used in the summer in the tankless water heater - majority of the water coming in is higher than the tankless is set for

I also have dial mixing valve to ensure the water is not too hot going to the facuets

Watched the video, pretty awesome!

 

[OzSolar]

It appears you're confused on the theory and your approach is very inefficient. The power is always available, its just takes what is not consumed in the moment (ie., no loads, batteries charged).

There are losses in the charge controller, the inverter and what do you do when you don't have more battery storage? Also, power transmission is efficient at high voltage -- most inverters/controllers max out at 48V and power is P = VI or P=I2R.

But it's already 100% wasted energy. Who cares if you get 3% or 83% of that back?

I will politely offer that it appears that you are the confused one. Energy that is put to no use has 0% efficiency. Feel free to derive that calculation if you want but it won't survive any sort of inspection.

OP has an array that is doing nothing for them, ALL of that energy is being wasted. Do all the math you want but once zero's show up it doesn't matter how cool your formula is, bad things happen to the overall results.

Installing an electric water heater to preheat the water is simple and reliable. Over complicating things and berating others for not going for your definition of perfect is counterproductive.

 

[Daddy Tanuki]

Here is the basic theory. Let's start at the panel current vs. voltage curve. The maximum powerpoint is when the voltage is highest and current is highest. With no load on panel, the voltage goes torward open circuit voltage. When a load current is applied, the voltage drops. A good assumption is that when panel voltage is above maximum power point voltage (Vmp) there is excess solar to be harvested.

This is how the circuit works: it monitors the panel voltage and when it goes above Vmp it turns on the heating elements. If the solar charger is also trying to get some energy the panel voltage will drop below Vmp and the heating element turns off and the charge controller uses MPPT to find the optimal power point. The heating element circuit is just responding to solar availability and current charge controller consumption. Of course the chosen voltage to toggle the heating element is going to be a little dynamic because of panel temperature coefficient, etc. A toggle voltage slightly above Vmp seems to work well for me.



View attachment 230885View attachment 230886

I designed my version to handle my panels (above Victron VRM screenshot) which operate at a Vmp of ~168V. I am running two 220V x 3500W heater elements in parallel. The resistance of the each element is ~15.4ohm or ~8ohms total in parallel. At 168V that is putting through ~21 amps and 3500W of heating power. I will be using this to heat a cypress hot tub in southern Chilean Patagonia (currently in testing in the US).

I chose this design to not cycle by LiFePO4 battery bank (no water heater power goes through the charge controller or the inverter). Also as @efficientPV points out you get the efficiency from the high voltage. 48V heating elements are totally useless.

I have 4000W of solar panels, a Victron MPPT Solar Charger (250/100), Cerbo GX and REC BMS connected to a 15kW battery bank (48v). We have so much solar in summer that we use an electric oven with no problems.

wrong post

 

[Daddy Tanuki]

I promote good design in heating water. I'm not limited t promoting any one circuit. Looking at PV array voltage is a good way to determining when excess power is available. Diverting from the array is efficient and cost effective. This method is very cheap and avoids using the resources of charge controller, inverter and battery. Nobody thinks they do, but every successful solar system has excess power which is free for the taking. This is a picture of my camp from the south, see any solar panels? I run a camp and heat water with the excess. I have three water tanks and one is just for laundry. All cycles use hot water. If you can't heat water, "Don't blame Desenex."
View attachment 230879
Here is a shot I took this morning, it is totally overcast and will rain today. Everyone is impressed with big numbers. Look at all those YT videos of the "bucket people", they stick an element in a bucket and connect it to a panel. It is bright and sunny and everything works fine. They never show you the rest of the day when it drops to nothing when the mismatch kills the power. They have no idea that the power is a function of the square of the current. I find this shot impressive, you probably won't.
View attachment 230880I run three water heaters and they all can be operating at the same time. Each has a set priority based on voltage. I test my board at 30A and it has half the capacitors. The board shown I think he will be trying 4KW. I designed a microprocessor for a guy in Canada that had a converted oil tank with five elements. Four were switched in and one had variable power. Dacian "matrix man" and I can't think of the board has a switched element system to match the array. Great idea, just about impossible to implement because factory heating elements don't exist in the values needed or your tank only has two. This is where a proportional system works out better.

OK you have my attention.

This looks like what I was shopping for two years ago but could not find a designer for. I put out requests for RFQ's to multiple places online for people who "prototype circuits" offering to pay for prototype boards that would do something similar (i had requested a temp circuit w/NPT to shut off if temp exceeded xx.xx for safety purposes). all i got back were people telling me if I gave them a circuit design, they could build it...that's not what I put the RFQ for!
I gave up looking for a digital controlled DC answer and settled on keeping it all AC where I could build the circuit with relays and standard temp sensors using the built in controls/circuits when available for the inverters and whatnot.

 

[dougbert]

I had solar thermal panels but in winter (-10F) didn't produce much. SInce my roof top space is limited, I removed the thermals and put in PV panels and gain better heating via electric both winter and summer

 

[efficientPV]

I'd like to use this video to illustrate the issues with matching a PV array to a heating element. This off grid system tries a number of methods to heat water and is still struggling to meet his needs.

@11:07 his open circuit array voltage is 281V. Power point voltage of that array at his temperature will be about 281 X 75% = 210V.

@16:12 the element voltage has dropped the array voltage to 41.7V with a current of 3.035A or about 130W. I consider getting 3A out of a high voltage array still pretty good. With a power point controller the power would be about 210V X 3A = 630W. That is almost a five times increase in power.

In my garage I have a manual transfer switch to change my array from parallel to serial doubling the voltage. This allows me to run that high voltage into my inverter directly to the high voltage section and run the washer right off solar panels. From watching, his system is very manual hands on. By adding a manual transfer switch to go from serial to parallel, the current would double to 6A and voltage to 84V resulting in the heating element producing 500W. That is a big power increase for little effort and this could be automated.

 

[LuxOfTheIrish]

I made one using a RPiPicoW (because it has two cores). One core processed the voltage from the batteries. If the voltage is above 51.6v then it allows more % AC power through using a Triac. The reading here shows the temperature. Every three seconds it will alternate to show the %power of the 3kw total. Because it uses a triac to allow very small changes in power usage it tracks the excess solar really nicely. I might upload more info if anyone is interested.

 

[efficientPV]

That adds new meaning to the term boom box.

How does the inverter like the TRIAC. I saw one commercial diverter that would rectify the AC on one side and PWM the sinewave. It was designed for use with PV inverters to give a better load. They must have done that for some reason. I only do PWM DC diversion. I don't know what it is like to have an inverter.

 

[LuxOfTheIrish]

That adds new meaning to the term boom box.

How does the inverter like the TRIAC. I saw one commercial diverter that would rectify the AC on one side and PWM the sinewave. It was designed for use with PV inverters to give a better load. They must have done that for some reason. I only do PWM DC diversion. I don't know what it is like to have an inverter.

Inverter has no issue with it so it seems . I included a snubber circuit and it has no noise. I was curious about that myself. I also have induction hobs and I was wondering how the inverter would cope with the power factor because it is quite noisy. Inverter copes fine. I've attached the graph from today and a bit of yesterday. I'm away from home so the only use is the immersion. You can see it utilises the solar well (until it heats up of course and the solar goes unused).

 

[efficientPV]

I have built a circuit (based on the design of @efficientPV) that heats water directly from my 4kW (4s2p) solar panels (2 heating elements 220v x 3500W). It sucks the energy right from the panels and never makes it past the SCC and thus never cycles the batteries. It was relatively cheap to manufacture (like 50USD all in). It uses a version of power point (i.e., turns on when the voltage of solar panels is above Vmp) that just uses whatever current available until voltage drops below Vmp. I have it wired in parallel with my Victron SmartSolar controller -- its been running for a week and works great. It's designed to handle upwards of 30 amps at 170V DC.

Here is a picture -- I can provide circuit diagram to whoever wants it. I suggest watching @efficientPV videos on Youtube to get a better idea of how this works -- its quite ingenious as it doesn't use a microcontroller (just analog electrical components -- ressitors, capacitors, MOSFETs, etc).


View attachment 230775

I just got twenty IR2153 from ebay, they were cheap at less than $10. Immediately tested them to see if they were fake and not do the correct shutdown. Good news was 11 were real. 9 would not oscillate at all and were totally dead. Still a decent deal for 11 of them. Most want a couple bucks each. I have had a number of fakes, never had one that wouldn't work before. This is the reason I hesitate recommending this design because they absolutely need to be tested.