7000 18650 cells in the model S tesla... smart or dumb?

[auraslip]

So the 2013 car of the year is the new Tesla sedan. Looks quite nice!

But the battery is huge and made from 18650 cells and the whole thing costs $20k. Why does this make sense? They seem to have poor energy density, and the fact that there is 7000 of them must raise the cost of producing the pack like crazy.

But then I've started thinking about how fickle cells are and how often they fail. RC lipo. Pouch cells in phones. 18650s in laptops. The li-mn cells that provide doc salvageable cells from failed mikita packs. Some here claim that high grade cells *won't* fail. I'm not so sure. Maybe Tesla isn't so sure either, and then how can they be when the technology in their batteries hasn't been around for more than 5 years? Maybe they listen to the battery companies claims of 10 years of use, and then plan on some of those cells failing anyways. Even if 5% of the cells fail in the tesla, range won't be effected but by a bit.

It starting to make sense to me! Maybe we're doing it all wrong here! Especially with our cheap headways, RC lipos, and Pings. Maybe we should ASSUME a 10% failure rate, and make it so that the cells are easy to replace, or small enough so that it won't matter. I'd love to see a setup made of a bunch of 18650 cells that require no soldering to swap out cells. Maybe just a screw or bolt to remove each cell. If one fails, then oh well. I lose 10% of my range until I can swap it out.

 

[neptronix]

Actually i believe those cells have the best energy density available in any commercial battery.

Tesla basically makes their batteries out of a giant parallel group of a dozen or so of these.. drops them into a machine, and out comes a giant parallel cell made up of a bunch of these.

I am sure that they cycle test them before assembling. You can't just stick a bunch of cells in a pack with a few % dud rate and warranty it for 5-10 years, you won't last as a company.

I still like big honkin' prismatics though.. less interconnections is good stuff.

 

[liveforphysics]

Actually i believe those cells have the best energy density available in any commercial battery.

Tesla basically makes their batteries out of a giant parallel group of a dozen or so of these.. drops them into a machine, and out comes a giant parallel cell made up of a bunch of these.

I am sure that they cycle test them before assembling. You can't just stick a bunch of cells in a pack with a few % dud rate and warranty it for 5-10 years, you won't last as a company.

I still like big honkin' prismatics though.. less interconnections is good stuff.

They cells are all cycled as a part of QC process at the panasonic factory as a part of the formation finishing process. Any that have even tiny fractional percentage deviations are culled to go for other uses, the ones with strong enough defects to get recycled seem to end up shrink-wrapped and called "ultrafire" or "trustfire" or a number of other funny names with wild capacity numbers written on the shrink.

IIRC, the Tesla pack has a PTC that can permanently disconnect in the event of an over-temp condition on any cell in the battery. This method obviously isn't possible to do when your pack is made of a single string of cells in series, but works fine when you've got something like 60-70 cells in parallel in each group, as a single cell failure in any 1 group only reduces pack capacity by something like 1.5%, and then you could also have a single cell failure in any/every other group and not have any additional pack capacity lost (because you can only discharge down to the capacity of the lowest cell in the group anyways, so 1 cell failure, or 1 cell failure in every series group is the same for pack capacity).

It seems like a lot of connections and complexity and potential failure points to me, but it seems they have perfected doing it this difficult way and it works for them. 18650's are still the cheapest $/Wh of anything, and when you're buying the right ones have excellent consistency and reliability, and Tesla's interconnect system seems to work well, and the thermal management seems to work well, so while it's not the way I like to do a battery, they do seem to have nailed the super giant 18650 cluster type battery very nicely.

 

[jonescg]

But the battery is huge and made from 18650 cells and the whole thing costs $20k.

For a battery that big, $20k is an awesome price. The closest I can get is 37 kWh for about $40k, which is still way too much.

 

[fizzit]

I have a feeling that the replacement pack price is less than they cost to make right now. They don't want to dissuade potential buyers by making them think that they'll have to pay half the car's price again in 10 years.

 

[mvly]

Anyone want to buy the $20K battery and disassemble them to sell here on ES? LOL seems like a deal. It should come out to be 3 bucks a piece after factoring tax, and cost of disassembly into pieces. This is much better than those $8-$10 on ebay.

Or maybe sell them in the parrallel package to save time on disasembly.

 

[Hillhater]

But the battery is huge and made from 18650 cells and the whole thing costs $20k. Why does this make sense?.....

There must be over 10,000 cells in the 85kWhr pack !
I assume you have read the paper about the thinking behind the tesla pack ?
Basically it sites safety as the primary reason for the choice of multiple 18650's vs other cell options.

Anyone want to buy the $20K battery and disassemble them to sell here on ES? LOL seems like a deal.

I suspect you would have to be a registered owner to buy one. !..and it would most likely be on an "Exchange" basis . :wink:
..but the "deal" is even better than you think....
http://www.mercurynews.com/business/ci_22090422/tesla-bumps-up-price-model-s-by-2

..The Model S offers three battery-pack options: 40 kWh, 60 kWh and 85 kWh battery options. Tesla also announced Thursday that it can now offer a battery replacement option. Customers can get a new battery anytime after the end of the year at a fixed price: $8,000 for the 40 kWh battery, $10,000 for the 60 kWh battery, and $12,000 for the 85 kWh battery.

 

[Hillhater]

BUT... can anyone explain this , on a pack that is supposed to have total protection ?...
http://jalopnik.com/5887265/tesla-motors-devastating-design-problem

....Tesla Motors' lineup of all-electric vehicles — its existing Roadster, almost certainly its impending Model S, and possibly its future Model X — apparently suffer from a severe limitation that can largely destroy the value of the vehicle. If the battery is ever totally discharged, the owner is left with what Tesla describes as a "brick": a completely immobile vehicle that cannot be started or even pushed down the street. The only known remedy is for the owner to pay Tesla approximately $40,000 to replace the entire battery. Unlike practically every other modern car problem, neither Tesla's warranty nor typical car insurance policies provide any protection from this major financial loss.....

.....A Tesla Roadster that is simply parked without being plugged in will eventually become a "brick". The parasitic load from the car's always-on subsystems continually drains the battery and if the battery's charge is ever totally depleted, it is essentially destroyed. Complete discharge can happen even when the car is plugged in if it isn't receiving sufficient current to charge, which can be caused by something as simple as using an extension cord. After battery death, the car is completely inoperable. At least in the case of the Tesla Roadster, it's not even possible to enable tow mode, meaning the wheels will not turn and the vehicle cannot be pushed nor transported to a repair facility by traditional means.......

.....The 340th Tesla Roadster produced went to a customer in Santa Barbara, California. In 2011, he took his Roadster out for a drive and then parked it in a temporary garage while his home was being renovated. Lacking a built-in Tesla charger or a convenient power outlet, he left the car unplugged. Six weeks later his car was dead. It took four men two hours to drag the 2,700-pound Roadster onto a flatbed truck so that it could be shipped to Tesla's Los Angeles area service center, all at the owner's expense. A service manager then informed him that "it's a brick" and that the battery would cost approximately $40,000 to replace. He was further told that this was a special "friends and family" price, strongly implying that Tesla generally charges more.

 

[flathill]

The biggest advantage, as Luke pointed to, is redundancy. With so many cells in parallel if you lose one it isn't a big deal. Each 18650 cell is triple "fused" for additional safety. Also note if you have one little bomb go off in a steel case the chances of it propagating is much less. This is the second big advantage.

Still, with tech advancing the need for active cooling/heating will be only be in extreme apps. Tesla is preparing for this eventuality...they are hiring battery engineers with prismatic experience that speak Korean :wink:

 

[neptronix]

What jalopnik conveeeeeeeeeeeniently left out, is that:

+ The Tesla car repeatedly warns you to charge it in this situation via your cell phone, email, or otherwise.
+ The owner's manual very explicitly states to NOT do that, in big red letters.
+ You leave any other battery completely drained with a BMS and constant thermal management left on and see what happens. Yep, it will drain.

Hey; they're Jalopnik; the internet equivalent of top gear. If it doesn't have a gas engine, it will be bashed and trashed via misinformation.

 

[SamTexas]

BUT... can anyone explain this , on a pack that is supposed to have total protection ?...
http://jalopnik.com/5887265/tesla-motors-devastating-design-problem

I certainly can't. The following is supposedly from Tesla (http://media.theunderstatement.com/021_roadster_manual_p5-2.pdf)

When fully charged, the Battery’s charge level can drop as much as 7% a day and 50% within the first week. When the Battery’s charge level falls below 50%, the rate of decline slows down to approximately 5% per week

7% a day when the vehicle is parked? That's unheard of. Typical lithium cells self discharge at about 5% A MONTH. Sounds like Tesla is doing something horribly wrong with their battery management system. What kind of parasitic load could this be?

 

[fizzit]

BUT... can anyone explain this , on a pack that is supposed to have total protection ?...
http://jalopnik.com/5887265/tesla-motors-devastating-design-problem

I certainly can't. The following is supposedly from Tesla (http://media.theunderstatement.com/021_roadster_manual_p5-2.pdf)

When fully charged, the Battery’s charge level can drop as much as 7% a day and 50% within the first week. When the Battery’s charge level falls below 50%, the rate of decline slows down to approximately 5% per week

7% a day when the vehicle is parked? That's unheard of. Typical lithium cells self discharge at about 5% A MONTH. Sounds like Tesla is doing something horribly wrong with their battery management system. What kind of parasitic load could this be?

I'm sure that this is a high estimate that they give in order to keep owners from ignoring the battery because they think that it'll take an extremely long time to discharge. But their BMS is very sophisticated and it probably does draw a decent amount of current watching over all those cells.

 

[Hillhater]

+ The Tesla car repeatedly warns you to charge it in this situation via your cell phone, email, or otherwise.
+ The owner's manual very explicitly states to NOT do that, in big red letters.
+ You leave any other battery completely drained with a BMS and constant thermal management left on and see what happens. Yep, it will drain.

The point is ,..there are inevitably situations where you either dont get the message, or even if you do, you cant respond. ( stuck overseas, or distant city, ..power system trips in the garage ?..etc etc )

Why wouldnt you have a "fatal voltage" isolator that completely isolates the CELLS if a critical voltage is reached.
What kind of BMS is it that actually destroys the battery !
..Or,..is this all negative propaganda ?
I get real annoyed when my $100 Pb starter battery fails after 3-4 years of abuse, and needs a jump start,.. so to have to face a $$$$'s bill for a Tesla pack on a new car would freak out most people.

At least lets hope the $10k replacement cost is real !

 

[grindz145]

Since they are a commodity, they have nearly the highest energy-density and at the lowest cell cost. Realistically, a high discharge rate is never necessary with the right sized pack. All of the water cooling and heating IMO will eventually be necessary in any automotive solution anyway. These guys are ahead of their time. The chemistry will change, but these guys know how to build a battery.

 

[auraslip]

Does anyone know if it's possible to get 18650s for a price near what tesla is selling them for? Right now the "decent" li-mn "panasonic" batteries that are rebranded as IMR go for around $10 each. Not cheap at all for anything buy ecigs and flashlights.

Actually, does anyone know a good source at all for 18650s? The only ones I can find are these "IMR" whatever IMR is. I hear it's just one guy that rebrands them to sell to the massive ecig and flashlight community.

At $10 a pop a 12s20ah pack would be something like $1200. Yeah....

I actually have a pretty good Idea for a battery holder for an 18650 pack. It could be made out of wood with shelves for each parallel group. Using the contact plates would maintain pressure with springs. Not springs on the batteries but on the contact bars. Each parallel group would have a strip of metal attached to the siding of the box with big thick springs. The side of the box you remove wouldn't need springs. When put the removable siding on the box it forces the cells down on the contact plates. This way, to remove a battery all you would have to do is remove the siding and pull the cell out. You could make the shelves out of a fireproof insulator to lower the fire risk. You could even do the same with each individual cell if you wanted too. Safe, easy, and theoretically pretty cheap. Would work really well with recycled cells since the labor would be taken care of once you have the battery box. Buy a thousand "bad" cells, throw out the junks, use the pack until enough go bad to where range sufffers, and then put new cells in. Hrm.

 

[Toorbough ULL-Zeveigh]

What kind of BMS is it that actually destroys the battery !

ULL uv zem!??

it's not even possible to enable tow mode, meaning the wheels will not turn and the vehicle cannot be pushed nor transported to a repair facility by traditional means

this one line is such over the top bull-crap it exposes the whole article for what it is.
i highly doubt any engineer would build a vehicle without a mechanical disengagement i.e neutral lever, but even if they were that stoopid;
tow-trucks all come equipped with 'traditional means' known as dollies or some such to haul away ICE vehicles that have lost their wheels completely in the extreme rare event of an automobile collision.
people need to think things thru a little critically.

The only ones I can find are these "IMR" whatever IMR is.

I=lithium Ion
M=Manganese, C=Cobalt, F=Ferro-phosphate - ICR, IFR
R=Round/cylindrical, P=Prismatic - ICP

 

[auraslip]

I=lithium Ion
M=Manganese, C=Cobalt, F=Ferro-phosphate - ICR, IFR
R=Round/cylindrical, P=Prismatic - ICP

sweet

 

[whatever]

my memory is a bit vague on this, but I remember the smaller the cell size the higher amp you could deliver,
its counter-intuitive I know, I cant remember the technical reasons on it, but that might be one reason for small cell sizes.

 

[agniusm]

Well, this is kinda right. Take A123 1100mah cell that delivers 3c constant and 20ah prismatic that delivers 10c constant. Now parallel 18 1100mah cells to get same capacity and that means that the group now can give away 18*3=54C constant and for 20ah thats 1080amps instead of 200 for prismatic 20ah 10C cell. This is how i imagine the situation.

 

[SamTexas]

my memory is a bit vague on this, but I remember the smaller the cell size the higher amp you could deliver,
its counter-intuitive I know, I cant remember the technical reasons on it, but that might be one reason for small cell sizes.

That's new to me. Thanks for mentioning it.
I would love to why, technically.

Just a wild guess on my part: Shorter travel distance from the anode to the cathode?

 

[Grey beard]

Allcell is building bike and scooter packs using the 18650 cells. Although the c rate is not high enough for a fast bike, the batteries are smaller, and lighter, than any other battery package that I've come across yet for the AHs.
http://www.allcelltech.com/images/act/ebike/allcell_48v_ebike.pdf

 

[fivari]

Well, this is kinda right. Take A123 1100mah cell that delivers 3c constant and 20ah prismatic that delivers 10c constant. Now parallel 18 1100mah cells to get same capacity and that means that the group now can give away 18*3=54C constant and for 20ah thats 1080amps instead of 200 for prismatic 20ah 10C cell. This is how i imagine the situation.

This is not correct. 18 cells with 3C rate parallelled are still 3C. The capacity increases 18-fold, but the rate is still 3C.
The big advantage of putting many small cells in parallel is the high redudancy. Furthermore it is probably quite easy to replace defect cells.

 

[liveforphysics]

What kind of BMS is it that actually destroys the battery !

I believe this would be known as, the vast majority of them. Most BMS's made are more aptly carrying the acronym Battery Murdering System.

There are great ones out there. They are just very few and far between.