Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Why do Li-ion Batteries die? And how to improve the situation?

This site may earn commission on affiliate links.

djp

Model 3 Performance
Aug 28, 2011
1,125
77
Toronto, Canada
Thanks to Saftwerk for finding this excellent presentation on Li-Ion battery life. It's long, but packed with information and is the closest we'll come to being inside Tesla's battery research lab. In fact, one of the grad students involved in this research project (Aaron Smith) is now running the Battery Lifetime Group at Tesla. It's a fascinating look at the science and technology behind the cells.

Why do Li-ion Batteries die ? and how to improve the situation? - YouTube

In general the research confirms many of the rules of thumb that have been discussed on the forum. Avoid spending long periods of time at high state of charge and keep the battery cool, and your pack should have a very long life.
 
Electrolyte composition may be different in the "A" and "B" packs of the Model S

Posted this on the original thread djp is referring to. Thought the significance to Model S owners is high. So here it is:

Amazing! This explains why model S owners are seeing minimal degradation in their batteries (assuming their packs are balanced). This research group tested Tesla's battery chemistry with different electrolytes and found different degradations. The types of electrolytes and electrolyte combinations make a huge difference in cycle life even in the same chemistry. Wow!

It makes me wonder if the "A" and "B" battery packs only differ by their electrolyte composition. One of the members of the group which made these discoveries of combining electrolytes to increase cycle life started working for Tesla right around the time the Tesla started changing the packs in the Model S. Interesting.
 
Excellent talk!

The only thing that he didn't come around to fully that I would have liked, is the effect of storage (he skipped over it and said it isn't part of his talk).

He obviously made a point that fast charge/discharge cycles isn't reliable measures of testing because it shows very little degradation. And he also made a point that storing at a high SOC is bad. H

However, is it actually better to literally charge faster, and thus leave it a low SOC for longer? i.e. Will you get better life if you charge with an HPWC (provided you do scheduled charging and drive as soon as it's done rather than leaving it at a high SOC), or will you get better life by charging with a slower charger?


Either way, it seems like whatever we can personally do will be pretty marginal. Tesla is doing all the heavy lifting.

There is such a huge difference though between the capacity loss over the first few cycles compared to the rest of the lifetime of the battery. I wonder if for PR sake if it would be better for Tesla to just hide the extra capacity before the initial drop-off. Then it will seem like their batteries last forever! Though they'll have to weigh that against marketing the car with a lower range, so it's a tough call.
 
There is such a huge difference though between the capacity loss over the first few cycles compared to the rest of the lifetime of the battery. I wonder if for PR sake if it would be better for Tesla to just hide the extra capacity before the initial drop-off. Then it will seem like their batteries last forever! Though they'll have to weigh that against marketing the car with a lower range, so it's a tough call.
The "hiding" is something we will probably see in the future. We as end-users will simply never be able to charge to 100%, 85% ~ 95% will be 100% for us as a end user.

This will prevent faster degradation, but also hide any degradation for a long time.

Obviously, the measurement instrument here is Amp Hours (Ah) and not %.
 
What was enlightening for me was that the charge/discharge cycle itself is actually very benign and doesn't on its own lead to degradation. Cycle count, depth of discharge, and mechanical stress from changes in cell volume are not the root cause of the problem.

The degradation comes from being at high state of charge, especially in high temperatures. At high voltages the cathode and anode are both highly reactive, and react with materials in the electrolyte to form a layer of "gunk" which blocks the flow of lithium ions. The longer the battery spends at high SOC, the more time for these parasitic reactions to destroy the battery.

The standard "cycle count" graphs from battery manufacturers are misleading since these are done as rapid cycles over a short period of time (30 days) which hides the effect of time at high SOC. This is what got Nissan in trouble with their poor choice of battery chemistry in the Leaf.

To see the effects properly you need to do extremely accurate energy and heat flow analysis to measure the parasitic reactions, which is what is presented in this lecture (and what Tesla hired his grad student to do).

The most effective way of extending battery life is by including additives in the electrolyte which slow down the parasitic reactions. He also found the reactions happen fastest at the top end of the charge cycle, especially above 4.08V. Interestingly, this corresponds to about 90% SOC on the Roadster, which is what Tesla chose as their Standard charge. He mentioned the Volt charges to 80%, which he called "decent".

Keeping the battery at 4.1V is better than 4.2V and 4.0V is better than 4.1V. He gave an example of a cell from 1999 which was kept at 20% SOC, and when tested in 2013 was "like new". (Storage mode on the Roadster is also 20% SOC)

I was impressed with the rapid progress being made on electrolyte additives, and the research confirmed my thoughts that it's best to keep the battery below 90% SOC.
 
Last edited:
Although very technical, and difficult to follow at times, this was an extremely informative lecture. I will be adjusting my daily charge from 90% to 70-80%, and increasing my charge current from 20 amps to 40amps as a result of viewing this. Thanks for posting. This should be a sticky!

It was also nice to see that Tesla's degradation line was nice and flat and low relative to the other manufacturers.
 
Although very technical, and difficult to follow at times, this was an extremely informative lecture. I will be adjusting my daily charge from 90% to 70-80%, and increasing my charge current from 20 amps to 40amps as a result of viewing this. Thanks for posting. This should be a sticky!

It was also nice to see that Tesla's degradation line was nice and flat and low relative to the other manufacturers.

I am really curious to know what a 90% charge equals in terms of cell voltage (i.e. 4.1 volts). Also, I am going to not bother "slow charging" to balance the cells when range charging, clearly just kills the cells. Might as well just go as fast as it can, and let it balance at the top of the charge. Yes I know range charging is bad, but is required for balancing as far as I can tell.

Also what was really illustrated by this lecture is that Heat is really bad too. High heat and high SOC really really bad.

Also, it looks like depth of charge is not as destructive as I once thought.
 
sent the Professor the following email, I'll post a reply if I get one:

"Professor Dahn, I just watched your very informative video and had a few questions. I would very much appreciate it if you had the time to answer them! First off, thanks for making your findings available, it really helped my understanding of li-ion batteries.

Since electricity appears to have become the defacto currency for renewable energy and I have a great interest in and now investment in renewable technologies (own a Nissan Leaf and Tesla Model S and a 9.8 kW PV system), I have been doing my best to learn whatever I can about energy storage and how batteries work, in particular, the way end user habits effect overall life.

I am wondering if your research has revealed what the best way to extend a batterie's life is. I have concluded, as you have noted, that high ambient temperatures degrade these batteries the most. Living in Seattle, we are blessed with a moderate climate, so I feel pretty fortunate in that regard.

Debate continues on whether charging habits will turn out to have much effect on battery life. I have started leaving all of my li-ion consumer products at about 50% charge when they are not in use. Tesla makes this very easy with a battery slider that allows the consumer to choose between 50% and 100% end charging and even has this feature integrated into the phone app to make it very easy to adjust upwards on the fly as long as the car is plugged in. I have installed an 80A (20 kWh) charger so that I can quickly add charge to the Tesla S, allowing me to minimize the inconveniences associated with leaving the pack at a low state of charge and thus increasing the amount of time the battery stays at a mid/low SOC. I tend to do mid pack cycling, discharging generally between 30 and 70% when convenient, while aiming to have the car sit for the longer stretches like overnight, at 50%. I'm not religious about this, just tend to aim in this direction generally.

In your opinion, do you believe it is worth the effort to keep li-ion batteries at 50% or thereabouts for the bulk of their resting time to extend their long term capacity and if so, do you have any idea how much of a difference this is likely to make over say a 8-15 year time period.

Also, given that Tesla limits power to the motor during extremes, like high and low charge and high and low temperatures, is there really any reason to avoid running the battery low, assuming Tesla doesn't let you discharge the battery all the way anyway and limits discharge rate as the charge level drops. I assume running the battery to zero (of what the manufacturer allows) does little to the overall life as long as it gets recharged soon after. I assume that low and high states of charge are more an issue if exposure is prolonged, is that correct?

Any opinion you may have on this subject will be much appreciated in helping me understand what the limits of this technology are and how to best treat the batteries in my cars and consumer products.

Kind Regards,
George"
 
Wow, just wow. Best 1.2h I have spent in recent time. The research methodology is really excellent and revolutionary. And while I had understood that Teslas BMS and chemistry choice is good in comparison to competition, the real tested results and his comments on the preferred tech at various conditions show that Tesla is lightyears ahead of others. The fact that you can have 20x difference in cycle life just from chemistry and additives choice in electrolytes is huge. And from what I understood in the Q&A section the LiTiO and LiCoO should have an order of magnitude better cycle capacity and the latter is in fact used by Tesla.

again, just wow...
 
Here is the response:

"Thanks for this e-mail. I have placed responses within your text using CAPS (not yelling) but so you can find my responses


Jeff Dahn, FRSC
Professor of Physics and Atmospheric Science
NSERC/3M Canada Industrial Research Chair
Canada Research Chair"

"Since electricity appears to have become the defacto currency for renewable energy and I have a great interest in and now investment in renewable technologies (own a Nissan Leaf and Tesla Model S and a 9.8 kW PV system), I have been doing my best to learn whatever I can about energy storage and how batteries work, in particular, the way end user habits effect overall life.

THIS IS AWESOME.

I am wondering if your research has revealed what the best way to extend a batteries life is. I have concluded, as you have noted, that high ambient temperatures degrade these batteries the most. Living in Seattle, we are blessed with a moderate climate so I feel pretty fortunate in that regard.

ABSOLUTELY. AVOID HIGH T WHENEVER POSSIBLE.

Debate continues on whether charging habits will turn out to have much effect on battery life. I have started leaving all of my li-ion consumer products at about 50% charge when they are not in use. Tesla makes this very easy with a battery slider that allows the consumer to choose between 50% and 100% end charging and even has this feature integrated into the phone app to make it very easy to adjust upwards on the fly as long as the car is plugged in. I have installed an 80A (20 kWh) charger so that I can quickly add charge to the Tesla S, allowing me to minimize the inconveniences associated with leaving the pack at a low state of charge and thus increasing the amount of time the battery stays at a mid/low SOC. I tend to do mid pack cycling, discharging generally between 30 and 70% when convenient, while aiming to have the car sit for the longer stretches like overnight, at 50%. I'm not religious about this, just tend to aim in this direction generally.

I THINK THAT IF YOU ARE KEEPING THE CELLS BELOW 4.0V (NO WAY FOR YOU TO TELL THAT, HOWEVER) THE BATTERY LIFE WILL BE VERY IMPRESSIVE. WE ARE TESTING CELLS BUILT IN 2002 THAT HAVE 2002 TECHNOLOGY (LIFETIME IS BETTER NOW) THAT STILL HAVE 75% OF THEIR INITIAL CAPACITY (CYLCED AT 37c THE WHOLE TIME). THESE CELLS WERE CHARGED ONLY TO 4.075V. MODERN CELLS LIKE THOSE IN TESLA CHARGED TO 4.0V SHOULD LAST A FEW DECADES, I SUSPECT, WITHOUT ANY ISSUE (SO KEEP YOUR CAR FROM RUSTING!). WHERE IS 4.0V RELATIVE TO STATE OF CHARGE? MAYBE 75%.

ONE OTHER THING I WOULD RECOMMEND IS TO AVOID HIGH RATE CHARGING AT TEMPERATURE BELOW 0C. ESPECIALLY WHEN THE CELLS ARE ABOVE 75% SOC. TESLA ELECTRONICS MAY PREVENT THIS

In your opinion, do you believe it is worth the effort to keep li-ion batteries at 50% or thereabouts for the bulk of their resting time to extend their long term capacity and if so, do you have any idea how much of a difference this is likely to make over say a 8-15 year time period.

KEEPING BELOW 4.0V MAY DOUBLE OR TRIPLE LIFE TIME COMPARED TO A FULL CHARGE EVERY CYCLE, I SUSPECT.

Also, given that Tesla limits power to the motor during extremes, like high and low charge and high and low temperatures, is there really any reason to avoid running the battery low, assuming Tesla doesn't let you discharge the battery all the way anyway and limits discharge rate as the charge level drops. I assume running the battery to zero (of what the manufacturer allows) does little to the overall life as long as it gets recharged soon after. I assume that low and high states of charge are more an issue if exposure is prolonged, is that correct?

I THINK THAT REALLY DEEP DISCHARGE SHOULD BE AVOIDED AS THEN THE GRAPHITE EMPTIES OF LI AND THE CELL POTENTIAL RISES TO THE POINT WEHRE THE SEI ON THE GRAPHITE SIDE CAN BE DAMAGED. KEEP THE CELLS ABOVE 3.0V PER CELL (NO WAY YOU CAN TELL THAT) BUT THAT WOULD BE ABOUT 98% DISCHARGED. SO DO NOT DISCHARGE BEYOND 98%.

Any opinion you may have on this subject will be much appreciated in helping me understand what the limits of this technology are and how to best treat the batteries in my cars and consumer products.

THE TECHNOLOGY IS REALLY PRETTY AMAZING WHEN YOU THINK OF IT. "
 
  • Like
Reactions: Gwgan
Just to clarify. Elon has mentioned (don't remember where) that Model S batteries don't use very much cobalt (expensive?), which leads me to believe that the Model S is using NCA (LiNiCoAlO2) cells (9% cobalt) vs LiCoO2 cells (60%) found in the Roadster. Or maybe LiCoAIO2 without the nickel. Since they are both cobalt based would the professors finding of the LiCo02 cells be that different than a LiNiCoAi02 or a LiCoAi02 cell? Would love Mario to chime in here.
 
http://www.teslamotorsclub.com/member.php/9901-100thMonkeyhttp://www.teslamotorsclub.com/member.php/9901-100thMonkey100thMonkey: Would you be willing to forward my above post to the professor? Would love to get some clarification on this. It is so exciting to get real data!

Here is the response:

"Thanks for this e-mail. I have placed responses within your text using CAPS (not yelling) but so you can find my responses


Jeff Dahn, FRSC
Professor of Physics and Atmospheric Science
NSERC/3M Canada Industrial Research Chair
Canada Research Chair"

"Since electricity appears to have become the defacto currency for renewable energy and I have a great interest in and now investment in renewable technologies (own a Nissan Leaf and Tesla Model S and a 9.8 kW PV system), I have been doing my best to learn whatever I can about energy storage and how batteries work, in particular, the way end user habits effect overall life.

THIS IS AWESOME.

I am wondering if your research has revealed what the best way to extend a batteries life is. I have concluded, as you have noted, that high ambient temperatures degrade these batteries the most. Living in Seattle, we are blessed with a moderate climate so I feel pretty fortunate in that regard.

ABSOLUTELY. AVOID HIGH T WHENEVER POSSIBLE.

Debate continues on whether charging habits will turn out to have much effect on battery life. I have started leaving all of my li-ion consumer products at about 50% charge when they are not in use. Tesla makes this very easy with a battery slider that allows the consumer to choose between 50% and 100% end charging and even has this feature integrated into the phone app to make it very easy to adjust upwards on the fly as long as the car is plugged in. I have installed an 80A (20 kWh) charger so that I can quickly add charge to the Tesla S, allowing me to minimize the inconveniences associated with leaving the pack at a low state of charge and thus increasing the amount of time the battery stays at a mid/low SOC. I tend to do mid pack cycling, discharging generally between 30 and 70% when convenient, while aiming to have the car sit for the longer stretches like overnight, at 50%. I'm not religious about this, just tend to aim in this direction generally.

I THINK THAT IF YOU ARE KEEPING THE CELLS BELOW 4.0V (NO WAY FOR YOU TO TELL THAT, HOWEVER) THE BATTERY LIFE WILL BE VERY IMPRESSIVE. WE ARE TESTING CELLS BUILT IN 2002 THAT HAVE 2002 TECHNOLOGY (LIFETIME IS BETTER NOW) THAT STILL HAVE 75% OF THEIR INITIAL CAPACITY (CYLCED AT 37c THE WHOLE TIME). THESE CELLS WERE CHARGED ONLY TO 4.075V. MODERN CELLS LIKE THOSE IN TESLA CHARGED TO 4.0V SHOULD LAST A FEW DECADES, I SUSPECT, WITHOUT ANY ISSUE (SO KEEP YOUR CAR FROM RUSTING!). WHERE IS 4.0V RELATIVE TO STATE OF CHARGE? MAYBE 75%.

ONE OTHER THING I WOULD RECOMMEND IS TO AVOID HIGH RATE CHARGING AT TEMPERATURE BELOW 0C. ESPECIALLY WHEN THE CELLS ARE ABOVE 75% SOC. TESLA ELECTRONICS MAY PREVENT THIS

In your opinion, do you believe it is worth the effort to keep li-ion batteries at 50% or thereabouts for the bulk of their resting time to extend their long term capacity and if so, do you have any idea how much of a difference this is likely to make over say a 8-15 year time period.

KEEPING BELOW 4.0V MAY DOUBLE OR TRIPLE LIFE TIME COMPARED TO A FULL CHARGE EVERY CYCLE, I SUSPECT.

Also, given that Tesla limits power to the motor during extremes, like high and low charge and high and low temperatures, is there really any reason to avoid running the battery low, assuming Tesla doesn't let you discharge the battery all the way anyway and limits discharge rate as the charge level drops. I assume running the battery to zero (of what the manufacturer allows) does little to the overall life as long as it gets recharged soon after. I assume that low and high states of charge are more an issue if exposure is prolonged, is that correct?

I THINK THAT REALLY DEEP DISCHARGE SHOULD BE AVOIDED AS THEN THE GRAPHITE EMPTIES OF LI AND THE CELL POTENTIAL RISES TO THE POINT WEHRE THE SEI ON THE GRAPHITE SIDE CAN BE DAMAGED. KEEP THE CELLS ABOVE 3.0V PER CELL (NO WAY YOU CAN TELL THAT) BUT THAT WOULD BE ABOUT 98% DISCHARGED. SO DO NOT DISCHARGE BEYOND 98%.

Any opinion you may have on this subject will be much appreciated in helping me understand what the limits of this technology are and how to best treat the batteries in my cars and consumer products.

THE TECHNOLOGY IS REALLY PRETTY AMAZING WHEN YOU THINK OF IT. "