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.

wk057

Former Tesla Tinkerer
Feb 23, 2014
6,499
17,123
X
EDIT/Update:
Near real time stats for my setup at wk057.solar. :)

EDIT/Update:
Since I started this thread this project has undergone many changes and updates and has turned into something larger than originally described. Most of that info is in later posts in this thread, and I'll try to consolidate some of it here later on.

My project has been featured on Bloomberg Business, Hack-A-Day (2), and quite a few other places around the interweb. So, I'll definitely work on getting a summary setup soon.

---

Some specs from later on in the thread:

  • Battery capacity
    • 36 modules from 2.25x Tesla 85kWh packs
    • 191.25 kWh (DC side)
    • ~4,200 Ah
    • 43.2V nominal @ 3.6V per cell
    • 15,984 cells (!)
  • Inverter capacity (8x Outback Radian GS8048A)
    • 240VAC @ 60Hz w/neutral
    • 64kW continuous AC output
    • 30 minute surge: 72kW
    • 5 second surge: 96kW
    • 100ms surge: 135.76kW
    • Grid->Battery Charging Capacity: 57kW
    • Expected AC output from pack after safe SoC window and efficiency considerations: ~160 kWh usable AC
  • PV Capacity (In Progress)
    • 102 Panels @ 435W (20% efficiency)
    • 44,370 Watts DC
    • Split into 17 sets of 6 panels (3 parallel of 2 in series)
    • 17 individual MPPT charge controllers (Midnite Solar Classic 200)

---

Update 2015-05-15: Roof portion of my system is completed.

Solar Install - 15.7kW Roof System - Final Compilation (Time lapse, drone shots, photos) - YouTube


Solar Install - 28.7kW Ground System (Residential) - Part 2 - Final - YouTube


---


So, I have this crazy idea about using an 85kWh Model S battery pack (from a salvage Model S) as the heart of a custom off-grid solar power setup somewhere around 25kW DC.

I assume others have thought about this...

I considered making a setup where the solar panels would be setup in a configuration to produce something like 450VDC+, then making a buck converter for high efficiency charging directly from the panels. The most I've seen the battery voltage was about 404V during a 100% charge at a supercharger, so, this should be possible. Little bit of microcontroller brains to keep things sane and all should be well.

Inverter should be simple, since HV inverters already exist. Might need to step the voltage down a hair since a lot do not operate at >400VDC input. But, that's not the hardest part of this project by a long shot.

That aside, does anyone have any information about the pack itself? I'm working on physically getting one to tinker with, which will help, but I'd be curious if anyone else has done any of this already and we could save each other some leg work.

I'm also wondering if Tesla would sell any of the parts needed to interface with it, such as the high voltage connectors. I might need to just get a full salvage Model S to do this (would be cool, because then I might end up with 170kWh of battery...). Things like hooking to the coolant loop and adding a radiator and things should be trivial. Accessing the existing BMS might be hard, though, but might not be needed. Since the pack will be stationary I'll probably pop the cover off and go from there.

Even if the pack+interface items cost say, $40k, thats only $470/kWh. I refuse to go lead acid again for an off-grid setup (my previous much more modest setup used lead acid, and it was just a pain...). But for lithium ion packs of comparable size, you're looking at a price of at least that, plus a lot of labor to build the pack itself. I came up with a cost of around $500-600/kWh using off the shelf DIY EV lithium ion batteries, not counting any assembly or wiring.

Anyway, my background includes a bunch of experience with solar and related items, as well as some electrical engineering, circuit design, programming, etc. So, this is a doable project assuming I can get the materials.

Just looking for input. :)

(Not sure if this is the right sub-forum, but it seems appropriate)

- - - Updated - - -

Just to note, part of the load of this system will be charging my Model S. I'd love to get it 100% off grid. :)
 
Last edited:
You should be able to configure your solar strings to fine tune the DC voltage (I would err on the side of a higher voltage and plan to trim it down to the correct level).

There should be no problem inverting 450 VDC (the Solectria inverters that I use operate from 200 to 550 VDC and have a maximum input amount of 600 VDC - they were running at 375 VDC today) other string inverters have even higher DC voltage capabilities.

(I have been thinking similar thoughts and would be very interested to see where your project goes.)
 
I admittedly hadn't looked into the HV input inverters in a lot of detail as of yet, but the majority of the ones I've seen are grid tie. Now I'm sure I could setup a microgrid and trick the anti-islanding nonsense into playing ball, but an actual off grid device would be preferred.
 
It is all possible.

Is there a compelling reason the OP wants to be off grid? I am happy using the grid as my backup. I am on a Tou EV rate with SCE (Southern California Edison) and on a Net Metering arrangement for my Solar. I have no bill but am a net consumer of kWhrs. This is because I generate power during the day (at $0.26/kWhr and charge my car at night (at $0.09/kWhr). I realize that my opportunity to game the system may be unique to my tarriff situation but that is the economic analysis that drove my decision. I didn't see any economic analysis in the OP's idea and that may be the difference between a "crazy idea" and a practical solution.

I don't know what rates you have in New Jersey, but feel free to share and you may get some cost effective ideas. Of course it all depends on your cost of capital. Again in my case, my cost of capital made the investment in solar attractive and the investment in batteries and inverters not attractice. To each his own.
 
Last edited:
  • Like
Reactions: h2ofun
It is all possible.

Is there a compelling reason the OP wants to be off grid? I am happy using the grid as my backup. I am on a Tou EV rate with SCE (Southern California Edison) and on a Net Metering arrangement for my Solar. I have no bill but am a net consumer of kWhrs. This is because I generate power during the day (at $0.26/kWhr and charge my car at night (at $0.09/kWhr). I realize that my opportunity to game the system may be unique to my tarriff situation but that is the economic analysis that drove my decision. I didn't see any economic analysis in the OP's idea and that may be the difference between a "crazy idea" and a practical solution.

I don't know what rates you have in New Jersey, but feel free to share and you may get some cost effective ideas. Of course it all depends on your cost of capital. Again in my case, my cost of capital made the investment in solar attractive and the investment in batteries and inverters not attractice. To each his own.

Actually the system wouldn't be in NJ, it would be in NC.

But the drive for off grid is that to me, the concept of net metering is more politics than anything. At any moment something with that arrangement could change and I could end up with a bill for grid use instead of using the grid as a battery, per se. Best to just use my own battery and let the grid be the grid.

I'd still have some power from the grid, but in the end I'd want it for just backup.
 
For true off grid battery system, instead of just battery back up system, I would go with nickel-iron batteries over lithium-ion. Nickel-iron last practically forever, with daily near full range charge/discharge cycles.

Check out Nickel-Iron NiFe batteries by Iron Edison


I've thought about going off grid, but it is still too expensive to do "just for fun". In CA net metering is locked in for current users for the next 20 years, so I have some time to think about it :)
 
Well, there you have it, it is unique, in this great country of ours, that you would find different solutions because of local politics. I have heard the term, "Energy Democracy" used to describe the choice to choose where we want to get our energy. That is why I have solar panels and drive electric cars.
 
For true off grid battery system, instead of just battery back up system, I would go with nickel-iron batteries over lithium-ion. Nickel-iron last practically forever, with daily near full range charge/discharge cycles.

Check out Nickel-Iron NiFe batteries by Iron Edison


I've thought about going off grid, but it is still too expensive to do "just for fun". In CA net metering is locked in for current users for the next 20 years, so I have some time to think about it :)

The grandfathering of NEM for 20 years may not be the guarantee that you are assuming...My understanding is that per the recent California PUC decision, the concept of NEM is indeed grandfathered (selling your daytime excess back to the grid and receiving retail pricing credit), but the terms and conditions that you are enjoying today are not locked in (using the grid as your battery for free, for example)...Expect the TOU peak period to shift to later in the day to mimic real world grid conditions (i.e. 2-9 pm), and a monthly customer charge of some sort to cover the grid services that solar customers are currently not paying for.
 
I can understand, in my case, the rational for a fee. But what about the people who hve solar and generate more than they use. At the end of each year, they get a paltry $0.04/ per kWhr. At that kind of rate the utility companies ought to be covering their grid costs.
 
The grandfathering of NEM for 20 years may not be the guarantee that you are assuming...My understanding is that per the recent California PUC decision, the concept of NEM is indeed grandfathered (selling your daytime excess back to the grid and receiving retail pricing credit), but the terms and conditions that you are enjoying today are not locked in (using the grid as your battery for free, for example)...Expect the TOU peak period to shift to later in the day to mimic real world grid conditions (i.e. 2-9 pm), and a monthly customer charge of some sort to cover the grid services that solar customers are currently not paying for.

That's fine. I will soon (next week actually)be producing very close to my total consumption. I don't need the TOU credit advantage to break even.

But not have net metering would break everything. Even generating 3x my consumption would not be enough if they only pay $0.05 per kWh I produce during the day, but charge $0.16 even for super off peak usage.

I'm also okay paying a reasonable amount for grid support. But if I do, they need to pay me actual peak rates when I generate power during high demand, etc.

If they start charging my neighbor $.49 for a kWh I generate while insisting I pay the grid and they only pay me $.05. I will be off grid as fast as I can get my electrician to install the equipment.
 
Last edited:
For true off grid battery system, instead of just battery back up system, I would go with nickel-iron batteries over lithium-ion. Nickel-iron last practically forever, with daily near full range charge/discharge cycles.

Check out Nickel-Iron NiFe batteries by Iron Edison


I've thought about going off grid, but it is still too expensive to do "just for fun". In CA net metering is locked in for current users for the next 20 years, so I have some time to think about it :)

To get an 85kWh pack from that company would be almost the cost of a brand new base model 85kWh Model S... so, no. lol

Also:
Recommended max charge rate: C/5
Recommended max discharge rate: C/5

Which would be 17kW charging/discharging with an $80k 85kWh pack... no thanks.
 
Expect the TOU peak period to shift to later in the day to mimic real world grid conditions (i.e. 2-9 pm), and a monthly customer charge of some sort to cover the grid services that solar customers are currently not paying for.

PG&E's EV-A TOU rate already defines the weekday peak to be 2-9 pm. This plan also includes a small monthly charge for the grid connection.
 
Well, there's the easy way, and the right way...

The easy way would be to run the solar panels in to a normal solar inverter, and then the output of that inverter in to a model S charger (from the same salvage vehicle you found the battery pack) and then put another inverter on the battery output.

The right way would involve direct DC charging.

I guess one of the big questions is around the Tesla BMS, how much of the battery management is done inside the pack itself vs elsewhere. If it's all inside the battery it should be pretty straight forward, but if a lot of the battery management is handled elsewhere then the pack may not have the same lifespan if used outside of the vehicle.
 
Well, keep in mind that we're talking loads on the order of a household here, where the Model S will draw hundreds of kW at times and charge at 120kW at times. I think in off-grid use it will be much easier on the battery. Even using it to charge a Model S with a HPWC @ 240V/80A (probably the largest load I personally would have) we're talking a load of like 0.25C, with no solar input. Flooring it in a P85 is something like 3.75C
 
I'm more worried that there'd be nothing stopping you from over-charging the battery, or running it completely flat. the rate of discharge should be almost negligible compared to what it's designed for. Not to mention that many BMS systems do cell level balancing (or at least small group of cell)
 
I'm more worried that there'd be nothing stopping you from over-charging the battery, or running it completely flat. the rate of discharge should be almost negligible compared to what it's designed for. Not to mention that many BMS systems do cell level balancing (or at least small group of cell)

Well, since I'd be designing the system from scratch, obviously I would set it up in such a way that I don't over-charge or over-discharge the batteries.

It would be awesome to interface with the existing BMS somehow, but, I'm doubting that it will be easy.