18650 stationary storage

[Solarpower]

Hello!

I've been searching on the forum a while now, but I can't find much information on the topic. I guess this should be in the "alternative energy" section, but I try here first

I wonder if there is anybody here with experience from larger 18650 battery setups? I've seen people using Nissan Leaf batteries for storing solar power, and wanted to find out the possibility of doing this with 18650. And yes, I know this would be a lot of work.

I think jehugarcia's YouTube channel is inspirational, and his "Tesla style battery" is a good way to go. If he uses 2500mah cells he's got about 1,3kWh in one of those batteries. So to get, lets say 20 kWh, I would need 15 of those!




Does anyone have some input? What would you do? Is 18650 not the way to go with these kind of sizes?


YouTube link:
https://www.youtube.com/watch?v=JQa5gn-7D74&list=UUcMfCkN1juSa49DJFYltOTw

 

[Ypedal]

No reason not to.. other than cost.

Stand-by power is usually something you do with less expensive packs, or one's that don't hold up to EV use anymore but still hold lots of capacity.. :wink:

Otherwise, it's the same as any other battery type.. parallel as many as you want for the capacity you need, and rock on.

 

[dnmun]

you can tell jerry garcia that he does not have to use those big wide metal straps to connect the cells in parallel as he does in that picture.

for parallel you need almost no current carrying capacity, and where he connects in series he only uses a single connection from channel to the next channel on each end it appears.

it should be just opposite. he can make connections in series from can to can, case to anode, and anode to case, sequentially. since these are 2.8Ah then he can get away with a series connection that carries the 10A or so which is easy enuff to do with a thin piece of copper, 1mmx4mm cross section and about 18mm long so it can span from the center on the anode to the edge of the case on the bottom of the next can.

 

[Punx0r]

Considering the risks of damage when soldering to 18650 cells I would not dare use this technique for a large pack.

 

[TheBeastie]

Hello!


I think jehugarcia's YouTube channel is inspirational, and his "Tesla style battery" is a good way to go. If he uses 2500mah cells he's got about 1,3kWh in one of those batteries. So to get, lets say 20 kWh, I would need 15 of those!


Does anyone have some input? What would you do? Is 18650 not the way to go with these kind of sizes?


YouTube link:
https://www.youtube.com/watch?v=JQa5gn-7D74&list=UUcMfCkN1juSa49DJFYltOTw

Yeah that is kind of inspirational to see such a modest sized battery pack driving that whole bus around.
Its a nice looking pack he built but yeah the standard is to use spot welding for connecting up the 18650 batteries, and I am assuming hes just testing that pack as it lacks balance wires.
I am still in testing with my own self made 18650 pack, but its a relatively tiny pack just for my ebike.

If per cell fuse design is important and I think it most certainly should be for really large packs ES member circuit has a pretty cool looking 18650 tab design, I think I read somewhere he is even selling them at below cost. (took me a while to find the link so please look at it :x )
http://endless-sphere.com/forums/viewtopic.php?f=14&t=55870
http://endless-sphere.com/forums/viewtopic.php?f=31&t=62344

 

[chilledoutuk]

Guys you do realize that picture is of a 1s150p module he hasn't connected the copper buss bar yet in the picture but if you watch his tesla style module video he is making a 1s module for his samba which uses i think 30s.

 

[Solarpower]

No reason not to.. other than cost.

Stand-by power is usually something you do with less expensive packs, or one's that don't hold up to EV use anymore but still hold lots of capacity.. :wink:

Otherwise, it's the same as any other battery type.. parallel as many as you want for the capacity you need, and rock on.

Yeah I know it will be expensive. But if I'm not wrong, a large 18650 battery pack, just like the Tesla pack, would last a lot longer than old school storage batteries.
If the day comes that I'm investing in battery storage, it would be this or an old EV pack.

you can tell jerry garcia that he does not have to use those big wide metal straps to connect the cells in parallel as he does in that picture.

for parallel you need almost no current carrying capacity, and where he connects in series he only uses a single connection from channel to the next channel on each end it appears.

it should be just opposite. he can make connections in series from can to can, case to anode, and anode to case, sequentially. since these are 2.8Ah then he can get away with a series connection that carries the 10A or so which is easy enuff to do with a thin piece of copper, 1mmx4mm cross section and about 18mm long so it can span from the center on the anode to the edge of the case on the bottom of the next can.

Okay, then I learned something. I thought that the parallel wires were carrying quite a bit of current, since the voltage is only 3,7V?

Yeah that is kind of inspirational to see such a modest sized battery pack driving that whole bus around.
Its a nice looking pack he built but yeah the standard is to use spot welding for connecting up the 18650 batteries, and I am assuming hes just testing that pack as it lacks balance wires.
I am still in testing with my own self made 18650 pack, but its a relatively tiny pack just for my ebike.

If per cell fuse design is important and I think it most certainly should be for really large packs ES member circuit has a pretty cool looking 18650 tab design, I think I read somewhere he is even selling them at below cost. (took me a while to find the link so please look at it :x )
http://endless-sphere.com/forums/viewtopic.php?f=14&t=55870
http://endless-sphere.com/forums/viewtopic.php?f=31&t=62344

Thank you for finding those links. I was thinking something similar to that design! With a pack this large, I totally want to use single cell fuses.

I think the way Jehu Garcia does it simple and it also looks good. But as dnmun here mentioned, how are you supposed to connect the fuses? Don't you need some thicker wire "connecting" all the fuses?

Picture of the Tesla fuse:
View attachment 1


Jehu's fuse: Do I need that big copper strip? Or can that wire be smaller?



Appreciate all the answers guys!

 

[Hillhater]

I suspect 18650's would be a very expensive option, even when considering long term costs.
Their main advantage is energy and power density, IE you can get a lot of energy in a small , lightweight, package, and it can provide higher discharge rates than other chemistries safely....at a cost premium $$/kWhr
BUT.. Those are not benefits for stationary storage applications where you need capacity, reliability, and low overall cost.
It would be very unusual to need to minimise physical pack size for stationary storage applications....to the point where you would be prepared to pay a heavy premium for it.
I suggest you do go and read through the "Alternative Energy" forum now. ,! :wink:

 

[knighty]

you can tell jerry garcia that he does not have to use those big wide metal straps to connect the cells in parallel as he does in that picture.

for parallel you need almost no current carrying capacity, and where he connects in series he only uses a single connection from channel to the next channel on each end it appears.

it should be just opposite. he can make connections in series from can to can, case to anode, and anode to case, sequentially. since these are 2.8Ah then he can get away with a series connection that carries the 10A or so which is easy enuff to do with a thin piece of copper, 1mmx4mm cross section and about 18mm long so it can span from the center on the anode to the edge of the case on the bottom of the next can.

Okay, then I learned something. I thought that the parallel wires were carrying quite a bit of current, since the voltage is only 3,7V?

Dnmun was wrong, it's a 150p1s pack, and he pulls 650amps through it

there's 8 parallel bars there, so about 80amps/bar

when he ran it in the VW, he must have put it in series with his original battery

 

[Hillhater]

Dnmun was wrong, it's a 150p1s pack, and he pulls 650amps through it

there's 8 parallel bars there, so about 80amps/bar

when he ran it in the VW, he must have put it in series with his original battery

Yes, that was just one of many 150p modules he planned to assemble using 6 Amp fuse wire links from each cell to those copper "collector " bars ...before he acquired some real Tesla modules.
Some folk should look close or watch the videos before criticizing.

 

[Jonathan in Hiram]

I did two individually fused 18650 packs 6s12p, I stripped cells out of old laptop packs I had a source for and tested to find good ones, my inspirations were DrkAngel here on ES and Jehu Garcia on Youtube. I used a very thin copper wire as a fusible link to connect each individual cell. The fusing works beautifully to protect the pack from cells that go bad, I made up a 1s12p pack and deliberately shorted one of the cells and the fuse wire blew almost instantly. On the other hand I've never had a fuse blow during ordinary operation.

For a stationary pack if you can get laptop cells cheaply they would work even better than they do in my ebike.. I had to go through a lot of cells that were otherwise reasonable for stationary use to get enough with a low internal resistance suitable for the high amp load an ebike pack sees.

Don't have any really good shots of the packs I get too focused on what I'm doing to get many pictures during the assembly process of things but here is one I took.

 

[Lory]

After two years collecting laptop batteries i got good usable 8 kwh capacity i soldered to modules. For sure cell fused.

Now, i´m wayting for budget go get the charge and reload to grid automation.

[youtube]bBzPAQejyTA[/youtube]

 

[circuit]

What BMS are you going to use?
For sure would be scary without one.

 

[Lory]

At the moment i balance when charging with revolectrix cellpro. I build 30 modules with 42cells each. The current setup is 6s210p (5 modules in paralell). Later i´ll use the modules with sunny island as 15s 84p (2 modules in paralell) and a BMS from REC that can communicate with the SMA Sunny Island.

BMS
http://www.rec-bms.com/BMS9R.html

 

[Syonyk]

If you're buying large numbers of 18650s, you shouldn't be paying much more than $300-$400/kWh - which is expensive compared to nameplate capacity of lead acid storage, but remember that you can actually use the full range of the lithium cells regularly, vs lead acid where you shouldn't use more than about 20% - so my 12kWh pack is, realistically, about a 3kWh pack in daily cycle use. Or less. So 3-4kWh usable capacity actually comes out comparable with lead acid in cost, and should be a good bit longer lived if you're using it daily.

One thing to consider is thermal management. Cold lithium batteries are not happy, hot lithium batteries are not happy, and lead acid mostly just doesn't care as long as you don't freeze it when discharged.

It's certainly something I intend to play with at some point here, since I've got a perfectly good lead acid pack and build plenty of lithium batteries in my spare time.

 

[Lory]

i don't buy new 18650. i recycle old laptop batteries. so my price per kWh is 50 Euro. i use 75-80% of their capacity (4,1-3,5v) they never get hot because i charge and discharge them with less then 1 amp per cell. in my celler the temperature is between 16 - 20 celsius. Lead acid where never an option. loos to many energy when charge and discharge, need liquid service, need fresh air system, low cycle count, low capacity usable 50% DOD vs. 80% LiIon.

 

[flippy]

i wawtched that VW bus project for a while and he does not any balancing with those packs wich is stupid to say the least.
that fuse wire stuff is nice but it will not blow when a cell is dying so it still pulls down your pack and with only recycled batteries with different IR's that is going to be an issue.

he makes nice video's but using crappy cells with the "tesla-style" packs is not a good combo. the tesla style needs a good balanced pack in order to work properly. not habing any balacing just tops off the stupidity. no way i am sitting in the back of that van with a few thousand dodgy cells under my feet.

 

[Lory]

I have a channel abo from the electric samba jehugarcia too. He dont use active BMS but he check the cells from time to time manual. He charge not to 4.2 cell voltage go have a buffer for drifting modules. Jehu has driven a few hundred miles without problems caused drift. So for his usecase it works without BMS.

I always will have full control and information about the voltage. So i´ll go with BMS. When i have a bad cell with very high self discharge maybe 1-2 amps the cell will discharge the hole pack. but this will need 40-70 hours. i charge every day with more than 25 amps. when i start charging. the module with 84 cells with 25 amp and one cell begin to drift the other 83 parallel cells will get the voltage up and if the load increase 2,5 amp the cell fuse will shoot out the bad cell.

If the BMS detect module voltage drift i´ll replace the module with a spare and start a revision job to the cells of the module.