Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Plan: Off grid solar with a Model S battery pack at the heart

This site may earn commission on affiliate links.
Was a little celebratory electrician on electrician crime yesterday... wish I had the GoPro in video mode...

pool-push-1.jpg


pool-push-2.jpg


pool-push-3.jpg


pool-push-4.jpg
 
wk057,

Wonderful project! I have been following it for quite some time. Congratulations for getting all the solar on line.

I was wondering if you had considered splitting it into two independent systems? You already have two load panels, each with their own transfer switch. The benefit would be in case of a failure, only half the circuits would be out (or running from the grid). You could even wire one of the charge controllers to be switchable between the two systems to ensure that neither battery bank gets too low of a state of charge.
 
wk057,

Wonderful project! I have been following it for quite some time. Congratulations for getting all the solar on line.

I was wondering if you had considered splitting it into two independent systems? You already have two load panels, each with their own transfer switch. The benefit would be in case of a failure, only half the circuits would be out (or running from the grid). You could even wire one of the charge controllers to be switchable between the two systems to ensure that neither battery bank gets too low of a state of charge.

I had considered it, however the peak loads on each panel would be more than half of the inverter capacity. Max is about ~48kW on the one panel and ~48kW max on the other. In total it would be more than my inverters could handle if every large load were maxed out, but with them as one system I don't really need to worry about it much at all. The chances of me charging both cars while running the over, dryer, both hot water heaters in resistive heat mode, resistive aux heat for all three HVAC units plus their heat pumps, along with the pool pump and other general loads........ is pretty slim.

I could see both cars charging while cooking and doing laundry though, while the HVAC was running which would be over 55kW.

Overall it made more sense to have the whole system be available for handling loads on either panel. My setup can do 64kW continuous, and 72 kW for 30 minutes as a surge rating. Without Model S charging included my max possible load is under the 64kW continuous rating. Since I'm redesigning the HPWC control circuitry to be smarter and work with my system better I will have a few lines of code in there that basically slow charging if charging plus other loads happen to push things above the 64kW mark for very long. In general, though, I bet I could just go about business as usual and it would probably never be a problem.

Now if only these clouds would go away today :(
 
Last edited:
I was wondering if you had considered splitting it into two independent systems? You already have two load panels, each with their own transfer switch. The benefit would be in case of a failure, only half the circuits would be out (or running from the grid). You could even wire one of the charge controllers to be switchable between the two systems to ensure that neither battery bank gets too low of a state of charge.

.... isn't each charge controller and each inverter already independent? Why would the failure of 1 charge controller or 1 inverter effect the rest of the system?
 
.... isn't each charge controller and each inverter already independent? Why would the failure of 1 charge controller or 1 inverter effect the rest of the system?

In theory. Although one charge controller dying would require a quick communication wiring change, and one inverter dying may or may not require a communication wiring change... but that's about it.
 
Awesome to see your panels up. Do you have a way to burnoff excess solar power when your batteries are full and not enough loads to keep charge controllers in constant MPPT mode? You are not planning back-feeding the grid, correct? There may be days when you cannot "absorb" all available solar energy and your daily solar power graph will not show all possible power you could have made which would skew your capacity factor lower than it could have been. I ran into that problem with my 3 panel experimental off-grid setup. To solve this I utilized Aux output on charge controller set up in diversion relay mode. The signal would drive power MOSFET switching on load resistors on DC bus at preset battery voltage. Here is video of it. I think this method can be improved using micro controller for some lithium battery management techniques to prolong battery life.
 
Awesome to see your panels up. Do you have a way to burnoff excess solar power when your batteries are full and not enough loads to keep charge controllers in constant MPPT mode? You are not planning back-feeding the grid, correct? There may be days when you cannot "absorb" all available solar energy and your daily solar power graph will not show all possible power you could have made which would skew your capacity factor lower than it could have been. I ran into that problem with my 3 panel experimental off-grid setup. To solve this I utilized Aux output on charge controller set up in diversion relay mode. The signal would drive power MOSFET switching on load resistors on DC bus at preset battery voltage. Here is video of it. I think this method can be improved using micro controller for some lithium battery management techniques to prolong battery life.

Very cool. I like your unistrut based panel racking. That was actually my initial idea as well, but too much work for so many panels.

Currently I have nothing really setup for handling excess power aside from the charge controllers just not taking it in. When that happens I'm going to set it up so that my graph changes color, indicating that it's not the maximum power available. If it turns out this happens often enough (likely will in the summer) then I'm going to investigate a dump into the grid, assuming they can actually pay me real money (not just credit, since that would be useless to me).

My current plan is to utilize existing household items as the initial diversion load in an automated way. Basically, drop or raise the temperature in the house a degree or two utilizing all HVAC systems (~9kW). Heat the hot water another degree or two (5kW if using the resistive elements). Top off the Model S's if they're below 90%, and maybe even bump them to 91 or 92%. Doing laundry or running the dishwasher when this happens would make sense too, although can't really automate that. I plan on getting an electric pool heater (probably 10kW) and will use that (even in the winter, maybe) also if the water isn't already too warm. Basically plenty of normal things I can think to dump power into.

If I still have excess beyond all that... well, not much I can about it I guess, currently. A DC based resistive heater up to 44kW would be pretty crazy. Could heat the neighborhood. :p But I don't really see the need to find a way to really do anything with the excess power if I can't do something useful with it. Dumping it into a resistor as heat is basically the same as just not utilizing the power from the array since the panels will get warmer since the power isn't being absorbed... and I don't really see any harm in that since it's what would happen if they just weren't hooked up anyway.
 
...But I don't really see the need to find a way to really do anything with the excess power if I can't do something useful with it. Dumping it into a resistor as heat is basically the same as just not utilizing the power from the array since the panels will get warmer since the power isn't being absorbed... and I don't really see any harm in that since it's what would happen if they just weren't hooked up anyway.

Absolutely. There is no harm not capturing all possible power panels can produce. The purpose of keeping panels at max utilization was for data logging sake only. I wanted to measure how much energy my panels could produce.
 
If the credit is sufficient and annual you could use it to reduce the cycling stress on the batteries and displace winter consumption... do you expect to need the grid in the winter?

Winter loads are significantly lower than summer loads (no pool, less HVAC use overall, less driving). So, probably won't need the grid much, if at all. I sized the system with winter usage and winter PV production in mind, but it probably will get a little close at time, especially if I do a lot more driving than normal one week.
 
I had considered it, however the peak loads on each panel would be more than half of the inverter capacity. Max is about ~48kW on the one panel and ~48kW max on the other. In total it would be more than my inverters could handle if every large load were maxed out, but with them as one system I don't really need to worry about it much at all. The chances of me charging both cars while running the over, dryer, both hot water heaters in resistive heat mode, resistive aux heat for all three HVAC units plus their heat pumps, along with the pool pump and other general loads........ is pretty slim.

I could see both cars charging while cooking and doing laundry though, while the HVAC was running which would be over 55kW.

Overall it made more sense to have the whole system be available for handling loads on either panel. My setup can do 64kW continuous, and 72 kW for 30 minutes as a surge rating. Without Model S charging included my max possible load is under the 64kW continuous rating. Since I'm redesigning the HPWC control circuitry to be smarter and work with my system better I will have a few lines of code in there that basically slow charging if charging plus other loads happen to push things above the 64kW mark for very long. In general, though, I bet I could just go about business as usual and it would probably never be a problem.

Now if only these clouds would go away today :(

How does the model S react to changes of the pilot signal that communicates max current allowed? I assume it will adjust down. Will it automatically adjust up?
 
Very cool. I like your unistrut based panel racking. That was actually my initial idea as well, but too much work for so many panels.

Currently I have nothing really setup for handling excess power aside from the charge controllers just not taking it in. When that happens I'm going to set it up so that my graph changes color, indicating that it's not the maximum power available. If it turns out this happens often enough (likely will in the summer) then I'm going to investigate a dump into the grid, assuming they can actually pay me real money (not just credit, since that would be useless to me).

My current plan is to utilize existing household items as the initial diversion load in an automated way. Basically, drop or raise the temperature in the house a degree or two utilizing all HVAC systems (~9kW). Heat the hot water another degree or two (5kW if using the resistive elements). Top off the Model S's if they're below 90%, and maybe even bump them to 91 or 92%. Doing laundry or running the dishwasher when this happens would make sense too, although can't really automate that. I plan on getting an electric pool heater (probably 10kW) and will use that (even in the winter, maybe) also if the water isn't already too warm. Basically plenty of normal things I can think to dump power into.

If I still have excess beyond all that... well, not much I can about it I guess, currently. A DC based resistive heater up to 44kW would be pretty crazy. Could heat the neighborhood. :p But I don't really see the need to find a way to really do anything with the excess power if I can't do something useful with it. Dumping it into a resistor as heat is basically the same as just not utilizing the power from the array since the panels will get warmer since the power isn't being absorbed... and I don't really see any harm in that since it's what would happen if they just weren't hooked up anyway.

Does temperature impact the longevity of the panels? (Solar newbie here). Sounds like you need more batteries :)
 
I'm wondering why you photo shopped your back yard by your pool and solar array with trees?

Hmm? You mean the simulation I did?

How does the model S react to changes of the pilot signal that communicates max current allowed? I assume it will adjust down. Will it automatically adjust up?

Not 100% sure, but Tesla says it follows the J1772 spec, which would mean that it would increase again. I haven't had time to test this yet.

Does temperature impact the longevity of the panels? (Solar newbie here). Sounds like you need more batteries :)

Not that I'm aware. Lower temperatures increase voltage output, but that's about it.

Later I'll fix my graph generator to change the color of the graph when power is being wasted, and maybe estimate how much power based on the voltages (higher PV voltage while wasting power).

Anyway, I've filled the house pack and have nothing left to do with the power. The taper that's happening right now is the constant-voltage stage (I have it set below the normal 100% point for these cells) happening and excess power is being wasted. :(

I guess that's a good problem to have with a setup like this.
 
Looks like a few of my charge controllers need their battery side voltage calibrated again. I dropped the constant voltage side voltage down a bit to make sure I don't overcharge the pack. Better safe than sorry.