Battery Degradation Scientifically Explained

[KenC]

Either way, isn't Tesla's strategy to minimize Cobalt, so higher on the nickel. Can we just a

So if we drive a model S... Am I to understand then the best practice is to charge to 60 percent... And drain to as close to 0 as safely possible (say 5%) before recharging back up to 60 percent? Any idea if that would also help undo any degradation that's happened after 6 years of charging to 90 percent?

The only way is to try and see. Doesn't sound like the cracks will heal in the cathodes, but who knows. I'd try 60-10, rather than 60-5 in a 6-yr old S, but that's just me.

 

[Renzo]

The only way is to try and see. Doesn't sound like the cracks will heal in the cathodes, but who knows. I'd try 60-10, rather than 60-5 in a 6-yr old S, but that's just me.

Well the research seemed to suggest to get it as low as possible... But I will try 60-10 I guess... My current battery health is at about 87% so I've lost about 13% capacity as the car aged 6 years. I bought it used from Tesla direct about 10 months ago so I've been trying to baby the battery as much, only charged to 100 3 or 4 times in the ten months I've had it for road trips.

 

[mechrock]

That is extremely interesting. Basically its WAY better to be 0-60% than 100% to 40%. A question that keeps popping up is what is 0% Voc, and what is 100% Voc? They're stating 2.7v and 4.3v, while this post implies 0% is ~2.8v and 100% <4.2v.

I'm also wondering what the NCA makeup of our exact Model 3 2170 cells are (I've only heard 8:1:1 ratio). They mentioned the S/X cells use NCA84 in their 18650 cells.

This is more about mechanical stress with really high nickel concentrations

And with NCA80 / NCA88

Very interesting! If I'm understanding this correctly, NCA80 is Model S/X and NCA88 is Model 3?
This would indicate Model 3 doesn't like to be fully charged even less so than Model S/X?
Fortunately this kind of validates my logic of only daily charging to 60% and only charging to 90% on trips.
I wish they would do this again, but using 90-10% DOD. It sounds as though 90% will have minimal cracking compared to 100% and if it's also true that just by getting 100% does the damage vs. letting it sit at 100% then that's really good to know.

Personally I've only charged to 100% twice and see little need to ever charge to 100% again.

 

[Jedi2155]

Very interesting! If I'm understanding this correctly, NCA80 is Model S/X and NCA88 is Model 3?
This would indicate Model 3 doesn't like to be fully charged even less so than Model S/X?
Fortunately this kind of validates my logic of only daily charging to 60% and only charging to 90% on trips.
I wish they would do this again, but using 90-10% DOD. It sounds as though 90% will have minimal cracking compared to 100% and if it's also true that just by getting 100% does the damage vs. letting it sit at 100% then that's really good to know.

Personally I've only charged to 100% twice and see little need to ever charge to 100% again.

We do not know the exact chemistry of the Model 3 AFAIK. The author of the above article is not affiliated with Tesla so they're using research cells with different concentrations of nickel to give a spectrum of potential degradation.

They did state that the S/X is at NCA84, but we don't know what the Model 3 is.

 

[Jedi2155]

Interesting article:

"The US-based EV maker will be using LG Chem’ 21700 type batteries using NCM811 that boast a nickel proportion of 80% or more." (emphasis added)

"Until now, Tesla has been supplied by Japan’s Panasonic, which uses NCA, which is another type of high nickel cathode material."

So the "even higher" nickel content in the NCMA battery is on top of the 80% or more nickel proportion in the NCM811. I've never heard of NCM811--M is for manganese. Here are some references:

2019 PREVIEW: NCM811 batteries 'not likely' to be used in electric vehicles in 2019 | Metal Bulletin.com

NCM 811 Lithium-Ion Cells Quickly Increase In Market Share

NCM 90: successor of NCM 811 battery cells - PushEVs

Industry typically uses the acronym NMC not NCM (I've worked in the industry for a number of years but not presently).

 

[Dr. J]

That is extremely interesting. Basically its WAY better to be 0-60% than 100% to 40%. A question that keeps popping up is what is 0% Voc, and what is 100% Voc? They're stating 2.7v and 4.3v, while this post implies 0% is ~2.8v and 100% <4.2v.

Maxim Integrated says,
"Most Li-ion batteries have a maximum charged voltage of 4.20V/cell, and every reduction in peak charge voltage of 0.10V/cell is said to double the cycle life. For example, a lithium-ion cell charged to 4.20V/cell typically delivers 300–500 cycles. If charged to only 4.10V/cell, the life can be prolonged to 600–1,000 cycles; 4.0V/cell should deliver 1,200–2,000 and 3.90V/cell should provide 2,400–4,000 cycles. A lower peak charge voltage reduces the capacity the battery stores. As a guideline, every 70mV reduction in charge voltage lowers the overall capacity by 10%. For absolute best longevity, the optimal charge voltage is 3.92V/cell (this may vary with chemistry). Reports I’ve seen say that this threshold eliminates all voltage-related stresses; going lower may not gain further benefits, and may induce other symptoms."
A Look at Tesla’s Latest Battery Technologies

If this is applicable and the relevant relationships are linear, and user-charging habits affect cycle life as indicated (it may be that peak charge voltage set by the mfg is all that matters)--if all that is true, charging to 1 - [(4.2V - 3.92V) / (4.2V - 2.8V)] = 80% is ideal for cycle life, "which may vary with chemistry."

If the relationship isn't linear, following the passage above (every 70mV reduction in charge voltage lowers overall capacity by 10%), 3.92V gets you to 60% SoC (-0.28V is 4X -70mV, so -40%).

Of course this doesn't address degradation specifically, only longevity.

 

[Jedi2155]

If the relationship isn't linear, following the passage above

The important thing to understand is that every single lithium chemistry variation has a slightly different charge/discharge curve so while you can use the 70mV/10% as an approximation, it is still just an approximation. Lithium Iron Phosphate (LFP) vs NMC (LG Chem's variation) vs NCA (Panasonic / Tesla) has quite a bit of difference.

 

[mechrock]

For absolute best longevity, the optimal charge voltage is 3.92V/cell

Appears to be working so far. I've kept my daily charge set to 60% since day 1. Charged to 100% 3 times ever, left immediately upon finishing. 90% <20 times, 10 of those being supercharging, all leaving immediately upon finishing.

Just charged back to 60% after a trip. 189 miles on my LR RWD Model 3. 195 miles would be new. 2% loss after 1.5 years and 17,000 miles. Or maybe I should say a 1% gain from 310 max since when I picked up the car, thanks Tesla!

 

[jaitch]

According to the excellent @EV-Tech Exp videos, and using his example resistance numbers for various components etc, the pack resistance of a nominally new 90 comes out at 47.3 mOhms. Using his voltage drop method at high current my specific pack resistance number comes out at 66.2 mOhms.

Are there any battery experts who have an opinion on what "excessive" pack resistance would likely be? Also any comments on whether the method to establish pack resistance is flawed or otherwise - i.e. use delta volts from off load to on load at high current to calculate R?

Thanks

James

 

[earthwormjim]

According to the excellent @EV-Tech Exp videos, and using his example resistance numbers for various components etc, the pack resistance of a nominally new 90 comes out at 47.3 mOhms. Using his voltage drop method at high current my specific pack resistance number comes out at 66.2 mOhms.

Are there any battery experts who have an opinion on what "excessive" pack resistance would likely be? Also any comments on whether the method to establish pack resistance is flawed or otherwise - i.e. use delta volts from off load to on load at high current to calculate R?

Thanks

James

What is the measurement procedure? It's pretty hard to measure mili-ohms accurately. Connectors within the car are going to contribute to increased resistance too as they age and see use.

 

[jaitch]

What is the measurement procedure? It's pretty hard to measure mili-ohms accurately. Connectors within the car are going to contribute to increased resistance too as they age and see use.

Not sure about it being a procedure but it was suggested in one of his videos that resistance could be shown by using volt drop vs current. See my attached excel pic that note summarizes what I did. I don't know if it is correct or valid or how thermal affects change resistance etc. Looking for some guidance, comments or otherwise!

 

[earthwormjim]

Not sure about it being a procedure but it was suggested in one of his videos that resistance could be shown by using volt drop vs current. See my attached excel pic that note summarizes what I did. I don't know if it is correct or valid or how thermal affects change resistance etc. Looking for some guidance, comments or otherwise!

View attachment 476915

Specifically I was wondering where you were measuring voltage across a shunt in your car. Do you disconnect the high voltage connection, then put a calibrated high power shunt in series with the connection?

 

[jaitch]

Specifically I was wondering where you were measuring voltage across a shunt in your car. Do you disconnect the high voltage connection, then put a calibrated high power shunt in series with the connection?

Oh right. I use ScanMyTesla to read the canbus data. Voltage and current etc etc is in there.

 

[mechrock]

Clarification needed when it comes to charging a "cold" battery.
Twitter Post
I got into this debate and wanted to make sure my information was sound. This is great advice if your battery is cold soaked, but should be completely un-needed if already warmed up correct? I know you can have lithium plating if you charge at a cold temp, but at 2c or above, I wouldn't think you'd see any benefit to warming the battery at AC charging speeds (11.5KW).

 

[ord3r]

According to this: http://xqdoc.imedao.com/16764c45aca1ea83fdfe1636.pdf article Tesla Model 3 cells have a 56% Nickel Cathode

Any thoughts on If this would mean that study on micro-cracks would apply to Model 3?

That article has a ton of other interesting information on Model 3 batteries and drivetrains, still reading/digesting.

 

[random155]

Just watched the video and learned a lot. However, while I understand that charging at a high state of charge for a prolonged period will contribute to increased degradation, what can we define as a high state of charge? Is supercharging really considered a high state of charge or just higher than charging at home? I would hope that Tesla would know and understand at what state of charge these negative effects happen to our batteries and would not allow charging levels to reach a point where batteries begin to degrade, even while supercharging. Furthermore, temperatures are mentioned, but what temperatures? My model s screams at superchargers when charging in warm weather while cooling the battery. My hope is that Tesla understands exactly where charging temperatures need to be in order to prevent excess degradation at superchargers. I am not a tesla fan boy, but a supercharger fan boy, that is why i raise these questions.

 

[Silicon Desert]

Is supercharging really considered a high state of charge or just higher than charging at home? I would hope that Tesla would know and understand at what state of charge these negative effects happen to our batteries and would not allow charging levels to reach a point where batteries begin to degrade, even while supercharging.

You might realize that high STATE of charge and high RATE of charge are 2 different things. And it seems Tesla has already been changing the charging patterns to reduce potential greater degradation to the dislike of some members.