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

[oneday]

As to your energy use, if you converted your 3 a.c./ heat pumps, pool pump, and possibly even your blowers to 3-phase I would guess you could save a few kWh per day right there. Have you thought about that at all?

- - - Updated - - -

In the spring, summer, and fall I expect to have an energy surplus of up to a bit over double my expected usage... so definitely going to need to get creative.

I doubt this will be very hard for you.

 

[mitch672]

Max solar charge rate will be ~45kW (under 1/4C), or about 1,100 amps DC at low states of charge. In the spring, summer, and fall I expect to have an energy surplus of up to a bit over double my expected usage... so definitely going to need to get creative.

I can just see the large pond/damn & storage tanks on wk's property, the neighbors ask "what's this for?", wk just says "I use the surplus PV power to pump water from the pond to the storage tank during the day, and at night it's reversed to generate power from the stored water"

 

[Robert.Boston]

I can just see the large pond/damn & storage tanks on wk's property, the neighbors ask "what's this for?", wk just says "I use the surplus PV power to pump water from the pond to the storage tank during the day, and at night it's reversed to generate power from the stored water"

Nah, he has a net surplus overall. He'll string wires over to his neighbors and take them off-grid, too.

 

[wk057]

Nah, he has a net surplus overall. He'll string wires over to his neighbors and take them off-grid, too.

I was actually talking to one of my neighbors about this. lol.

Turns out I'm short on crimpable lugs for making the rest of my battery connections. I was able to hook up 20 modules before I ran out, though, for 106kWh of capacity hooked up so far.

Since I have some time to kill until my additional lugs arrive (possibly tomorrow) I decided to do an extended test under normal conditions. At 7AM this morning I switched the whole house over to battery power using the 20 out of 36 modules I have wired. I had brought all of the modules up to 24V (4V per cell) several days ago using my standalone charger to make sure they would be balanced when hooked up in the rack.

Started with 48V (4V per cell), not quite fully charged. Now 16.5 hours later I'm still at over 43V (around 3.6V per cell) and have output about 44kWh AC power from inverters, as measured at my load centers, using the 20 modules. Taking the inverter's 93% efficiency into account, that would mean I've pulled about 47kWh from the modules so far.

For those interested in the outback inverters, turns out with the latest update the low battery cut off point can be set as low as 36V. (Mini grid and grid zero and other settings still have a minimum of 44V for some reason, though). I went ahead and set it at 38V, which comes out to slightly above the cell voltages my P85D was at when I pulled into the supercharger with 1 rated mile the other week, according to the pre-input voltage shown on the dash (303V). I figure a discharge to this level should answer a few questions, primarily how the inverter behaves in real world conditions at this low input voltage (2V below spec sheet minimum).

I had run it down to this point before in my test setup, but from a lower starting voltage and less than ideal conditions for testing, just to see if the inverter would indeed keep operating. It did, but I didn't run it until it cut off. I plan to see exactly what the inverter does around this voltage. Will it cut off until the batteries reach the cut-in voltage? (42V), or will it continue to operate slightly below? Don't know for sure, but that's what I'm going to find out. I have two different voltage monitors online, both of which will alert me when the pack voltage gets to a few alert points.

Some highlights for today:

• Got hot today for the first time this year. Almost 80F. Upstairs got pretty warm, about 75 (beyond my comfort temp), and for fun I kicked the A/C on for a few to drop it back to 72.
• Drove to town in the P85D to grab lunch. Came back and charged at 20kW for ~25 minutes. Voltage drop on the car dash was something like 2-3V under full load. Impressive voltage regulating by the inverters. Grid doesn't do that!
• While the P85D was charging, some laundry switching was done. Washer and dryer both running, which caused the hot water heater to kick on also. Hit almost 30kW total house draw for a short time. Batteries didn't break a sweat (little over 1/4C draw... less than 1A per cell. Almost nothing, basically).
• Since the beginning of the test: 7 kWh for 24/7 network/server equipment; 5.8 kWh for my desk area (PC, monitors, switch, UPS, speakers, wifi AP, printer, etc); 5.2kWh of water heating; 3.2kWh of HVAC; 3.0 kWh of pool pump (was off most of the day); 1.7kWh of laundry; 1.5 kWh for my fiance's desk area; rest is misc (lights, tv, etc)

So far so good.

 

[cashie]

Great first test,sounds like the system is running as designed... :biggrin:

 

[MassModel3]

In a zombie apocalypse I'll still have power (until the zombies muck up my panels...).

It definitely looks like you'll have more than enough juice to power an electric fence or more high tech deterrence systems (rail guns, directed energy lasers, etc.) to keep the zombies away. :biggrin:

 

[FlasherZ]

Max solar charge rate will be ~45kW (under 1/4C), or about 1,100 amps DC at low states of charge. In the spring, summer, and fall I expect to have an energy surplus of up to a bit over double my expected usage... so definitely going to need to get creative.

Wardenclyffe NG. That is all.

 

[pgiralt]

Max solar charge rate will be ~45kW (under 1/4C), or about 1,100 amps DC at low states of charge. In the spring, summer, and fall I expect to have an energy surplus of up to a bit over double my expected usage... so definitely going to need to get creative.

This is going to be wk057's house soon...

 

[andrewket]

This is going to be wk057's house soon...

Hilarious. I want to know when tours are available (I'm only half kidding.. )

 

[okashira]

I was actually talking to one of my neighbors about this. lol.

Turns out I'm short on crimpable lugs for making the rest of my battery connections. I was able to hook up 20 modules before I ran out, though, for 106kWh of capacity hooked up so far.

Since I have some time to kill until my additional lugs arrive (possibly tomorrow) I decided to do an extended test under normal conditions. At 7AM this morning I switched the whole house over to battery power using the 20 out of 36 modules I have wired. I had brought all of the modules up to 24V (4V per cell) several days ago using my standalone charger to make sure they would be balanced when hooked up in the rack.

Started with 48V (4V per cell), not quite fully charged. Now 16.5 hours later I'm still at over 43V (around 3.6V per cell) and have output about 44kWh AC power from inverters, as measured at my load centers, using the 20 modules. Taking the inverter's 93% efficiency into account, that would mean I've pulled about 47kWh from the modules so far.

For those interested in the outback inverters, turns out with the latest update the low battery cut off point can be set as low as 36V. (Mini grid and grid zero and other settings still have a minimum of 44V for some reason, though). I went ahead and set it at 38V, which comes out to slightly above the cell voltages my P85D was at when I pulled into the supercharger with 1 rated mile the other week, according to the pre-input voltage shown on the dash (303V). I figure a discharge to this level should answer a few questions, primarily how the inverter behaves in real world conditions at this low input voltage (2V below spec sheet minimum).

I had run it down to this point before in my test setup, but from a lower starting voltage and less than ideal conditions for testing, just to see if the inverter would indeed keep operating. It did, but I didn't run it until it cut off. I plan to see exactly what the inverter does around this voltage. Will it cut off until the batteries reach the cut-in voltage? (42V), or will it continue to operate slightly below? Don't know for sure, but that's what I'm going to find out. I have two different voltage monitors online, both of which will alert me when the pack voltage gets to a few alert points.

Some highlights for today:

• Got hot today for the first time this year. Almost 80F. Upstairs got pretty warm, about 75 (beyond my comfort temp), and for fun I kicked the A/C on for a few to drop it back to 72.
• Drove to town in the P85D to grab lunch. Came back and charged at 20kW for ~25 minutes. Voltage drop on the car dash was something like 2-3V under full load. Impressive voltage regulating by the inverters. Grid doesn't do that!
• While the P85D was charging, some laundry switching was done. Washer and dryer both running, which caused the hot water heater to kick on also. Hit almost 30kW total house draw for a short time. Batteries didn't break a sweat (little over 1/4C draw... less than 1A per cell. Almost nothing, basically).
• Since the beginning of the test: 7 kWh for 24/7 network/server equipment; 5.8 kWh for my desk area (PC, monitors, switch, UPS, speakers, wifi AP, printer, etc); 5.2kWh of water heating; 3.2kWh of HVAC; 3.0 kWh of pool pump (was off most of the day); 1.7kWh of laundry; 1.5 kWh for my fiance's desk area; rest is misc (lights, tv, etc)

So far so good.

I assure you Tesla takes the cells well below 3.1667V (38V). The cells would be perfectly happy going down to 2.5V (30V) under load; Tesla probably goes to ~2.8V under load before a total shut down.
Please report back on how the inverters perform at 38V!

 

[wk057]

Love the comments!

30 hours 40 minutes on battery power using only 55% of the modules. Pretty successful test if you ask me.

Down to about 41V now (3.42V per cell)

In particular, even though I'll have to recharge from the grid this time, I find this chart satisfying:

From my grid panels.

 

[electrodacus]

I assure you Tesla takes the cells well below 3.1667V (38V). The cells would be perfectly happy going down to 2.5V (30V) under load; Tesla probably goes to ~2.8V under load before a total shut down.
Please report back on how the inverters perform at 38V!

For long battery life in stationary storage applications it makes no sense to go below 3.2 or 3.3V. If you look at a discharge curve for this type of batteries you will see that is less than 5% capacity left at 0.25C, the additional stress on battery is not worthed for this 5% or less.

 

[Pollux]

Max solar charge rate will be ~45kW (under 1/4C), or about 1,100 amps DC at low states of charge. In the spring, summer, and fall I expect to have an energy surplus of up to a bit over double my expected usage... so definitely going to need to get creative.

Electric lawn mowers.

Electric fences.

Neighbors.

Tesla coils.

Electric catapults.

Make your own aluminum.

Lasers (with or without cats and/or sharks).

 

[okashira]

For long battery life in stationary storage applications it makes no sense to go below 3.2 or 3.3V. If you look at a discharge curve for this type of batteries you will see that is less than 5% capacity left at 0.25C, the additional stress on battery is not worthed for this 5% or less.

There is 15% capacity at C/3 between 3.3V and 2.5V. The cells also degrade more slowly at lower voltage so I'd choose to use the bottom 90% over the top 90% of capacity if I was being conservative.

 

[wk057]

I've been working through the numbers, and as of now I'm actually planning on using ~8% to ~92% most likely, in the final setup, for about 155-160kWh usable with an 8% buffer on either end. I'll be able to manually override this, if absolutely needed, but I don't see why I would ever really need to.

Considering that the pack will be discharged somewhat every night (not fully, obviously) I figure it will never be at either end of the SoC spectrum very long. At most a handful of hours until the sun goes down. For low SoC cutoff, this also shouldn't be for very long, either. Let's say the pack dies in the middle of the night after a few stormy days with minimal solar input. Well, the sun is going to rise again, and will input something into the pack even on a rainy day. So again, at most only a handful of hours at this level.

I have a module that is damaged that I'm going to break the cells out of and do some charge/discharge curve testing.... likely to the point in which some are permanently damaged or even have a thermal event. Yep, I want to know what these babies are capable of!
But, since there are 444 of them... with most of them intact.... I don't care about destroying a few. *cue video of touching 4/0 cable to each terminal*

I may throw some up for sale cheap (mostly shipping and my time making sure they are at ~50% SoC for shipping) by request for some folks here looking to do their own tests/experiments. Since they're unprotected I'll only sell them to folks who actually have an idea what they're doing with them, and only a few of them still, maybe sets of four. I wouldn't use them in a 18650 flashlight, for example.

 

[MassModel3]

I have a module that is damaged that I'm going to break the cells out of and do some charge/discharge curve testing.... likely to the point in which some are permanently damaged or even have a thermal event. Yep, I want to know what these babies are capable of!

Oohhh, now that sounds like a fun side project! We want videos, please, especially if you can make one pop like an over-charged capacitor! :scared:

 

[electrodacus]

There is 15% capacity at C/3 between 3.3V and 2.5V. The cells also degrade more slowly at lower voltage so I'd choose to use the bottom 90% over the top 90% of capacity if I was being conservative.

I see about 7% or 8% here from 3.2V to 2.5V at C/3 is not worthed to go below 3.2V

 

[FlasherZ]

I may throw some up for sale cheap (mostly shipping and my time making sure they are at ~50% SoC for shipping) by request for some folks here looking to do their own tests/experiments. Since they're unprotected I'll only sell them to folks who actually have an idea what they're doing with them, and only a few of them still, maybe sets of four. I wouldn't use them in a 18650 flashlight, for example.

A single-cell flashlight probably isn't a problem... but multi-cell OTOH.

That said, I have a 2-cell light in which I occasionally use 16850B's (while my protected cells are recharging). Livin' life on the dangerous side!

 

[wk057]

A single-cell flashlight probably isn't a problem... but multi-cell OTOH.

That said, I have a 2-cell light in which I occasionally use 16850B's (while my protected cells are recharging). Livin' life on the dangerous side!

Just also keep in mind these cells have no insulation around the casing, either. So, the casing is negative. Probably bad for a metal series tube

 

[FlasherZ]

Just also keep in mind these cells have no insulation around the casing, either. So, the casing is negative. Probably bad for a metal series tube

Good point, forgot about that. I have some leftover Tesla stickers that will work. Or shrink wrap.