Direct to water heating.

[houseofancients]

I forgot to talk about the Loadmaster. It is a design made by a software guy and some of his claims are a little iffy. I don't think it works as well as he says because he hasn't really tested it, just made guesses. Rather than losses in heat, it might not be able to transfer all the power the panels are capable of creating. Analysis of the AC component of the input would be needed. Solar panels are current sources. Panel current changes over a short period of time indicate the circuit is not able to store or use all the power a panel can create. The techluck claimed "97% (pass thru)" which is saying when it is not actually functioning at full power. Nothing was ever said about 50% panel power efficiency or any other range. It has three capacitors the size of your pinkie to store energy and provide all that current to the heating element. I just wonder how long those can last running all day for years being rated at .58A. Internal heating will dry them out. The internal welds of a capacitor are the weakest part. Often these are just a crimp. Repeated current spikes can cause them to fail. The standard rating for capacitors at temperature is 4,000 hours. Physical size and current rating of capacitors does matter. Sure would like to see a thermal scan of those at 50% power. There are differences in design that mean more to me than others. All these designs should work well enough to gain extra power over direct connect. The techluck which might never appear again had a severe weakness in the power supply making prone to extreme failure from a near lightning strike. I worked on
one and the only semiconductor on the board that didn't fail was the reverse voltage protect diode. For that owner it was a total loss. The ACTii solved this issue by making the power supply a seperate item. Really nice small packaging which makes it cheap to ship. Even though the photo boost module may have some length of life issues because of size, these would be easy to repair and make even more robust. Current is limited to 8A because it is made for a european marker where water heaters are smaller and lower wattage. I designed my board for 10A and that is only limited by the capacitor bank for convenient physical board size. The ACTii uses a fan because of its small packaging. Fans are always a failure point.
I use two FET in parallel which reduces heat by four times that of just one. In normal operation there is no noticible heat rise. The techluck also is 8A and uses a IGBT which has a high saturation voltage causing 15W of heat. Unknown if ACTii uses IGBT or FET.

Those with small camps don't realize that they can add hot water with a very minimal investment using only excess power from their existing array. A small array will possibly need only one or two extra panels. No need for for a battery or increase charge controller or inverter capacity. A common 6 gallon 120V water heater doesn't need the element changed and can heat water with only one KWH of energy a day. I use a 50L (13 gallon) tank from China that was only $150 shipped. It meets my needs and I have a dishwasher. Cheap, but needs extra insulation to make it effective. I gave up my propane heater years ago. I was refilling several tanks each summer. Waking up and having hot water at the sink really changes the camping experience and it costs next to nothing. This idea just has not gained acceptance as this group are usually beginners. You do need to use an array of 60-90V and they are usually battery voltage systems.

The ACTii is capable of operating in two modes. Full MPPT in a stand alone system or as a fixed voltage power diverter. The fixed mode has to be used when in parallel with a charge controller as two MPPT would fight each other. The voltage can be ajjusted seasonaly to account for temperature. My board has an external temp sensor that can be attached to a panel or a small square of metal exposed to approximately the same sun conditions. This calculates the natural power point.I was amazed at how many want to have stand alone hot water systems which to me seems very wasteful. Any successful PV system has to have excess power to recover from bad days. Why waste that power. These systems are amazing to see operate
and once you have used one would never be without one. The picture below is a half hour graph of my garage system. It uses only excess power not used by the house system and the house water heater. This 40 gallon water heater in the garage is only for the laundry and only hot water is feed into the cold inlet of the machine so all cycles use hot water.

This is a 60V array and at temperature the MPPT ideal voltage should be about 57V. This parallel array has a lot of shading and you'll see it drops to 44V at the very end, likely the refrigerator has turned on. I set the fixed voltage (green trace) higher to insure the house systems have priority. The garage also has two panels facing east and the distance to the house systems
is about 125 feet. The higher voltage insures more power goes to the house in the morning. I could be more agressive, but I get adequate hot water as it is. The dips are from clouds and/or demands on the house MPPT charge controller. Peaks above 61V are because the two 5500W elements in parallel can not take all the excess power available. Since it peaks out at less than 5A, the upper thermostat has opened up leaving just one element. Again I could be more aggressive and replace the heater elements with lower resistance ones. The clothes come out of rinse cycle steaming so why bother. I have a LG front loader and these are notorious for the mold problem. The detergent dispenser is always spotless. At home I have the same machine and have to pull the dispenser apart and clean mold out of it periodically. I should do full hot water at home. Blue trace is the element current. Yellow is the heater power. Notice given varying cloud conditions in this half hour view excess power is instantly delivered to the water heater or removed. It makes this decision up to 100 times a second. The washer in the garage also runs off these panels with no battery. When the washer goes into a low current fill cycle, excess power goes back to heating water. Free laundry is pretty neat although it has now become my responsibility to do it. In the event of a large cloud, the washer will stop. I'm generally out in the garage working so that isn't a problem. Just restart it when the cloud passes. That doesn't happen often. We line dry and only do laundry on really nice days. Last picture is the laundry system in the garage. You are not seeing things! Those are jumper cables connected to the water heater. The 10A pulses are ideal for desulfating a lead battery. Battery is in series with heater element. I just clip them together when I don't have a battery. I've been picking them up at the town recycling and experimenting with restoration. Our town has a lot of people who ignore batteries over winter and just buy new ones. Been lucky, can't save them all.

Tall tanks are really the way to go. The top 15 gallons heats quickly and the lower section acts as a pre heater. This stratified heating method has drawn the attention of adaptive control builders as a way to economize tank performance. Lower tank heat loss drops to almost nothing due to temperatures not much more than ambient. Larger 40 gallon tanks are almost the same cost as less popular small tanks.

thanks for you extended explatiin, however, this is not from the techluck guy.
he and nodge had similar designs, but both are not ( or no longer ) for sale, and i am not aware they ooensourced their designs

 

[houseofancients]

thanks for you extended explatiin, however, this is not from the techluck guy.
he and nodge had similar designs, but both are not ( or no longer ) for sale, and i am not aware they ooensourced their designs

actually seeing your pictures, you actually might be nrodge yourself

 

[copec]

Any old MSW inverter would be easy to hack using just the H bridge section. You feed array voltage into the HV caps and provide 12V to the control circuitry. Most old inverters use a TL494 to generate pulse widths. I've used anywhere from 20V to 140 to power them. Circuit just doesn't care. Add extra capacitors to the array to store power during off times to make it efficient. Just change the circuit for constant array voltage. As electronics goes it is quite simple. If this makes your head spin, there isn't really any help I can give you. There are hundreds of inverters and they are all slightly different. The output sections are quite robust. a 1,000W inverter could easily power 2,000W into a heating element with no heating issues.

Thanks! I really only need good starting information with what/where to look. I do have access to virtual unlimited APC UPS’s with dead batteries, I’ll have to look and see what I can use.

 

[Hedges]

Thanks! I really only need good starting information with what/where to look. I do have access to virtual unlimited APC UPS’s with dead batteries, I’ll have to look and see what I can use.

Are those MSW?
If you can vary the pulse width, do that to maintain PV around Vmp under varying illumination.

 

[copec]

Are those MSW?
If you can vary the pulse width, do that to maintain PV around Vmp under varying illumination.

I doubt they are MSW, because they are intended for electronics that are probably expecting a continuous waveform to convert from, but that sounds like a perfect idea to me.

 

[Hedges]

Or amplitude. That is probably easier, given that it is sampling some voltage-divided feedback point.
If you can varying it with a potentiometer, then could use a microcontroller with ADC and connected to a digital pot.

Or, control a dimmer circuit. Not sure how well inverters do feeding that.

 

[copec]

Or amplitude. That is probably easier, given that it is sampling some voltage-divided feedback point.
If you can varying it with a potentiometer, then could use a microcontroller with ADC and connected to a digital pot.

Or, control a dimmer circuit. Not sure how well inverters do feeding that.

I was just thinking altering the feedback would be the key to controlling it without making my own logic from scratch.

 

[Hedges]

It makes a sine wave. The higher the amplitude, the more power delivered to water heater (and taken from PV).
It can only maintain full voltage and deliver full power if PV can deliver that much. So I imagine implementing MPPT algorithm, or a simpler one to just maintain PV voltage at Vmp. It needs to vary power delivered to water heater to accomplish that.

 

[efficientPV]

I was just thinking altering the feedback would be the key to controlling it without making my own logic from scratch.

Starting from scratch and doing it the right way almost always tends to be the easiest and lowest cost. I do run my LG washer from a MSW inverter right off my 120V array with no batteries. Only needed to solder three wires.

 

[copec]

It makes a sine wave. The higher the amplitude, the more power delivered to water heater (and taken from PV).
It can only maintain full voltage and deliver full power if PV can deliver that much. So I imagine implementing MPPT algorithm, or a simpler one to just maintain PV voltage at Vmp. It needs to vary power delivered to water heater to accomplish that.

That’s exactly what I was talking about earlier that the cyboinverter does.

 

[efficientPV]

You don't know it makes a sine wave. creating a sine wave would be wasteful if it did.

 

[Hedges]

You don't know it makes a sine wave. creating a sine wave would be wasteful if it did.

Don't know about all their models, but first APC UPS data sheet I pulled up was sine wave.

These days, many computer loads have power-factor correction and may require power input which isn't too far from a sine wave.
Newer Dell computer, for instance, require PSW, while older models were OK on MSW.

PSW uses a switching power supply circuit to synthesize sine wave, and efficiency can be in the mid if not high 90%.

https://download.schneider-electric.com/files?p_enDocType=Brochure&p_File_Name=EcoStruxure+Ready+Smart-UPS+with+APC+SmartConnect+Brochure+Americas+V6.pdf

 

[copec]

I am interested in doing real DIY direct from PV resistive heating and want to lab experiment with it and then use it at scale - but does anyone in this thread just use the dry contacts on their inverter with a programmable battery voltage/SoC to control a DC relay from the battery (through a breaker) directly to a heating element (matched for duty cycle)?

I have a lot of excess but only sized my battery for if the grid is down at night, but run all the big loads during the day when the sun is out.

My plan for my house is to make a thermal storage tank to dump the excess into, for both preheating the hot water and to supplement heating in the winter. The house is adequately heated just though solar gains during the day, but drops during the night.

 

[efficientPV]

I was answering that the cyboverter did not use sine waves to heat water.

Energy should be diverted when it is made and not thru batteries. Modifying inverters feedback loop is not as easy as it seems. I am inherently lazy and have never bothered to do it. I've bought 495 and 3235 $5 inverter boards and they are a real pain with double sided traces. That 300W inverter board shown , nobody has been able to follow the instructions to modify it. It is just easier to take a 1 1/2 sq inch perf board and put a half dozen components on it. I've been working this problem for over 5 years and there is nothing out there which can easily be cobbled together from amazon. If you have a MSW inverter with a TL494 it is far easier to adjust the dead time control to change duty cycle using a TL431 as the voltage detect.

A TL431 and FET driver makes a fairly simple HW control. Just be sure the driver has a under voltage lockout. A NANO is the easiest driving a opto isolator to a FET. All these circuits have high current pulses and noise will be a problem with current loops an induced voltages. Efforts should be made to eliminate short pulses that result in FET heating. Here is my first water heater control that was uno based for my outside tank. It took 2 hours to build and the bubble wrap was protection from weather.

 

[copec]

I was answering that the cyboverter did not use sine waves to heat water.

The cyboinverter does use sine waves: http://www.cyboenergy.com/products/cimini1200h_spec.html and I've actually seen the output on an oscilloscope. It varies the voltage (amplitude) of the sine wave as (mppt) panel output changes.

Energy should be diverted when it is made and not thru batteries.

It's not through the battery per-say, It would only be on when the battery was at >80% SoC and the solar output was higher than what the heating element outputs. Then the *current loop* is basically an mppt connected to a heating element, and it would seamlessly work with what I have setup at my house right now.

Like I said though, I'm working on an off-grid homestead, and out there I'm really interested in the most direct panel to heating.

 

[Honuz]

The cyboinverter does use sine waves: http://www.cyboenergy.com/products/cimini1200h_spec.html and I've actually seen the output on an oscilloscope. It varies the voltage (amplitude) of the sine wave as (mppt) panel output changes.


It's not through the battery per-say, It would only be on when the battery was at >80% SoC and the solar output was higher than what the heating element outputs. Then the *current loop* is basically an mppt connected to a heating element, and it would seamlessly work with what I have setup at my house right now.

Like I said though, I'm working on an off-grid homestead, and out there I'm really interested in the most direct panel to heating.

Are you looking, like I do, for a solution that works in parallel to the inverter, meaning in paralleled with an existing mppt..?
I can't see any good solution that works that way, most of them work alone, meaning PVs are only there for this heating purpose (bad and non optimal way of using them) or for some of those solution with a divertion to an other device when the water is hot.. Giving priority to heating water, not acceptable in off grid system. Well...it could be if only a small quantity of water needs to be heat, but even then it would be far from optimal.

In a good of grid system, all available energy needs to be harvested.. Or at least till all "systems" are full (battery, hot water tank, pressure tank, slab temperature desired reached...etc). In my list, priority are those :

1 - Satisfy DC circuit (got one DC and one AC circuit)
1.1 - Lights DC 24V
1.2 - Water pump DC 24V (Water pressure)
1.3 - Charging computers, phones DC 24V

2.1 - All that is in use on the AC circuit, tower computer, tools, kitchen devices..etc
2.1 - Charge Batteries
2.2 - Heat Water tank, subsequently heat slab if needed.

It means, in my case, heating water is the last way of using energy that would be wasted otherwise. And batteries can't ever be used for this task. In this priority list, the only manual thing is to disconnect inverter at night if needed.
All items on this list can be satisfied at the same time if sun is shining, if not enough power available on PVs, 2.2 is cut.... Still not enough... Charging batteries is cut... Not enough.... Inverter is cut.
The only manual thing is to decrease or energy use in case of bad weather, no AC use most certainly means batteries will charge

 

[Hedges]

Giving priority to heating water, not acceptable in off grid system. Well...it could be if only a small quantity of water needs to be heat, but even then it would be far from optimal.

In a good of grid system, all available energy needs to be harvested.. Or at least till all "systems" are full (battery, hot water tank, pressure tank, slab temperature desired reached...etc). In my list, priority are those :

1 ...
2.2 - Heat Water tank, subsequently heat slab if needed.

For my AC coupled system, optimum implementation would apply variable load to AC in order to keep frequency in a certain range.
Nominal frequency is 60 Hz. Sunny Island increases frequency to request curtailed power production, and sometimes sits below 60 Hz to keep electromechanical clocks and timers on time (if that option is enabled.)

Default for Sunny Boy is deliver 100% of available power up to 61 Hz, linearly decrease to 0% at 62 Hz. UL-1741-SA frequency/watts is similar but frequency range is closer to 60 Hz.

Less important load like water heater could ramp up from 0% at 60.5 Hz to 100% at 61 Hz. That way, all power is available for battery charging and normal loads, but when not needed for those the battery inverter begins raising frequency and stabilizes wherever water heater consumes all surplus or its maximum wattage rating. If excess power is still available, frequency increases above 61 Hz and PV production is reduced.

This could be implemented electronically, ideally with a 1.0 power-factor buck converter taking in sine wave and delivering variable sine wave to water heater.

An electromechanical implementation would be to use a variac, with a motor driving its shaft servo-controlled to keep frequency at 60.5 Hz. That would vary sine wave amplitude feeding water heater.

A couple additional features could be disabling server motor when thermostat opens so variac setting remembers last position used, and using a load-shed feature below say 80% battery SoC to disable heater in case servo/variac gets stuck on.

 

[Honuz]

For my AC coupled system, optimum implementation would apply variable load to AC in order to keep frequency in a certain range.
Nominal frequency is 60 Hz. Sunny Island increases frequency to request curtailed power production, and sometimes sits below 60 Hz to keep electromechanical clocks and timers on time (if that option is enabled.)

Default for Sunny Boy is deliver 100% of available power up to 61 Hz, linearly decrease to 0% at 62 Hz. UL-1741-SA frequency/watts is similar but frequency range is closer to 60 Hz.

Less important load like water heater could ramp up from 0% at 60.5 Hz to 100% at 61 Hz. That way, all power is available for battery charging and normal loads, but when not needed for those the battery inverter begins raising frequency and stabilizes wherever water heater consumes all surplus or its maximum wattage rating. If excess power is still available, frequency increases above 61 Hz and PV production is reduced.

This could be implemented electronically, ideally with a 1.0 power-factor buck converter taking in sine wave and delivering variable sine wave to water heater.

An electromechanical implementation would be to use a variac, with a motor driving its shaft servo-controlled to keep frequency at 60.5 Hz. That would vary sine wave amplitude feeding water heater.

A couple additional features could be disabling server motor when thermostat opens so variac setting remembers last position used, and using a load-shed feature below say 80% battery SoC to disable heater in case servo/variac gets stuck on.

Hello Hedges,

Mmm how i see it, but perhaps i'm wrong, if the inverter AC side is starving, i think it will instantly take power from batteries therefore keeping frequency at 60-61Hz. My inverter which is a chinese all in one, do that ... still i've never checked the variation of frequencies on the AC side.

And.... the variac seems a pretty nice device, but i can't keep thinking that this thing is expensive ...

 

[efficientPV]

When I am off grid and divert from array voltage that is parallel to a charge controller. I will say that depending on solar 100% of the time for anything is costly. You can store in a tank or battery and that has its costs. Tall tanks which will stratify can give fast recovery and excess goes to lower tank to preheat water. I also did this with two smaller tanks in series. All I can say is that this PV diversion is tried and tested and while everyone is running around in circles. Batteries work, but there are several conversion losses in series and using a battery has real costs in terms of KWH life. Everything in a battery system has to be oversized to accommodate water heating. People do it because IT IS EASY and they don't know any other way.

Sine waves are quieter, but doing that makes those systems more expensive. High current pulses will always make a little noise.

I'm still looking for someone who is capable of installing one of my boards.

 

[Honuz]

When I am off grid and divert from array voltage that is parallel to a charge controller. I will say that depending on solar 100% of the time for anything is costly. You can store in a tank or battery and that has its costs. Tall tanks which will stratify can give fast recovery and excess goes to lower tank to preheat water. I also did this with two smaller tanks in series. All I can say is that this PV diversion is tried and tested and while everyone is running around in circles. Batteries work, but there are several conversion losses in series and using a battery has real costs in terms of KWH life. Everything in a battery system has to be oversized to accommodate water heating. People do it because IT IS EASY and they don't know any other way.

Sine waves are quieter, but doing that makes those systems more expensive. High current pulses will always make a little noise.

I got a 1000l tank, this make a virtual 60 kwh "heat battery" (20°C-90°C), not counting loses that are heating house anyway.
This 60kwh "battery" cost a little less then 1000€. I use this water to heat house (Low side exchanger) and to heat shower / kitchen water (top side exchanger).
So, for me ... i better put in this battery every watt that will anyway be "not used" (surplus).

In summer i stop heating earlier cause i do not need the bottom part of the tank , i perhaps use only top 25% of the tank.
PS : My tank is, as we call it here, a dead water tank, water inside can't be use, there is no water coming in or out of it... better stratification and no limestone on the electric resistor.