Solar PV is now the most cost efective energy source.

[cmhansen]

I might research if dirt can be used for adsorption as adsorbent (seems unlikely):

Adsorption is a surface-based process while absorption involves the whole volume of the material

but maybe for absorption... more to learn.  I read silica gel lasts 30 yrs for adsorption, it can run as low as 60C, seems like water loop cooling is standard but surely ambient air exchanger would work too. 

Minisplits are really efficient but I read some reports of mold with humidity, maybe due to no antimicrobe filter like you find in central air.

Hmmm makes sense high-efficiency voltage conversion needs a 'storage' device being cap, coil or both.  Or, what about using / switching battery cells directly for this...  the more cells the finer grain switching can be and can 'adjust' the combined cell voltage to the array, e.g. charging a 370V battery if 100 cells were switched.... it's interesting to think about what you have in that light as you could say your DC-DC 'matcher' router/hub/switch 'converter' just moves the 'cap / inductor storage' off board.  I wonder if you are switching both panels and loads to make the right voltage matches.  The whole system makes a DC-DC converter.  A stand-alone converter between panels and loads is just a bottleneck. 

Why not grid tie too?  Good if it's CE and UL, in fact UL is required even for off-grid in the US as far as I know, some municipalities are banning off grid even.  I have a $.50/watt rebate available where I'm at.... net metering is nice that net is averaged per-month so battery and thermal storage is pretty much unnecessary, normal rate is about $.13 and excess sells back for $.02.  I think it's about $25 in base fees so ~1.25MWh excess is needed to truly get it to 0, but that isn't that feasible to do on 1/4 acre lot really.  I was allotted about 12.45kW capacity calculated from previous year all-electric use, at winter angle this gives me 1.35MWh / month winter average and those months I use ~3.1 or so, so 3 or 4 $220 bills plus $300 base I estimate $1200 / yr or so utilities minus water which isn't bad, ~1500sqft 'leaky' older home.

[Kleinstein]

Filling the thermal buffer all the way to full, only if wether report is calling for a longer no sun phase is not very reliable. This way the effective buffer capacity is reduced. Even if the next few days have not much excess energy predicted one would have to keep the buffer full as there might be a cold phase after that. So not filling up the buffer increases the risk of running out of heat, or one would need a larger buffer or PV. It is only if the predicted weather allows to fill up the buffer all the way later, that one could safely not put the energy in the buffer, without increased risk of running out of heat. With something like a 5 days reasonable reliable weather forecast this won't happen very often in winter.

Without much control over heat release it already is a problem, that a very hot buffer will send out quite a lot of heat and will thus initially loose its temperature faster than normally needed. So the upper temperature part is not that effective for storage. This might not be a problem if there is another thermal buffer that has much better control for the heat release, so initially one can use heat from the concrete only and use the more controllable buffer only later.

To get the heat out of the concrete buffer one would need it to be considerably warmer than the room. So using the energy from cooling to 16 C would only work if the room temperature really falls down, maybe to 12 C - maybe Ok for a few rooms, but not very comfortable. Even when using the floor for preheating fresh air, one would still need a second heat source to compensate the other heat loss. So I think the buffer capacity of the concrete floor is overestimated, especially at the lower temperature end. Without a heat pump (could be used for cooling too), temperature lower than room temperature (plus a little extra) has a low value. Possible use like air preheating or maybe water preheating would need extra investment.

I would more expect the concrete buffer part to be between 20 C and maybe 35 C (if the structure allows). The higher temperature part would have a small draw back: To compensate for to much heat released to the building one would need to cool be something like extra ventilation. So there would be extra heat loss if the buffer is near full, but still some extra heat could be stored for a short time, with little extra costs (ventilation for cooling).
It could be a good idea to have something like a thick carpet to slow down the heat release in a way that could be removed in a worst case period. By design the structure is only a short time storage, especially if not isolated against the room. Ideally coupling is variable - but this costs extra.

Anyway the concrete buffer should only be a part of the thermal buffer - maybe half, so you have at least some heat available you can control and keep for longer than maybe a 3-4 days.

[cmhansen]

To get the heat out of the concrete buffer one would need it to be considerably warmer than the room. So using the energy from cooling to 16 C would only work if the room temperature really falls down, maybe to 12 C - maybe Ok for a few rooms, but not very comfortable. Even when using the floor for preheating fresh air, one would still need a second heat source to compensate the other heat loss. So I think the buffer capacity of the concrete floor is overestimated, especially at the lower temperature end. Without a heat pump (could be used for cooling too), temperature lower than room temperature (plus a little extra) has a low value. Possible use like air preheating or maybe water preheating would need extra investment.

PV for heating is ideal ok.  Evacuated tubes fall linearly on efficiency if you look at the data (and heat flux formula):
http://www.heliodyne.com/wp-content/uploads/2016/03/Evacuated-Tube-Comp.pdf

https://en.wikipedia.org/wiki/Square-cube_law#Engineering_examples

This is the reason why underground molten salt storage can reach claimed 99% annual thermal efficiency once a large enough size is reached.  Volume is indeed important and it's actually advantageous to increase the heat buffer size in addition to using the highest specific heat possible (basically water), maximizing both minimizes both the temperature delta needed and also the percentage lost at the surface, of course insulation is still warranted. 

For a ~130m^2 home in cold climate I don't think even 20kW PV is enough heat buffer, one really should consider thermal collectors even assuming $.50/watt panels, also should be possible to use about half the area if it's 34% vs 17% efficient.  Those solar pool collectors / heaters show benefit:

http://www.houzz.com/photos/37108213/SunHeater-Universal-2-x-20-Solar-Heating-Panel-80-Sq-Ft-Set-of-2-traditional-heating-and-cooling

You get about 14.5m^2 for $194 shipped, at 20% efficiency that's almost 3kW!! or over 7x less installed cost assuming a 388W solar capacity for 194 ($.50/watt) shipped.  If they last 10* years, that's still over 2x less amortized cost.  True you'll need glycol and a pump so that adds to BOS.  If you have 30kW thermal, you've saved ~$13k up front and ~$7.5k over 30yrs.

Chances are that one won't last 10yrs. One could for even less make monolithic water panels from xypex waterproof concrete with lots of fiberglass to stop cracks.  Small enough concrete panels shouldn't crack and can be repaired. 

Water allows easy heat transfer in and out when needed.  If PV really is afforded for such a large cold climate heat buffer, place heating elements inside the water tank.  That should be more economical than a heated slab, you'll still have the inside watersource handler heatpump but you need, at least want, an airhandler anyways.  Circulate water to an insulated hole lined with a pool liner to make a tank (or use a tank).  Basically no loss at night.  Run the inside heatpump as needed. 

You couldn't tell by a single data point (i.e. device of only one size tested) or even the heat flux formula, but due to the 2/3 power law that applies to well everything physical, scaling anything becomes slightly non-linear.  This can add up, in a good way hopefully.

[electrodacus]

Filling the thermal buffer all the way to full, only if wether report is calling for a longer no sun phase is not very reliable. This way the effective buffer capacity is reduced. Even if the next few days have not much excess energy predicted one would have to keep the buffer full as there might be a cold phase after that. So not filling up the buffer increases the risk of running out of heat, or one would need a larger buffer or PV. It is only if the predicted weather allows to fill up the buffer all the way later, that one could safely not put the energy in the buffer, without increased risk of running out of heat. With something like a 5 days reasonable reliable weather forecast this won't happen very often in winter.

As mentioned before having the buffer at around 75 to 80% most of the time is OK and if weather forecast (that unfortunately is not always reliable) predicts multiple consecutive bad days then buffer can be charged to 90 even 100% even if less comfortable for that one or two days.

Without much control over heat release it already is a problem, that a very hot buffer will send out quite a lot of heat and will thus initially loose its temperature faster than normally needed. So the upper temperature part is not that effective for storage. This might not be a problem if there is another thermal buffer that has much better control for the heat release, so initially one can use heat from the concrete only and use the more controllable buffer only later.

To get the heat out of the concrete buffer one would need it to be considerably warmer than the room. So using the energy from cooling to 16 C would only work if the room temperature really falls down, maybe to 12 C - maybe Ok for a few rooms, but not very comfortable. Even when using the floor for preheating fresh air, one would still need a second heat source to compensate the other heat loss. So I think the buffer capacity of the concrete floor is overestimated, especially at the lower temperature end. Without a heat pump (could be used for cooling too), temperature lower than room temperature (plus a little extra) has a low value. Possible use like air preheating or maybe water preheating would need extra investment.

I understand now what you try to say. My house ambient air temperature is always the same as concrete floor temperature with no more than 0.5C difference between ambient air and concrete floor.
The amount of loss is extremely predictable and only based on delta between inside temperature and outside.
So if the concrete floor gets to 16C then ambient temperature is also at least 15.5C
So when I was mentioning absolute max range of 12C delta in the example I considered the +16C to +28C floor temperature range that will equate to a 15.5 to 27.5C ambient temperature unless you affect that by opening a window or something like that and after you close the window the ambient will stabilize again at around the floor temperature in no more than 1 to 2h
The wall that are 10cm solid wood and only thermally insulated on the outside additional mass that I did not added in the example to keep things simple but is not insignificant.
The wall is made out of about 8.5 cubic meter of wood that has a thermal heat capacity of about  0.35kWh per cubic meter per degree Celsius so almost additional 3kWh per degree Celsius additional to the concrete floor that has 8kWh per degree Celsius.
So while 12C delta for concrete floor may be a bit highs for comfort the storage capacity of the house is still probably at around 100kWh when also wood in the walls and Drywall are included at probably just 8C delta.
The example was just that showing how to calculate the thermal mass storage and the fact that a large floor can have significant amount of thermal storage at no additional cost since is part of the structure anyway and has the role of floor and foundation.   

I would more expect the concrete buffer part to be between 20 C and maybe 35 C (if the structure allows). The higher temperature part would have a small draw back: To compensate for to much heat released to the building one would need to cool be something like extra ventilation. So there would be extra heat loss if the buffer is near full, but still some extra heat could be stored for a short time, with little extra costs (ventilation for cooling).
It could be a good idea to have something like a thick carpet to slow down the heat release in a way that could be removed in a worst case period. By design the structure is only a short time storage, especially if not isolated against the room. Ideally coupling is variable - but this costs extra.

Anyway the concrete buffer should only be a part of the thermal buffer - maybe half, so you have at least some heat available you can control and keep for longer than maybe a 3-4 days.

As mentioned above there is no significant difference between concrete floor and actual indoor air temperature.
I just dislike carpets that is why plan was always for ceramic tiles and in floor heating.
I will say Concrete floor can easily be all the thermal storage needed even if you need to add more concrete than minimum required for structural integrity.
Yes a smaller 15 to 20% of the main thermal storage can make the ambient more constant for better comfort.

This is the reason why underground molten salt storage can reach claimed 99% annual thermal efficiency once a large enough size is reached.  Volume is indeed important and it's actually advantageous to increase the heat buffer size in addition to using the highest specific heat possible (basically water), maximizing both minimizes both the temperature delta needed and also the percentage lost at the surface, of course insulation is still warranted. 

For a ~130m^2 home in cold climate I don't think even 20kW PV is enough heat buffer, one really does need thermal collectors even assuming $.50/watt panels, also should be possible to use about half the area if it's 34% vs 17% efficient.

Water allows easy heat transfer in and out when needed.  If PV really is afforded for such a large cold climate heat buffer, place heating elements inside the water tank.  That should be more economical than a heated slab, you'll still have the inside watersource handler heatpump but you need, at least want, an airhandler anyways.  Circulate water to an insulated hole lined with a pool liner to make a tank (or use a tank).  Basically no loss at night.  Run the inside heatpump as needed. 

You couldn't tell by a single data point (i.e. device of only one size tested) or even the heat flux formula, but due to the 2/3 power law that applies to well everything physical, scaling anything becomes slightly non-linear.  This can add up, in a good way hopefully.

I do not doubt that you can create a large seasonal thermal storage with realy good efficiency outside the house but seasonal thermal storage makes no economical sense with inexpensive PV solar.

Maybe I already mentioned this in this forum and are sure present in my paper but here are the economical reasons.

This are just rough numbers based on 25 year amortization period.

PV panels  2.4 cent/kWh  (based on 80 cent/Watt acquisition cost, 25 years amortization and amount of solar energy at my location) it assumes you will be able to use all available energy and in my case I will be using only around 40 to 45% in average over a year.

LiFePO4 battery is a bit harder to calculate since it depends on model and how well it is used but a realistic number will be around 25cetn/kWh (a bit lower with DMPPT)

And last is thermal storage short therm few days estimated at 0.5cent/kWh

Best from an economic perspective is to maximize the thermal storage and PV array and reduce as much as possible the LiFePO4 storage.

Considering my example total cost for heating around $12k big part of that is the PV array at about 8k then much less thermal storage 2k and wires for heating and power transmission + DMPPT and some other accessories make up the rest.

Now thermal storage is say good for around 3 to 5 days and if you want seasonal storage then that will need to be able to store energy for 60 to 90 days and that means a thermal storage that is in a simplified way 15 to 30x more expensive and that means a cost of 30k to 50k for a seasonal thermal storage with savings of just 4k by reducing the PV array needed to half.
Of course that will not make sense for me in my climate and probably will also make no sense for anyone in any climate.
Even if you have horrible solar compared to my location adding extra PV capacity will still probably be a better option.

My house is 65m^2 and in a realy cold climate so a house that is build at same standards as mine ad same location that is 130m^2 will need probably just 16kW of PV since the larger the house the more efficient it is assuming the same level of thermal insulation.
As you seen in my comparison thermal collectors are actually more expensive than PV panels (just the panels) and when you consider the entire system that gap gets even larger so the only advantage is reduced collector area needed to half but at significant extra cost and complexity also reduced reliability to name just a few.

[Kleinstein]

With the thermal buffer so close coupled to the room temperature, it could rather fast get uncomfortable warm or cold. In this case even an 8 K or even 6 K temperature window is rather high and thus the storage capacity is getting rather small.
Adding more PV has only limited capability to compensate for too little storage. The worst case week can be rather bad, even if on average the conditions are good.

The solar condition for the OP are extremely good, so this only applies to the few good places with plenty of sun in winter. Here in Germany we are at something like half over the year and maybe 1/5 in winter.

[electrodacus]

With the thermal buffer so close coupled to the room temperature, it could rather fast get uncomfortable warm or cold. In this case even an 8 K or even 6 K temperature window is rather high and thus the storage capacity is getting rather small.
Adding more PV has only limited capability to compensate for too little storage. The worst case week can be rather bad, even if on average the conditions are good.

The solar condition for the OP are extremely good, so this only applies to the few good places with plenty of sun in winter. Here in Germany we are at something like half over the year and maybe 1/5 in winter.

I absolutely age with you that thermal storage and PV array size will need to be dimensioned for each particular case. I used my house and my location as an example and showed how thermal mass needs to be calculated.
Germany has fairly bad solar resource that is probably why people there where considering even seasonal thermal storage.
Still if space is not an issue then over-sizing the PV array is up to a point better than over-sizing the thermal storage and there is an optimum between the size of PV and thermal storage but that is realy specific to both location and house.

Yes a floor used as thermal storage is closely coupled to room temperature as long as there is no secondary system to control the ambient air. Since my concrete floor storage is marginal the plan is to add an additional thermal storage connected to heat recovery ventilation to have a better control of the air temperature thus maintaining a more constant room temperature.

If I stop all heating in the house the temperature drop in 24h will be between 1C and 2C depending on external temperature. The last 4 days where extremely cloudy here in January and I did no heating for the past 4 days.
The starting temperature was just +21C and I ended up now with just +16.5C after full 4 days without heating. This is an unusual year compared to average you seen from PVWatts and the outside temperature is also way above average for this period with an average of just around -5C over this last 4 days (actually it was warm even before that and will stay above until the end of the month at least). The average temperature for this month should be -16.5C and will probably not be the case this year. The beginning of the month was realy clod with maybe average around -20C but will still not compensate for this warm weather.
With current heating solution is difficult to heat the floor above 20 to 21C because the propane heater has a temperature limit and I use some small DC pumps and a heat exchanger plus a smaller 200 liter water buffer before sending water trough a PEX tube embedded in the concrete floor.
As it was a temporary solution I did not try to improve on the current system. I'm quite impressed by the fact that this cheap propane heater was able to last as long with this being the 4 winter and it seems it will probably make it.
The burner looked quite bad when I checked last year so a fail can happen at any time and then I will need a quick replacement.
You can see here the most recent photos of the temporary heating system https://plus.google.com/+DacianTodea/posts/VXbijJHXYmT  I know it looks bad but when I installed I was thinking it will be for one or two winters and now is almost 4 
This is an even older post https://plus.google.com/+DacianTodea/posts/SDVyfKjAGev and here you can actually see the burner https://plus.google.com/+DacianTodea/posts/8EKdxc3sD7j that was when it was newly installed so it looked good.
This is an inexpensive propane water heater for camping showers or stuff like that was not intended to heat a house

On one hand Germany has way less solar in winter than I have here but on the other hand the climate is much more moderate so a house of same size and level of thermal insulation will need much less energy to heat.
So for example Berlin in January will get just 228kWh with the same 10kW PV array where I get 1165kWh in same month
On the other hand Berlin average temperature in January is 0C where at my location is -16.5C thus almost twice as much energy is needed to keep the same house heated at 20C at my location than Berlin.
So my house if moved to Berlin will need a 25kW PV array and also probably about 2x the thermal storage capacity because of more consecutive cloudy days possible there.

Still if there is enough space for the larger PV array then it will still probably be the most cost effective heating option since I'm guessing here that natural gas will be more expensive than at my location.
If you use natural gas let me know what is the cost at your location for one unit of energy. If natural gas is 2 to 2.5x more expensive than here then solar PV will still be even there the most cost effective solution.

[Kleinstein]

The propane heater does not even look that bad. It might still work as a backup. Still it looks like there could be quite some heat loss to the outside. So the efficiency may not be that good and thus the need for heat is overestimated.

With an extra thermal buffer, the floor can be used as a limited thermal buffer. But there really should be a second one, where you have more control over the heat release. I would be surprised if there are not occasionally bad weather phases of something like a week or so. It is a relatively normal correlation that cloudy days will not be very cold and very cold days are often sunny. This helps a little, as one needs the buffer only for the not so could days. Still it looks like the floor is good for maybe 4 days and this with rather low temperature at the end.

Here in Germany it is not unusual to not see the sun for a month or even longer in winter. So it takes more than twice the buffer. There it also quite some variation from winter to winter. Average January temperature may be between about + 8 and -12 C. There is quite a lot of variation, because it depends on the weather pattern: sometimes the cold only comes til Poland and sometimes there is even snow in London. It depends on how large to cold high pressure region over Russia gets - so some years it's really cold and other years not. This makes even seasonal storage difficult, as one might need twice the normal heat in bad winters. So one has to be careful with just looking at averages.

[electrodacus]

The propane heater does not even look that bad. It might still work as a backup. Still it looks like there could be quite some heat loss to the outside. So the efficiency may not be that good and thus the need for heat is overestimated.

With an extra thermal buffer, the floor can be used as a limited thermal buffer. But there really should be a second one, where you have more control over the heat release. I would be surprised if there are not occasionally bad weather phases of something like a week or so. It is a relatively normal correlation that cloudy days will not be very cold and very cold days are often sunny. This helps a little, as one needs the buffer only for the not so could days. Still it looks like the floor is good for maybe 4 days and this with rather low temperature at the end.

Here in Germany it is not unusual to not see the sun for a month or even longer in winter. So it takes more than twice the buffer. There it also quite some variation from winter to winter. Average January temperature may be between about + 8 and -12 C. There is quite a lot of variation, because it depends on the weather pattern: sometimes the cold only comes til Poland and sometimes there is even snow in London. It depends on how large to cold high pressure region over Russia gets - so some years it's really cold and other years not. This makes even seasonal storage difficult, as one might need twice the normal heat in bad winters. So one has to be careful with just looking at averages.

The temperature of the flame is so high compared to glycol mix that absorbs that and with ambient that losses are not that high and not that dependent on ambient temperature. I also made rough measurements and base on that efficiency is not lower than 75 to 80%.
The burn rate is fixed and what I can do is just heat the water in the 206 liter barrel and then measure the water temperature increase over a specified period of time and calculate how much energy I get inside. Then I also can calculate how much time each propane take will last and the amount of energy it contains is known.
The thing is that the propane tank containing about 100kWh costs 25 to 30 CAD so that is about 10x more than the PV panels cost amortization if you are able to use all solar energy available.
Having an oversized PV array is still better than even covering just 10% of the energy with propane.
Yes an average year is just that and there are variations and I can see that this year compared to an average year.
Fore example than few consecutive cloudy days with very low solar output a few days ago look more like something you see here in November but the average temperature for those few days where also same as November and not January.
Right now as I write this there are +2C here and last night was a minimum of +1C and that is way above average in January here (-16.5C) and this is not a short period it started in the middle of January and will still last a few days so the second half of January this year looked more like a typical November in both temperature and solar output.

Yes I'm quite fortunate with the amount of sun here especially compared with Germany and other countries in Europe but solar will still be a good solution there since temperature is much milder and fossil fuel options cost 2x more there than here.
Average on each year for a month has little fluctuation no more than 10 to 15% more or less energy for a certain month in a different year.
For example at my location for January I had
2017  -14.3C (so far just 27 days)
2016  -11.2C
2015  -11.1C
2014  -15.2C
2013  -14.7C
2012  -10.0C
2011  -15.1C
2010  -13.3C
2009  -17.6C
2008  -15.0C
2007  -13.2C
2006   -6.4C
2005   -18.8C
2004   -18.0C
2003   -15.4C
2002  -13.0C
2001  -10.4C
2000  -15.6C

So I will say that using that average of -16.5C is good enough for design as the coldest year average was -18.8C in 2005 and there was that 2006 extra warm January with just -6.4C but not important.
So in your case I will design for a -8C average and all should be fine even in the coldest years that are probably also with more sun.
While a -8C average January seems like much different from a -12C when you look at energy needs the difference is much smaller since is based on delta between inside say +20C and outside so you have a delta of 28C to a delta of 32C so just about 14% more energy needed in the colder year than average.
The thing with solar PV is that it will produce energy even in those realy bad overcast day. And I noticed that having snow on the ground helps quite a bit when the panels are at a high tilt angle since the reflected light from snow can double the PV solar output . The fact that I had less than average solar those overcast days recently had also to do with the fact that all snow melted and everything was gray and dark no reflective surfaces around.

[CCitizenTO]

As for residential solar for heating, why not just heat your storage medium / working fluid directly? Why go through the electrical step? It will make substantially more efficient use of the sunlight. The technology for this has existed for a very long time.

I'm guessing he needs some electricity generation to run some appliances other than the furnace/heater/whatever (Fridge/Freezer/TV/etc...). That said I have seen solar powered hot water systems... It'd just be a matter of switching that into a radiator type system to heat the house. Would probably greatly improve the efficiency of the system too if the bulk of the energy is being used for heating as opposed to you know other stuff as well. With such a setup they may be able to get away with a smaller solar installation.

[Kleinstein]

Solar thermal collectors are possible too. They are mainly known for hot water, but can also be used to support room heating.

With the price for PV going down so far, PV plus heaters could be come an alternative.

The solarthermal collectors can be a little cheaper, but they require relatively expensive pipe-work and more maintenance for the liquid and pumps. Depending on the needed temperature the efficiency can be higher (maybe twice), but lower intensity light can not be used well for higher temperatures (e.g. 60 C). A large distance from the collector to the buffer also can be a problem.

As far as my estimate goes, it depends on the local conditions wether PV + heater or direct solar thermal is more cost effective. The lower costs for installation and maintenance compensates for more expensive collectors. It also depends on what is done DIY - some can do electrics, some can do pipes.

[electrodacus]

I'm guessing he needs some electricity generation to run some appliances other than the furnace/heater/whatever (Fridge/Freezer/TV/etc...). That said I have seen solar powered hot water systems... It'd just be a matter of switching that into a radiator type system to heat the house. Would probably greatly improve the efficiency of the system too if the bulk of the energy is being used for heating as opposed to you know other stuff as well. With such a setup they may be able to get away with a smaller solar installation.

Yes I need electricity also and the combination of the two makes the electricity part lower cost (that is an additional advantage).
But even if you only need heating solar thermal is more expensive and to get the exact numbers check the page 5  http://electrodacus.com/DMPPT450/dmppt-presentation-v01.pdf
The title is "Solar PV is now the most cost effective energy source" this means is also cheaper than thermal solar.
If thermal solar will have been a better option I will have selected that since cost was may main selection criteria for the house.

[Seekonk]

It is shocking when you see systems like http://www.sunbandit.us/ selling almost a conventional water heater in excess of $9,000 for the system. Just how does that pay off?  It is a conventional water heater with one extra heater for grid connect (3 total) a couple of panels and enphase converters.  I heat water with PV and the home made PWM converter to keep the panels at power point only costs about $25. Failing to keep the panels at power point will result in 50% power loss.  Your basic cost is almost just the panels.  If you design the system as supplemental heating and use only the lower heating element for PV, nearly 100% of the panels potential power generation is used 365 days a year. That surpasses almost anything else you can do with PV at a much lower hardware cost.  $500 for just two panels is a cheap way to get into solar.  Shamefully simple.  Just a nano,  a couple FET and some capacitors as the storage bank.

[electrodacus]

It is shocking when you see systems like http://www.sunbandit.us/ selling almost a conventional water heater in excess of $9,000 for the system. Just how does that pay off?  It is a conventional water heater with one extra heater for grid connect (3 total) a couple of panels and enphase converters.  I heat water with PV and the home made PWM converter to keep the panels at power point only costs about $25. Failing to keep the panels at power point will result in 50% power loss.  Your basic cost is almost just the panels.  If you design the system as supplemental heating and use only the lower heating element for PV, nearly 100% of the panels potential power generation is used 365 days a year. That surpasses almost anything else you can do with PV at a much lower hardware cost.  $500 for just two panels is a cheap way to get into solar.  Shamefully simple.  Just a nano,  a couple FET and some capacitors as the storage bank.

The name is quite clear "SunBandit"
Where do you see the price on that website ? I was not able to find a price. And what version of the Kit is that $9000 for ?
In any case they sell you a few components in the Kit.
I will suspect the most expensive part is the water heater tank.
Since they use microgrid inverters I do not even see the benefit since you need grid and you can just connect that directly to your house so any device can use that energy including the water heater.
I seen something in the past from Germany I think where they had an inverter but offgrid (no need for grid) but was also expensive.

The PWM thing works not sure what version is $25 but it can be relatively inexpensive for low power 500 to 1000W of PV panels that is more than sufficient for a normal hot water tank.
I seen one version of this PWM water heater on eBay for 250 USD and supports I think up to 3 to 4 panels in series about 750W to 1000W. Is that what are you referring and $25 was a typo ?

[Seekonk]

I would like to think this is a technical board and someone has the electronic nouse to build something simple.  Though, reading many posts I question that premise. "Just a nano,  a couple FET and some capacitors as the storage bank." refers to what you would have to buy for $25.  That $275 one has capacitors that can't handle the current and won't last long.  It also uses a IGBT, not a big issue if you want to fool around with getting rid of 10W of heat.   Everyone would like to have a tank like this http://www.akvaterm.fi/eng/Accumulators/Akva.38.html   Tank losses tend to be around 100-150W each hour.  Backup heating could be supplied by a low voltage transformer of only 250W.  A system is workable if you can get a 36V (50+V power point) string or higher and use a standard $10  2,000W 120V heater element.  Most electronic wall warts will operate at 50V giving an easy supply for micro.

On another board someone was quoted a complete install for $9,000 of the sunbandit.  This was a discounted price of normally $11,000 and they encouraged him to act quickly.

[electrodacus]

I would like to think this is a technical board and someone has the electronic nouse to build something simple.  Though, reading many posts I question that premise. "Just a nano,  a couple FET and some capacitors as the storage bank." refers to what you would have to buy for $25.  That $275 one has capacitors that can't handle the current and won't last long.  It also uses a IGBT, not a big issue if you want to fool around with getting rid of 10W of heat.   Everyone would like to have a tank like this http://www.akvaterm.fi/eng/Accumulators/Akva.38.html   Tank losses tend to be around 100-150W each hour.  Backup heating could be supplied by a low voltage transformer of only 250W.  A system is workable if you can get a 36V (50+V power point) string or higher and use a standard $10  2,000W 120V heater element.  Most electronic wall warts will operate at 50V giving an easy supply for micro.

On another board someone was quoted a complete install for $9,000 of the sunbandit.  This was a discounted price of normally $11,000 and they encouraged him to act quickly.

Sorry I somehow missed the fact that you mentioned a nano (as in AVR based mini arduino) and implied people will build their own version with parts worth about $25.
Is quite likely that those capacitors will not last much in that $275 PWM device.
I do not think that $25 is a realistic number if you want a quality electrolytic and mosfets and you will most likely need some other parts.
If you have a tank with 150W losses then you need 3.6kWh in 24h just to maintain the temperature and that is in average what 3 x 250W panels will be able to produce so they will be barely able to compensate for the losses.
 
A standard 2000W 120V heat element will have a resistance of around 7.2Ohm.
So with say two 60 cell PV panels (250W x 2) in series that have a max power point of around 60V you will get 60V/7.2Ohm = 8.33A so when sunny at noon you will have the FET fully on maybe and the rest of the time you will need to PWM to get closer to max power point but of course there will be losses on both FET and capacitor in this conditions (fairly significant).
With this fixed 2000W 120V fixed element nothing other than 2x250W will be able to work at max power point so you can not use 3 panels for example.
And with 500W you will get in an average sunny day around 2.4kWh/day so for a large water tank this will maybe just cover the losses of the tank.
This will quickly become expensive even if you can buy parts for $25 and you will assemble for free (for each 500W of PV) if you need 10kW of PV as I will need or 14kW the DMPPT450 supports to use for space heating.

[Seekonk]

Dacian, I think I deserve to use as crappy a numbers as you do.  I can buy 3 Canadian Solar 270W locally for $160 each. I said $500 in panels. That gives me 90V which is a good match.  I also said that this is supplemental  so that 100% of the panels output is always used, not heating 100% of all water with solar.  By doing this you get a cost performance better than with a heat pump.  Adding a heat pump to this would be ideal energy savings but it would never pay off the heat pump with reduced use.  I do heat water at my camp with just 900W of panels.  I can divert 2500WH a day since I have to run everything else at the camp, like refrigeration, with this same power. In that case I have two tanks and can switch between them and use both.  Take a course in reading comprehension. My numbers are real and a lot better than some of the ones you use.  In a world of jellybean designs where you can't tell a Hyundai from a Mercedes, I have an appreciation for something a little different.  I just wish you had someone to bounce some of your ideas off before you go into production.  They really aren't mature designs.  That said you ought to build one of these so you have something that is marketable.  Best of luck.

[electrodacus]

Dacian, I think I deserve to use as crappy a numbers as you do.  I can buy 3 Canadian Solar 270W locally for $160 each. I said $500 in panels. That gives me 90V which is a good match.  I also said that this is supplemental  so that 100% of the panels output is always used, not heating 100% of all water with solar.  By doing this you get a cost performance better than with a heat pump.  Adding a heat pump to this would be ideal energy savings but it would never pay off the heat pump with reduced use.  I do heat water at my camp with just 900W of panels.  I can divert 2500WH a day since I have to run everything else at the camp, like refrigeration, with this same power. In that case I have two tanks and can switch between them and use both.  Take a course in reading comprehension. My numbers are real and a lot better than some of the ones you use.  In a world of jellybean designs where you can't tell a Hyundai from a Mercedes, I have an appreciation for something a little different.  I just wish you had someone to bounce some of your ideas off before you go into production.  They really aren't mature designs.  That said you ought to build one of these so you have something that is marketable.  Best of luck.

Panels can be used with PWM for your system or with mine so panels will cost the same.  DMPPT450 is designed for a max 14kW PV array so not for small system to supplement domestic hot water.
If you where to need 14kW of PV panels using PWM method will be more costly and way more nasty (I'm referring to all that electromagnetic noise generated by switching 14KW)
If the DMPPT450 will have been just for heating the cost will have been 40% less than it is now so even less expensive but a large part of the DMPPT450 is related to battery charging also.

So say you just need 25$ for each 800W of panels then you need about 17 of this $25 devices to get to same 14kW capability of DMPPT450 and cost will them be $425 same as DMPPT but only have one function and done worse plus you need to build all those yourself is not already build.  Since DMPPT450 has no need for electrolytic capacitors it will most likely outlast those DIY $25 worth of parts devices be much smaller and much more efficient (way lower TDP).

On top of that DMPPT450 when used for both heating and Lithium battery charging together with SBMS will help reduce the battery capacity needed to at least half making savings of a few thousand $.

If you do not mind can you provide info about your setup. Like what panels are you using in that 900W array what capacitor and mosfet (TVS or other important parts). I can take a look at those and make a more clear analysis.

Also curios how you use the Battery (what type) and heating on the same 900W PV array?

[NiHaoMike]

LiFePO4 battery is a bit harder to calculate since it depends on model and how well it is used but a realistic number will be around 25cetn/kWh (a bit lower with DMPPT)

And last is thermal storage short therm few days estimated at 0.5cent/kWh

That really drives home why thermal storage makes the most sense for the three biggest loads in most homes - HVAC, refrigeration, and hot water. The remainder - mostly lighting and other electronics - is easily handled by a few kWh of batteries. Maybe only a few hundred Wh for a minimalist. Too bad lithium batteries and LED bulbs were really expensive back when I went to college, since the $40/month utility service charge in that area over the course of two semesters would easily buy nearly 2kWh of lithium batteries nowadays. (Multiply that by up to 4 years of college!) Take them on campus to charge, take them back home and have a pseudo off grid setup. (HVAC need is minimal to nonexistent over much of the school year, refrigeration could be done without as the grocery store was well within walking distance, and hot water is provided by the campus CHP setup.)

What I would really like to see more of are stationary bicycle generators. While limited to 100W or so, that's actually a lot for modern electronics. Make a portable version for charging portable electronics and we'll have a way to turn the obesity epidemic into a renewable energy source 50 times as energy dense as lithium batteries.

[electrodacus]

LiFePO4 battery is a bit harder to calculate since it depends on model and how well it is used but a realistic number will be around 25cetn/kWh (a bit lower with DMPPT)

And last is thermal storage short therm few days estimated at 0.5cent/kWh

That really drives home why thermal storage makes the most sense for the three biggest loads in most homes - HVAC, refrigeration, and hot water. The remainder - mostly lighting and other electronics - is easily handled by a few kWh of batteries. Maybe only a few hundred Wh for a minimalist. Too bad lithium batteries and LED bulbs were really expensive back when I went to college, since the $40/month utility service charge in that area over the course of two semesters would easily buy nearly 2kWh of lithium batteries nowadays. (Multiply that by up to 4 years of college!) Take them on campus to charge, take them back home and have a pseudo off grid setup. (HVAC need is minimal to nonexistent over much of the school year, refrigeration could be done without as the grocery store was well within walking distance, and hot water is provided by the campus CHP setup.)

What I would really like to see more of are stationary bicycle generators. While limited to 100W or so, that's actually a lot for modern electronics. Make a portable version for charging portable electronics and we'll have a way to turn the obesity epidemic into a renewable energy source 50 times as energy dense as lithium batteries.

Out of those mentioned hot water probably needs the most energy then HVAC and refrigeration not as much and can be reduced even more with better thermal insulation for the fridge.
Yes high energy density lithium cells can be had for about $220/kWh but those have lower cycle life 500 to 1000 cycles vs LiFePO4 that cost almost double at 350 to $400 but have a cycle life of 3000 to 6000 cycles.
So cost amortization is still better for LiFePO4 plus that is much safer in operation and also much safer for the environment.

Not sure if I mentioned here about the bicycle generator but I sure had discussions about this a few times.  I think that human generated power like a bicycle generators are a bad idea if energy is what you are after and not exercise.
Average human can not maintain more than 100W for long periods (hours) so what can be produced by even a fit human is extremely negligible probably similar to what a 30 to 40W PV panel that cost 30 to $40 and work for 25 years producing everyday as much as a fit human working hard to do that.  Of course the bicycle generator is also not free and most likely more expensive than a 40W PV panel.

[NiHaoMike]

With a heat pump water heater, the energy usage is quite small - on the order of a few hundred Wh per day. I actually built a heat pump water heater with a 1/2 ton compressor and that actually seems to be quite a bit bigger than necessary. There's little good reason to use a compressor smaller than that, since there's no real cost savings and the 300W or so it uses is a small load for any good sized off grid setup.

The main advantage of a stationary bicycle is that it would work when solar would not. Hence it would be a good complement to solar.

[electrodacus]

The main advantage of a stationary bicycle is that it would work when solar would not. Hence it would be a good complement to solar.

Solar actually works in all whether conditions.
A typical 250W panel will produce around 1.2 to 1.4kWh in a sunny day and at least around 0.1kWh in an overcast day so is never zero and most of the time much more than that minimum of 0.1kWh that only happens 5 to 8 days per year at my location. So a standard PV panel of that size 250 to 280W will produce in worst solar day as much as a human with one hour of pedaling and that panel will be 160 to $200 probably as much as a realy inexpensive bicycle generator. And 90% of the time that panel will produce an order of magnitude more than a single human on bicycle generator can do. People time is way to valuable for this even animals are way to valuable for this that is why you do not see a horse driven generator anywhere. Even if you do not care about the horse the food you will need to provide for him will cost more than the energy is worth.

[NiHaoMike]

I have in mind more of a backup in an off grid setup. Even if it's really just to ease the minds of those who are just starting to get into off grid, as an answer to "what if I run out of power and can't wait for the next day".

It is also worth noting that a 250W panel is definitely too big to be portable and even a 65W panel is bigger than what most people think of as portable. In contrast, there are already stationary bicycle generators that fold up into a very small package, intended for camping use. And then there are far more common crank chargers but most of those are designed cheap and don't actually work very well.

[HackedFridgeMagnet]

Wood can be cheaper if you live in a forest. And it warms you twice and keeps you fit.

https://grhgraph.wordpress.com/2010/02/11/chop-your-own-wood-and-it-will-warm-you-twice-henry-ford/

[electrodacus]

I have in mind more of a backup in an off grid setup. Even if it's really just to ease the minds of those who are just starting to get into off grid, as an answer to "what if I run out of power and can't wait for the next day".

It is also worth noting that a 250W panel is definitely too big to be portable and even a 65W panel is bigger than what most people think of as portable. In contrast, there are already stationary bicycle generators that fold up into a very small package, intended for camping use. And then there are far more common crank chargers but most of those are designed cheap and don't actually work very well.

There are portable PV panels that fold and are very light weight but of course they are at a premium in therms of cost because of the much lower production volume.
You can not just run out of power since you have SOC indicators so you know exactly how much energy you have stored and will know ho to use.
I live offgrid for about 4 years now with solar only and no other backup.

[electrodacus]

Wood can be cheaper if you live in a forest. And it warms you twice and keeps you fit.

https://grhgraph.wordpress.com/2010/02/11/chop-your-own-wood-and-it-will-warm-you-twice-henry-ford/

Even if you live in a forest and can get the wood for free the cost of heating will still be higher. You need to consider the burner that has a cost and limited life and if to that you add your time then cost is considerably higher than solar PV.
You will be fit in summer when you collect the wood for a few weeks but the air quality if you have the stove inside the house will more than negate the benefits