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

[Ampster]

Are those batteries lead acid types?

No, they are Voltronix (rebranded Thundersky) LIFEPO4s, which I found used for $125/kWhr. I have 3 parallel groups and 16 groups in series. They have been great. I initially bought a battery management system but they have stayed balanced without it. Looked at used Volt and Leaf cells as well.

 

[miimura]

You should be able to AC couple them in the event of a grid failure... it's not ideal long-term since the Radian can't control the charge rate of the batteries but it'll work in a pinch. The Radian can keep the batteries from getting overcharged... it will increase frequency to 60.6hz which will cause your Enphase inverters to trip off.

Outback AC coupling

I've read the documentation for the Radian. It is possible to program the charging rate of the non-interactive grid input. If you did a little programming and had either an energy monitor or a SmartMeter that you could read in near real-time, you could program the system to absorb your grid export during TOU periods of lower reimbursement and allow all the solar to go to the grid during periods of higher reimbursement. You would of course power your loads from battery during the peak rate hours. I also have an existing Enphase solar system, so I've thought about this before.

The SMA HVDC inverter linked above is somewhat underwhelming compared to an 8kW Radian. The SMA one is only 2,500W. It seems to be designed to pair with something like a PowerWall that itself has lower output limitations. A salvaged intact EV battery pack would easily be able to service 8kW-20kW peak. I think there would be a market for "adapters" that would provide cabling and simple logic for interfacing with EV battery packs. The adapter would only have to do simple things like provide a mating pack connector and activate the contactor inside the pack when the inverter says it's OK. If it had integration with the pack's BMS, even better.

 

[SolarWannabe]

I think this video is probably pretty useful for noobs like myself wanting to replicate J's battery bank.

 

[nativewolf]

Very interesting week for Tesla watchers. I wonder if the Model 3 reservations have just put a giant kick in the nuts to the powerpack and powerwall roll-outs? I mean, it will be 300k or close to it in only a week. That's a huge demand increase.

 

[nativewolf]

Very interesting week for Tesla watchers. I wonder if the Model 3 reservations have just put a giant kick in the nuts to the powerpack and powerwall roll-outs? I mean, it will be 300k or close to it in only a week. That's a huge demand increase.

EDITED: relevancy here is that maybe I won't be able to source a powerpack afterall. That was my plan prior to the launch of the m3, now...

well there will be a gap between production and the production of the powerwalls and powerpacks in volume.

 

[Witzgall]

Hi, I have read much but not all of this tread. I am interested in the potential use of these batteries as a house battery for our sailboat. We live aboard. On our previous boat, we used lifepo4 for a 12v system.

This boat is 24v nominal. To the OP, do you have a discharge curve you could share? The nominal voltage of these 85d modules is about 2v less them what our system was designed for. I am trying to get a handle on the real world voltages as the modules are drawn down in a fractional c senario. I should not have a system voltage lower than 20v, so at what capacity am I at when we are at 20v?

We currently have a victron 24v inverter charger, and I will be adding solar soon. This is used while plugged into shorepower, and run from the 4kw genset when not, providing up to 70a of current. It appears that the charger settings can be changed to values that will work with the tesla packs.

We also charge with an alternator on the mail engine, that regulator should play nice with the packs as well.

Here is a link to the our system as it stands right now, not much to show yet
https://vrm.victronenergy.com/site/share/0ceb96ae?mainTabId=live-feed

Chris

 

[redevries]

Hi, I have read much but not all of this tread. I am interested in the potential use of these batteries as a house battery for our sailboat. We live aboard. On our previous boat, we used lifepo4 for a 12v system.

This boat is 24v nominal. To the OP, do you have a discharge curve you could share? The nominal voltage of these 85d modules is about 2v less them what our system was designed for. I am trying to get a handle on the real world voltages as the modules are drawn down in a fractional c senario. I should not have a system voltage lower than 20v, so at what capacity am I at when we are at 20v?

We currently have a victron 24v inverter charger, and I will be adding solar soon. This is used while plugged into shorepower, and run from the 4kw genset when not, providing up to 70a of current. It appears that the charger settings can be changed to values that will work with the tesla packs.

We also charge with an alternator on the mail engine, that regulator should play nice with the packs as well.

Here is a link to the our system as it stands right now, not much to show yet
https://vrm.victronenergy.com/site/share/0ceb96ae?mainTabId=live-feed

Chris

Unfortunately, the OP isn't (very) active on this forum anymore. A lot of details are available though the links in the initial post, which point to Jason's (the OP) website. HTH.

 

[Cruisin]

Thats a pretty sane number. I'll probably stick closer to C/4 with C/2 as a peak discharge rate.

I sell "NEW" Volt battery modules with a much lower cost per kw and easier to install. Additionally, the modules have a socket for 3rd part BMS systems which work very well with the VOLT batteries. Used primarily for off grid systems, most customers are in Florida, California and Porta Rico. I can be reached at [email protected]

 

[Cruisin]

Hi, I have read much but not all of this tread. I am interested in the potential use of these batteries as a house battery for our sailboat. We live aboard. On our previous boat, we used lifepo4 for a 12v system.

This boat is 24v nominal. To the OP, do you have a discharge curve you could share? The nominal voltage of these 85d modules is about 2v less them what our system was designed for. I am trying to get a handle on the real world voltages as the modules are drawn down in a fractional c senario. I should not have a system voltage lower than 20v, so at what capacity am I at when we are at 20v?

We currently have a victron 24v inverter charger, and I will be adding solar soon. This is used while plugged into shorepower, and run from the 4kw genset when not, providing up to 70a of current. It appears that the charger settings can be changed to values that will work with the tesla packs.

We also charge with an alternator on the mail engine, that regulator should play nice with the packs as well.

Here is a link to the our system as it stands right now, not much to show yet
https://vrm.victronenergy.com/site/share/0ceb96ae?mainTabId=live-feed

Chris

When you chose to use LifePO4 Chinese junk, you didn't do your homework. How many EV companies use that chemistry?

 

[nwdiver]

When you chose to use LifePO4 Chinese junk, you didn't do your homework. How many EV companies use that chemistry?

LiFePO4 isn't used in EVs due to it's lower Energy and Power density. That chemistry is generally more robust than what EVs use. Keeping it at a high SOC is also less detrimental then EV chemistries.

Different chemistries have different benefits... IMO the BMS is more critical than the cells...

 

[googlepeakoil]

This project is absolutely amazing, lets hope we can get Elon to chime in with his opinions once it's up and running. So annoying that TSLA being public limits what we can ask of the CEO and what he can comment on.

Question... When this thread first started I had assumed the non-battery hardware would be relatively simple, but now that it's progressing there are rows and rows of junction boxes, miles of conduit and it looks like a NASA lab or something. Can someone explain to the electrically illiterate why this setup requires so many different pieces of hardware seemingly in parallel? Is it to push charge out to the thousands of individual cells?

Thanks and again, amazing work you're doing here.

Because the system is about 6-10x larger than most domestic installations - he's had to kinda "8x" everything. Hence the massive duplication.
A Tesla Powerwall can supply about 4kw to a house - peak. That's enough to run 40x 100w bulbs - but only 70% of a 5.5kw oven.
A powerwall can store about 8kwh - this guy has about 20x that capacity!!
This guy's setup can provide about 7x the current so he could run all appliances in the house + charge a Tesla at full capacity off of stored battery alone.
Also enough power to run most houses off-grid for 2 weeks - or his big / high-draw house for approx 3-5 days. Less if he's doing long-drives in the 2 Tesla's each day.

 

[AntronX]

WK posted 1 year update on his site. Main points:

• 43,929 kWh generated by solar array and utilized (Average 120 kWh/day, usable AC power)
  
• 20,576 kWh discharged from the batteries
  
• 114 equivalent cycles on the storage batteries
  
• 11,259 kWh used for Tesla Model S charging (over 33,000 miles)
  
• Best production day: 2016-03-21 @ 260.7 kWh (Charged three Model S that day)
  
• Worst production day: 2016-01-22 @ 0.02 kWh (All panels covered in several inches of snow)
  
• On 222 out of 366 days (60.7%) the storage batteries reached full capacity before sunset and some amount of incoming power was wasted.
• 703 kWh bought from the grid (1.6% of total usage).

Looks like his solar capacity factor was only 11.3% (44.4kW DC solar system), while PVWatts estimates 16.4% or 64,000MWh/year for grid-tie system at his location (Hickory, NC). That's 31% of system potential that was uncaptured due to inability to store that energy and possible loss due to less that optimal panel placement.

44,000kWh generated at $0.1/kWh grid power cost saved $4,400 in one year. But according to my napkin math, system cost was ~$100k for major components + ~$20K for installation. At that rate it will take 36 years to pay for itself. If you try to replicate this project with non-DIY commercial system, it would take the cost into $300K range at today's prices.

 

[Cosmacelf]

Nice napkin math, AntronX. The system would be much more economical in a southern climate. Far fewer zero solar production days, which would result in needing a smaller battery pack, and probably a smaller solar array.

But even so, this system is very complex with lots of parts that must be monitored and kept running. If you factor in cost of the time spent on designing, building and maintaining this system, it makes it even less economical.

But it's a great hobby!

 

[AntronX]

But it's a great hobby!

Totally agree! If I had the money I would do the same thing.

 

[SeminoleFSU]

I hope installers are looking at what Jason did and finding ways to stand on his shoulders. My hope is to keep my MS until the wheels fall off, and hopefully by then someone can offer a service to remove the old battery and repurpose it as home energy storage... would be very nice if they could figure out a way to leverage the pack as is, without the tear down... that way you could maybe leverage the existing cooling and other parts (sensors, etc) I think Jason tried this at the beginning but couldn't find an inverter that would work with the pack's high voltages. All of this is way over my head, I just know what I'd like to see out of it, someone creating a business out of repurposing old Model S packs as home storage as a service

 

[Cosmacelf]

The killer in any business is labor costs. Tearing apart an old pack is very labor intensive, to the point where the most economical use of an old pack may very well be what Tesla is planning on doing with them. Which is stuff them into an automated mill that crushes them and separates out the useful metals and recycles the metals into new batteries.

 

[Cosmacelf]

Solar plus batteries just isn't an economical alternative today in the US mainland to grid electricity. it is economical today in remote areas that don't have grid electricity, or where grid electricity is very expensive, like on an island which heretofore relied on import diesel and generators.

 

[SeminoleFSU]

The killer in any business is labor costs. Tearing apart an old pack is very labor intensive, to the point where the most economical use of an old pack may very well be what Tesla is planning on doing with them. Which is stuff them into an automated mill that crushes them and separates out the useful metals and recycles the metals into new batteries.

That's why I said use the pack as is. As an intact pack ;-)

 

[SeminoleFSU]

Solar plus batteries just isn't an economical alternative today in the US mainland to grid electricity. it is economical today in remote areas that don't have grid electricity, or where grid electricity is very expensive, like on an island which heretofore relied on import diesel and generators.

Doesn't have to be economical for there to be enough business to make sense for someone to try at least as a niche service. Lamborghinis aren't economical and plenty of folks by those

 

[miimura]

Outback Power made a follow-up announcement about their next generation hybrid inverter system called SkyBox. They are going for much more integration so that the installation cost is less. The PR is pretty light on details, but I believe they are integrating the solar charge controllers inside the "box" and they will be using their Ultra 300 technology which will allow 300VDC solar strings instead of the 145VDC max of their current products. They also tout the ability to install the solar and grid connections first and easily add storage later. Hopefully this will include a HVDC battery interface so that intact salvage EV packs can be connected. The battery interface is not mentioned anywhere in the links below.

Outback Power Inc. - SkyBox

SkyBox Press Release