PV->hot water

[wattmatters]

I installed a PV diverter for our water heater last year so that we use our grid-tied solar PV instead of importing from the grid. Works well. Heat pump water heaters suitable for our home were too expensive for us, our hot water energy consumption (~5 kWh/day) isn't high enough to justify one.

 

[efficientPV]

@Stewfish Sorry, I'm not good at doing things wrong. Plenty of others here are good at that. The trouble with direct connect is that it performs worst when you need it most. Yesterday it was cloudy and rain and I was still able to do a dishwasher load and my wife dye her hair.

I haven't provided a water heater board to anyone in two years. But you may be in luck. I'm thinking of having a contest in about two months when I get some time. The winner will get a FREE water heater board set up for their system, whether stand alone or operating in parallel with a MPPT controller to harvest excess power. Selected individual will have to meet certain requirements.

 

[daklein]

A PV diverter does not have to be expensive or too awful complicated. I have a system that is usually operating off-grid, with AC-coupled solar, and controls to turn on dump loads or charge batteries when solar is available. I used a very cheap dimmer control, modulating the 240vAC to the water heater element. It has a noisy and poor power factor, but aside from me noticing the different humming noises, it works just fine. https://www.amazon.com/10000W-Voltage-Regulator-Controller-Thermostat/dp/B085W43M7B It claims to handle 10kw, eh ok, maybe, but my water heater element is only 3.8kw, so it does handle that fine.

I connected a wifi relay board with resistors, in place of the pot. This heats a hydronic water heating buffer tank, on the lower element, if there's excess solar. Box on the right is the SCR dimmer control for lower element, powered from off-grid house panel, wifi relay board varies the power in 16 steps depending on available excess solar. The small black and white wire from the relay board to the SCR is the replacement for the pot. The original pot I mounted on the side of the box, just unplugged inside. I did that for testing first, so I could control it either manually or later with the relay board. An arduino or ESP32 that would do a pwm output would be more elegant, but I already had these relay boards figured out to use with my controls. Other stuff in the picture is off-topic.... Upper box goes to a HPWH for domestic hot water. Timer box with aube relay is for control of the hydronic tank upper element as backup heat source from grid. An airsource heatpump is the main heat source for the hydronic.

 

[SeaGal]

I connected a wifi relay board with resistors, in place of the pot.

for simple but effective DIY solution.

 

[sparrowhawk]

When I was in the planning stages for a small back up solar system. I considered many schemes to heat water and all were to complicated & cumbersome. But I was limiting my options because the back up systems I was considering were small. So when I found out about diversion water heating I made the decision to think bigger.

Diversion water heating looked promising and had some advantages, but it also came with several disadvantages. After lots of research and kicking around the idea I decided it wasn't worth it for me to go that route.

In the end I decided the simplest thing was to go with the largest inverter, battery and panel set up I could afford and power my water heater from my system. My fiances limited my system to 6kw inverter, 25kwh of batteries and 6kw of panels which was not large enough to power my water heater along with all the other loads without some changes.

So one solution for my systems limitations was to replace the 4500 watt elements in the water heater to 2200 watt ones which put me in the doable range. The other solution was to put the water heater on a timer so that it only ran in middle of the day when solar power was most abundant. It's worked out great for me and I'm extremely pleased with everything.

I know my setup is not for everyone, but it's a lot simpler and has eliminated some additional potential failure points by not adding additional hardware to the mix. The way I see it if you're already using an inverter. Why would you want to add more complexity?

 

[damo green]

There are a variety of water heater controls like this in Europe that use voltage directly from array https://www.ebay.co.uk/itm/263190867548 where this is more of the norm.

Heating from batteries is a rather poor way to do it. Even from a 48V system the currents are enormous with serious losses. I heat all my water with PV using a similar diversion method that allows me to use a charge controller in parallel. Easy and cheap to build, but nobody here builds anything.

I saw some of your comments on their youtube video... Do you think their products is OK.... is the capacitor bank sufficient in size and quality? Perhaps I might change out the capacitors in terms of quality.... not sure if a capacitor capacity increase would worjk with their design..... if any change is even needed. I am also considering building one of "Loadmaster" 's boards. Very busy with the maior solar project so just thinking ahead about this for now.

 

[efficientPV]

With MPPT searching programs, the total capacitance may change tracking slightly. Capacitors have a current rating related to internal series resistance. Just as in wires, resistance causes heating and heat is the long term killer of capacitors. The welds from the foil to the exit conductor is also the weakest part of the construction and a failure point. You can get away with a lot of abuse for a while and some designs count on that. Capacitors are typically rated at 4,000 hours at temperature. As a water heater operating half the day could do that in 2 years. Lowering the temperature greatly extends the life. The techluck controller uses three capacitors in parallel rated at 0.68A each. People are reporting that these controllers become erratic after several years. I use five capacitors in parallel (6,000uF total) rated at 3.2A each. That results in a very low internal resistance and a long life. Other European designs are better and everything is made to a cost point. It is fine to just use them till they wear out. That can be hard to detect without observing input voltage waveform and saving it for a reference.

The loadmaster is a stand alone unit with IV tracking. I have focused on diversion of excess power and that requires a fixed voltage and temperature tracking. Everyone has excess power that they could use. My excess of the excess heats water for the laundry. Having laundry with hot water is worth it alone. My front loader a camp doesn't have the mold problems the one at home has At the time I looked their software didn't include a fixed voltage option. There are lots of designs for controllers. I avoided inductor designs because of the added complexity and difficulty in obtaining without real performance improvement. It also isn't a full MPPT with a resistor load. The load resistance should be half of ideal for full benefit in an inductor design. It is not really designed for twice the current.

 

[efficientPV]

Something simple can get you just as much performance. Buy spares and program them because when they fail, replacements may no longer be available. I've had a couple NANO taken out by lightning. Many FET from China are fake. I've had many of them. They should be tested for on resistance. I prefer a simple designs with common IC that will be available for the next 20 years. This is a $10 board from China that makes a water heater control. Too bad you never know what they will send you, but they all can be made to function. Some significant modification is required. This can be used with existing water heater thermostat in the high current element line without fear of arcing. Efficient water heating isn't hard, it just takes some talent.

 

[k490]

If you don't already have an electric water heater consider a Heat Pump Water Heater. They are more expensive but only use about 25% of the energy of a resistance element in traditional electric water heaters. The ones I know from Rheem and Ruud (same company) have timers so you would not even need a diverter. Since they only draw 500 Watts I have mine timed to run from 10AM to 4PM to avoid peak rates and charge from solar. If you don't want to do Net Metering you can buy a GT inverter which has an option for no export. You just have to put the Current transformer close to the meter so that the HPWH uses the power before it gets to the meter.

I have a 65-gallon Rheem heat pump water heater. When it's on energy save mode, it pulls 22 amps at 240V to heat it. If I change it to high demand same current pull 22 amps. About 5280 watts of power it's pulling with an outside temp is 55F where I save money is in the summer I don't need the element. My garage gets to about 102F I set it to heat pump only all that heat in the garage makes it heat water up very quickly. Not sure what it pulls in the summer I'll take a current reading next July see what it actually is pulling. In the wintertime I only run it during super off peak 10am-3pm. It holds enough heat all night to get one 10-minute shower although I removed the water restrictor in the shower head might go longer without that removed. If we need it I can always turn it on with my phone. I got it for free through a pilot program my utility was doing if I had to buy one I don't think I would. I think you get better ROI from just extra insulation and spending lot less on a dual-element heater or go with a solar hot water heater on the roof if I could get rebates from my utility to cover most of the cost. Econet software on my phone dose show KWH usage I read it was not accurate I believe that. It shows 2200kwh a day when the unit draws over 5000 watts.