TMC is an independent, primarily volunteer organization that relies on ad revenue to cover its operating costs. Please consider whitelisting TMC on your ad blocker or making a Paypal contribution here: paypal.me/SupportTMC

18650 Batteries

Discussion in 'Tesla Motors' started by NotTarts, Dec 15, 2012.

  1. NotTarts

    NotTarts Member

    Joined:
    Aug 20, 2012
    Messages:
    72
    As others have mentioned, the Roadster uses LCO (Lithium Cobalt) batteries whilst the Model S uses NMC (Lithium Nickel Manganese Cobalt). NMC batteries have improved life span, power output, safety, and temperature performance compared to what's used in the Roadster. You can read about it here:

    Types of Lithium-ion Batteries – Battery University

    LCO
    li_2(1).jpg

    NMC
    li_6(1).jpg

    Also remember that Tesla essentially purchased commodity cells intended for laptops when building the Roadster, so it's pretty incredible that they lasted even 250,000km. With the Model S they've partnered with Panasonic to produce a cell suited for automotive use.
     
  2. Zzzz...

    Zzzz... Member

    Joined:
    Aug 19, 2012
    Messages:
    867
    Location:
    East Asia/Canada
    That is not true. Model S uses NCA, nickel cobalt aluminum(LiNiCoAlO2) cathode chemistry from Panasonic.

    It is a specialized automotive optimized cells/formulation of NCA, and not the one Panasonic produce for retail market.
     
  3. NotTarts

    NotTarts Member

    Joined:
    Aug 20, 2012
    Messages:
    72
    Thanks, I just assumed that they used NMC as I couldn't find anything about the batteries apart from 'Nickel-based'.

    Interestingly NCA is the same chemistry used by the Smart Fortwo EV battery which is apparently rated at 400,000km.
     
  4. Dan5

    Dan5 Member

    Joined:
    Oct 13, 2011
    Messages:
    422
    Location:
    Delran, NJ
    I agree with ZZZZ... It's not NMC it's NCA

    Nissan is romoured to be using the NMC in their next gen EV. These are all more stable than the plain cobalt flavors though.

    A good way to think of it is that Nissan isn't liquid cooling and is considering using a combo. Tesla is liquid cooling AND using a similar, but not the same chemistry as Nissan. I would say Tesla is treating the battery very gently, and would expect Nissans to fail or degrade well before the Tesla battery
     
  5. Zzzz...

    Zzzz... Member

    Joined:
    Aug 19, 2012
    Messages:
    867
    Location:
    East Asia/Canada
    And one important point to add, performance of Panasonic NCA cells optimized for EV use was published "220th ECS Meeting, © 2011 The Electrochemical Society Development of High power and Long life Lithium secondary batteries"
    With pdf publicly available. You can see there that cells could handle 3000+ cycles without substantial degradation. While specs of the cells of the similar chemistry that Panasonic produce for general market(NCR18650A or something) perform MUCH worse even after just 500 cycles....

    You can find link to pdf with the study here: Hansjörg von Gemmingen Roadster pack out of juice? - Page 3
     
  6. vfx

    vfx Well-Known Member

    Joined:
    Aug 18, 2006
    Messages:
    14,792
    Location:
    CA CA
    Given the certain amount of planned obsolescence or acceptable battery lifetime in relation to consumer replacement, do you think that the batteries made for cars have been given much longer life because Tesla is buying in such large quantities?

    And would it be worth putting the cars cells in flashlights and other consumer items that accept that form factor?
     
  7. Zzzz...

    Zzzz... Member

    Joined:
    Aug 19, 2012
    Messages:
    867
    Location:
    East Asia/Canada
    No! They, the cells were given much longer life because customer wanted it. Customer == Tesla Motors.

    Rest of the crowd.... How to say it - they definitely not evil. But producers of Smartphones or Notebooks DO NOT CARE if battery would last 3 years or 15. Furthermore, big players(top 5 control what? 60% of worldwide market or something?) would prefer battery that would last 2 to 3 years before dieing.... Not cuz they evil or conspirators! It is just customers wont pay extra $1 for smartphone that would last for 15 years instead of three or four. And provided the same price, biggies would pick battery that would last 2 years and a half or something, this would mean incentive for a client to buy new notebook in few years. So why battery producers disadvantage themselves?

    Yeah, that would be awesome! But most of the market do not "buy" 16850 cells. Consumers buy notebooks, flashlights etc. Not cells. And that creates market for "cells for laptops" which is dozens/hundreds of times bigger then consumer one. Have you replaced your battery or your gadget itself when you noticed past warranty problem(s) last time?

    And now lets talk about price. Take this with a grain of salt but here is my IMO. From my understanding(limited). I personally follow A123. Just after they announced possibilities of bankruptcy, in a week or two they reported "our science team archived record energy density of the cells!" That did not make me laugh, but I indeed smiled, at least in my mind and sarcastically when I was reading the "news". From my understanding strong science team that specializes in secondary batteries have no problem producing Li-ion cells that have 10 000 cycles life. Or 600Wh/Kg cells. First one was shown by A123 itself in 20x0, where x is between 3 and 7. And that might have been 15 000 cycles. I remember counting myself - how many year such battery would last... More the 25! I was impressed, and I bet many were at the time. But reality is, it is not the cell that you can create in the lab that counts... But only cells that could be economically produced.

    So what about 600Wh++/kg? This one was shown many hundreds of times... Practically. But who cares that after first 5 cycles capacity drops to 300Wh/kg and after 30 cycles drops to less then a 150Wh/kg? Definitely not a crowd who reads about breakthroughs! They are impressed by top numbers.

    But bottom line... Cell manufacturers DO have a space to maneuver as to regards of some parameters of cell production. And if Panasonic got NCA manufacturing process - they do have some freedom as to what resulting cells would be optimized for: price, power density, specific energy, longevity... List could go on and on. And some parameters contradict each other - like cell that have low IR would probably get less stable electrolyte. And that mean shorter lifespan of the cell itself.

    Clients - pretty much any ones who order cells directly from Panasonic do have a say in by what parameters they want cells to be optimized... And by how much. Tesla, Dell, GM, DEWALT etc and so on do have battery scientists on payroll and they know/could reliably estimate how much this or that one chemistry adjustment would cost to industry player like Panasonic or LG etc...

    PS. This dose not mean that Tesla Motors picked up most long lasting NCA chemistry/process offered by Panasonic. But I don't have any doubt that Tesla's Panasonic cells do not match NCR18650A aging profile... So while Panasonic can produce EV "optimized" cells that would last 3000 cycles with 10% degradation, and could produce same chemistry notebook use cells that last 500 cycles with 30% degradation.... It was and IS up to Tesla Motors to decide which cells they really want.
     
  8. stopcrazypp

    stopcrazypp Well-Known Member

    Joined:
    Dec 8, 2007
    Messages:
    7,039
    #8 stopcrazypp, Dec 16, 2012
    Last edited: Dec 16, 2012
    The cells Panasonic (Matsushita) produces for the "retail market" uses NCA (under the new marketing name of NNP, it was previously marketed as NCA). And Tesla uses a "retail market" cell, specifically, the NCR18650A (the only 3100mAh 18650 battery that Panasonic makes). Tesla is not using some special formulation by Panasonic (I have not seen any evidence presented that they are, but there's plenty of evidence pointing to Tesla using specifically the NCR18650A, they even had a press release that said as much announcing Tesla was using Panasonic 3.1Ah batteries made at 300 mil. units per year AKA the NCR18650A).

    http://powerelectronics.com/mag/601PET06.pdf
    http://www.engadget.com/2010/04/23/panasonics-3-1ah-batteries-to-be-used-in-the-tesla-model-s-hav/

    As for the original article. I think the owner should at least let Tesla look at the pack to see if it's unbalanced or have a couple of damaged cells. The range dropping off a cliff from about 200km to 100km in such a short period suggests there might be a balance issue or only a couple of failed cells. The fact the guy insists on getting a free replacement pack or a free Model S without even allowing the Tesla to diagnose the problem on the pack tells me it's the customer being unreasonable. And the pack is likely out of warranty so it's not like Tesla is even obligated to respond.

    Practically all 18650 cycle tests cut off at 500 full cycles, so the fact the pack survived to 700+ is already pretty good. It's over 150k miles which is probably past the miles most people own their cars for.
     
  9. tdiggity

    tdiggity Member

    Joined:
    Oct 3, 2012
    Messages:
    721
    Location:
    Bay Area, CA
    Quick question: what are cycle tests and why do they cut off at 500?
     
  10. JRP3

    JRP3 Hyperactive Member

    Joined:
    Aug 20, 2007
    Messages:
    10,103
    Location:
    Central New York
    Usually testing ends when capacity drops to between 80-70% of new. I think 80% is standard but some test to 70%.
     
  11. Zzzz...

    Zzzz... Member

    Joined:
    Aug 19, 2012
    Messages:
    867
    Location:
    East Asia/Canada
    Yes.

    Whatever 3.1A Panasonic cells Tesla is using, they do have different formulation/production specifics, so they are optimized for automotive use. There is no point for Tesla not to use cells that have longer calendar life/cycle life compared to NCR18650A optimized for notebook market.

    Could you provide link for such claim?

    Dosen't make ANY sense for Tesla not to use readily awailable technology from Panasonic.

    Opps, you are not serious, are you? Those links prove exactly my point, Tesla is using 3.1A NCA cells from Panasonic. And since Panasonic do have NCA chemistry that is optimized for EV(see link to pdf in my posts above), why would Tesla order NCA cells with much shorter calendar and cycle lifespan?

    Please provide links that make you thinks so.
     
  12. stopcrazypp

    stopcrazypp Well-Known Member

    Joined:
    Dec 8, 2007
    Messages:
    7,039
    I assume you mean this link:
    http://www.teslamotorsclub.com/attachment.php?attachmentid=10719&d=1350500751

    That link compares older (LiCoO2, Li(Ni,Mn,Co)O2) chemistry with the NCA (Li(Ni,Co,Al)O2) chemistry used in the 3.1Ah NCR18650A (and previous 2.9Ah NCR18640). There is no third "EV optimized" chemistry mentioned at all in that link!

    In fact, the link talks specifically about the 140 cell module is mentioned here in this press release, which actually uses the older 2.9Ah NCR18650 (which uses the same chemistry). The 3100mah NCR18650 wasn't even in production yet (production started in April 2010, see the engadget link I gave before announcing production). The 140 cell module has 58Ah of capacity using 7 rows of 20 cells in parallel; 58Ah/20 = 2.9Ah per cell.
    http://panasonic.co.jp/corp/news/official.data/data.dir/en091001-3/en091001-3.html
     
  13. Zzzz...

    Zzzz... Member

    Joined:
    Aug 19, 2012
    Messages:
    867
    Location:
    East Asia/Canada
    #13 Zzzz..., Dec 23, 2012
    Last edited: Dec 23, 2012
    NCR18650A have what? More then 20%+ capacity drop after 300 cycles. Info from Panasonic dot com:
    http://www.panasonic.com/industrial/includes/pdf/ACA4000CE254-NCR18650A.pdf

    Panasonic's EV optimized NCA chemistry process have what? Less then 10% capacity drop after 3000!!! cycles. Info from paper published by Panasonic Corporation.

    Excuse my graphic design skill, but if you do not see the obvious, here I highlighted main points using red arrows:

    pan.jpg

    So what Tesla have chosen? Slightly cheaper(few percent) NCA that barely lasts 300 cycles or EV optimized NCA that lasts 3000 cycles?

    Besides... Do you realize that same li-ion chemistry could be optimized literately in thousands ways per customer request? For example Tesla Roadster was using mainstream LiCoO2 li-ion chemistry. The one that fails in notebooks after 2 or 3 years. And after couple years LiCoO2 battery in your laptop would be lucky to provide 70% of capacity it had when notebook was new.

    Roadsters run on the roads for 4+ years already. And here we come on topic: do you see mass failure of Roadsters batteries? This topic is an exception, and looks like person behind the claim, Hansjörg von Gemmingen have his agenda. But in fact most old Roadsters batteries still hold 94%-97% of initial capacity...

    Do you even realize that TM have got major purchase power when dealing with battery producers? And that most top players in the industry probably have dedicated teams trying to convince TM to deal with them? I could collaborate with numbers if THAT is not clear.

    Do you even realize that some Model S rolling of the production line most likely have got cells that were not produced by Panasonic? Tesla were reinforcing message that only some!(read Panasonic) relationships with battery cell producers were publicized. And with scale of Tesla orders it do not make any sense to be stuck with just one supplier of key component. TM perfectly in position to have three or four cells suppliers. But most likely they are dealing with an issue "Apple way", having two suppliers for every major component for particular model...
     
  14. stopcrazypp

    stopcrazypp Well-Known Member

    Joined:
    Dec 8, 2007
    Messages:
    7,039
    #14 stopcrazypp, Dec 23, 2012
    Last edited: Dec 23, 2012
    I guess you don't get my point. They are the same chemistry (Li(Ni,Co,Al)O2) tested under different conditions! The first graph was tested cycling from 2.5V to 4.2V (0-100% SOC). Cycling conditions were 0.885A charge and 2.95A (1It) discharge.

    Panasonic also has a 500 cycle graph (link is dead though if you go to the older thread), but I have it here:
    attachment.php?attachmentid=1479&d=1347276205.jpg
    http://www.teslamotorsclub.com/showthread.php/5220-Panasonic-cells-for-Model-S

    edit: found a working link:
    http://industrial.panasonic.com/www-data/pdf2/ACI4000/ACI4000CE25.pdf

    The pdf you linked (2nd graph in your post) tested the cells from 3.6V to 4.05V (~60-90%SOC). Cycling conditions were 1It and 2It (no mention of charge conditions). This is much milder testing because it avoids the high and low SOCs that a standard 18650 test (like the one in the first graph) would have to go through. Bog standard 18650 cells using the older LiCoO2 chemistry have lasted 100k+ cycles by using small 20%DOD even though in standard 100%DOD their testing cuts off at the standard 300 or 500 cycles with ~70-80% capacity left (see Table 1 on page one in link below):
    http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA515369

    DOD for your second graph is about 30% so you have to at least multiply the results by that much to get the equivalent "full" cycles (that would mean 3000*30% = ~900 cycles equivalent).

    There is no indication that your PDF was talking about a third different chemistry. It mentions the same Li(Ni,Co,Al)O2 that the NCR18650 and NCR18650A currently use and not a different one (which would have a different chemical structure).
     
  15. JRP3

    JRP3 Hyperactive Member

    Joined:
    Aug 20, 2007
    Messages:
    10,103
    Location:
    Central New York
    Stopcrazypp is correct.
     
  16. Zzzz...

    Zzzz... Member

    Joined:
    Aug 19, 2012
    Messages:
    867
    Location:
    East Asia/Canada
    #16 Zzzz..., Dec 23, 2012
    Last edited: Dec 23, 2012
    I guess that is you who do not get my point.


    I repeat my questions once again. Do you understand that, for example same chemistry (Li(Ni,Co,Al)O2) also known as NCA, also known as Lithium Nickel Cobalt Aluminum Oxide, also known as LiNiCoAlO2 etc can have thousands different formulations?

    And that such chemistry could be optimized on the fly for clients needs? And that such practice quite common in industry? As long as cells producer got above mentioned NCA process he could tweak it left and right for client needs.

    And my point was that NCA based Panasonic NCR18650A, NCA based whatever previous generation NCA 2.9A Panasonic cell, the Panasonic once again NCA that mentioned in this pdf and NCA cells that TM actually buy from Panasonic are four different formulations of NCA. While being "the same chemistry".

    Do you understand that TM is aiming at purchasing around 3% of all cells produced for electronics worldwide in 2013? And around 6% of cells produced worldwide in 2014? Provided estimated cells production for computers, communication, and consumer electronics is around 4.6 billion units in 2012 and 20,000 Model S produced in 2013 having 7000+ cells each. And 40 000 S and X in 2014? I think actual numbers would be higher(definitely hope so, given high demand and already archived production rate). And taking into account higher specific energy of each TM cell plus the fact that most cells are produced for cheap cellphones TM could claim 10%+ worldwide market share of 18650 cells in 2014...

    And with such purchase power TM is in position to dictate how every step of cells production is done by Panasonic...

    Are you aware that it make 0 business sense for TM to stuck with one supplier of key component(NCA li-ion cells) of Model S and most likely TM already have another supplier of NCA cells? Taking into account TM's purchase power and how crucial cells are, plus the fact that TM claimed that "only some" relationships with cells suppliers(read with Panasonic) were publicized...


    Could you provide the link or any evidence supporting your point?
     
  17. JRP3

    JRP3 Hyperactive Member

    Joined:
    Aug 20, 2007
    Messages:
    10,103
    Location:
    Central New York
    Stopcrazypp already provided the evidence that the cycling differences you are seeing come from different test protocols. Shallow discharge cycles vastly extend the life of cells.
     
  18. Zzzz...

    Zzzz... Member

    Joined:
    Aug 19, 2012
    Messages:
    867
    Location:
    East Asia/Canada
    So what? That is in no way support his point that NCRxxxxA is the exactly the same cells that TM ordering from Panasonic...

    - - - Updated - - -

    And I explained why...
     
  19. JRP3

    JRP3 Hyperactive Member

    Joined:
    Aug 20, 2007
    Messages:
    10,103
    Location:
    Central New York
    You have speculated but offered no definitive proof.
     
  20. Zzzz...

    Zzzz... Member

    Joined:
    Aug 19, 2012
    Messages:
    867
    Location:
    East Asia/Canada

Share This Page