Range Loss Over Time, What Can Be Expected, Efficiency, How to Maintain Battery Health

[Steve446]

Thank you! Interesting that you charge immediately and then your car sits at 60% for overnight (half day?) and cycle is 60% to 50% back to 60%. Was wondering how your battery had handled lithium plating and SEI buildup through your procedures. Lot of literature on how some of these help maintain capacity. I don’t know the details, but interested if other know.

On my 2022 M3P, I have been charging only to 50% and don’t drive everyday and keep my SoC super low. Sometimes sits for days at 20-30% in enclosed garage and even lower and I am showing more degradation (was 7% but recently jumped to 8%+) than I would have thought. Yes we have a hot summer in Vegas here but I don’t do most of my driving until the evenings. Sometimes I just charge to 30-40% and don’t go to 50% because it’s enough for what I do.

Interesting and great to read all of the comments which reinforce that low SoC and low temperature are good for low battery capacity reduction, as predicted by the scientific data that @AAKEE has presented. For my own experience, where I live, the average annual temperature is about 15C. Over the summer months where my car's average battery temperature is around 25-27C I can see a small monthly degradation rate. Over winter when the average battery temperature is 14-15C, it's sometimes hard to observe any capacity loss. I acknowledge that "battery calibration" may play a role here since I can see around a +/- 0.5-1kWh difference in NFP when SoC is made "0" or 100% on screen. I assume though that since calendar aging is a chemical process, capacity loss at lower temperatures still occurs, though more slowly. What about though the reported slow down in capacity loss after the first 5% or so? is that real or just an artifact? Does anybody have any thoughts on that?

 

[AlanSubie4Life]

What about though the reported slow down in capacity loss after the first 5% or so? is that real or just an artifact? Does anybody have any thoughts on that?

@AAKEE has addressed this, he can give the actual correct information or you can search for his posts. Basically is expected a square root (fitted - I am not sure there is anything physical about that) relationship with time.

As I understand it (very simply - this could be wrong): Physically the SEI formation inhibits further SEI formation. So as cyclable lithium is used up, it tends to reduce the tendency for more cyclable lithium to be used up. Seems to me like that would be more of an exponential decay sort of behavior, but that wouldn't look that different than a square root relationship at the beginning. And the actual mechanisms may be more complex than described.

I assume though that since calendar aging is a chemical process, capacity loss at lower temperatures still occurs, though more slowly.

Yeah that is what the test data shows. There are diminishing returns when things get cold enough since the rate of loss is quite low already, getting colder doesn't help too much more (there's nearly nothing left to gain). And things are quite bad indeed at very very high ambient temps.


My understanding is if you take a car that has been very cold and at low SOC and start using it in very warm conditions and higher SOC, capacity loss rates will go way up, even if the car is years old, and then the loss rates will decay down over time as the SEI is built, just as it would have originally. So that wouldn't look like a square root overall, but the parameters of the square root decay scaling depend on conditions.

 

[AAKEE]

What about though the reported slow down in capacity loss after the first 5% or so? is that real or just an artifact? Does anybody have any thoughts on that?

Not sure I got the question right, but anyway:

Calendar aging reduces the rate with the square root of time. Calendar ging builds Solid Electrolyte Interphase (SEI), which itself acts as a protection against further calendar aging. Initially the calendar aging is rapid but the SEI reduces the rate.
What happened after 1 month will be doubled after four months.
What happened after 1 year will be doubled after 4 years.

So, yes the rate slows, but not exactly after 5%

The blue line is my plan for my Plaid. Origin is 99.4kWh capacity and time set so the calendar aging started when the car was built (the batteries seem in general built not long before).
Amber = actual Nominal Full Pack from the BMS. Green, Nominal remaining at 100% charges.

*) The capacity was high at delivery but the BMS was off track (hadn't tuned itself on the NFP maybe). Despite showing 95.7-96 kWh or so, the real capacity was above 98kWh.

 

[Steve446]

Thanks to both of you for the prompt replies. I think I (maybe!) understand better now. So do I understand that the consequences of SEI build up on slowing capacity loss is either a titration issue or more likely a permeation effect. @AAKEE your Plaid battery parameters for NFP and Nominal remaining vary within 1 kWh, maybe it's difficult to resolve that data more precisely than that since both come from the BMS?

 

[AAKEE]

@AAKEE your Plaid battery parameters for NFP and Nominal remaining vary within 1 kWh, maybe it's difficult to resolve that data more precisely than that since both come from the BMS?

The nominal full pack is an estimated capacity, built on the BMS readings of the pack. It might/could be slightly conservative, but whennthe BMS is spot on the capacity it calculates the real capacity very close to a manual calculation of the capacity.
The nomimal remaining at a full charge has always been around 0.5 kWh or so above the NFP, but not always 0.5 so it varies between ~ 0.3 to 1.0 above NFP.

I saw I had a typo in february, Nominal remaining was wrong.

 

[DrChaos]

What about though the reported slow down in capacity loss after the first 5% or so? is that real or just an artifact? Does anybody have any thoughts on that?

That is apparently real and observed in all the scientific publications on calendar aging. There are multiple mechanisms chemically for capacity loss from calendar aging (and even more mechanisms for other kinds of aging) but the primary mechanism apparently is loss of mobile lithium which sticks to the graphite in a growing permanent film. There's less lithium to move back and forth and thus lower energy capacity.

My guess is that a virgin graphite anode has lots of places for the lithium to stick to, but after the initial burst of lithium deposits the rest is somewhat less attractive for the chemical reaction and so it happens slower. At higher states of charge there is more lithium near the graphite side, and so the reaction proceeds faster.

In the battery literature the simplest model over time is roughly sqrt(time) though it really could be C*sqrt(time*A + B) for various coefficients A and B and C, and even this is an empirical approximation and not derived from first principles.

It could also be an addition up of multiple mechanisms with different aging rates and limits, like a few exponential initially plus a long term slower linear added up together which can be fit well to a sqrt() in initial phases. Nobody knows for sure and they can't be easily distinguished.

 

[DrChaos]

@AAKEE has addressed this, he can give the actual correct information or you can search for his posts. Basically is expected a square root (fitted - I am not sure there is anything physical about that) relationship with time.

As I understand it (very simply - this could be wrong): Physically the SEI formation inhibits further SEI formation. So as cyclable lithium is used up, it tends to reduce the tendency for more cyclable lithium to be used up. Seems to me like that would be more of an exponential decay sort of behavior, but that wouldn't look that different than a square root relationship at the beginning. And the actual mechanisms may be more complex than described.

playing armchair chemist......

if loss L ~ t^(1/2) then dL/dt = 1/2 t^(-1/2) = 1/2 * 1/L

like the attraction of any free lithium to graphite is inversely proportional to thickness of already existing solidified film (1/L) between free lithium and the graphite.

 

[Steve446]

That is apparently real and observed in all the scientific publications on calendar aging. There are multiple mechanisms chemically for capacity loss from calendar aging (and even more mechanisms for other kinds of aging) but the primary mechanism apparently is loss of mobile lithium which sticks to the graphite in a growing permanent film. There's less lithium to move back and forth and thus lower energy capacity.

My guess is that a virgin graphite anode has lots of places for the lithium to stick to, but after the initial burst of lithium deposits the rest is somewhat less attractive for the chemical reaction and so it happens slower. At higher states of charge there is more lithium near the graphite side, and so the reaction proceeds faster.

In the battery literature the simplest model over time is roughly sqrt(time) though it really could be C*sqrt(time*A + B) for various coefficients A and B and C, and even this is an empirical approximation and not derived from first principles.

It could also be an addition up of multiple mechanisms with different aging rates and limits, like a few exponential initially plus a long term slower linear added up together which can be fit well to a sqrt() in initial phases. Nobody knows for sure and they can't be easily distinguished.

Hmm the maths is not my strong point But, what I retain from the comments from all of you and my curiosity, briefly reading round the subject is that a stable SEI is essential for good capacity retention and cycle life, this is best achieved by "storage" at lower temperatures. Higher temperatures lead to a thicker more unstable SEI which can lead to reduced capacity and reduced cycle life. That's the end of my 101 since there's probably more to it than that. The practical results though have been well exposed with all the published data available here!

 

[Steve446]

The nominal full pack is an estimated capacity, built on the BMS readings of the pack. It might/could be slightly conservative, but whennthe BMS is spot on the capacity it calculates the real capacity very close to a manual calculation of the capacity.
The nomimal remaining at a full charge has always been around 0.5 kWh or so above the NFP, but not always 0.5 so it varies between ~ 0.3 to 1.0 above NFP.

I saw I had a typo in february, Nominal remaining was wrong.

I was just interested to know if your NFP was impacted by exposing the battery to "0" or 100% SoC. I do that around 4-6 times per year but only for road trips and the 100% is just prior to leaving. I also check the energy screen values at that time. A few hours later, I generally, but not always, see an increase in NFP of about 0.5 kWh. Nominal remaining generally aligns with the NFP (did I get the real SoC right?). Below is an example from end April as I started my Irish road trip.

Data Point (SMT data)	0h	+11h and after supercharging
FPWN (LG M48 in M3LR 2021 MiC)	74.5
Screen SoC	100	89.5 (real SoC ~90.1)
Energy Screen	71900
NFP	72.1	72.5
Nominal Remaining	72.1	65.3 --> 65.3/0.9 = 72.5
Buffer	3.24	3.26
Usable	68.9	62

After a few weeks (~6), the NFP has dropped back by about the same amount although any supercharging can impact that. All this is is just my curiosity getting the better of me

 

[Snos Rap75]

Hi.

12 months ago, 100% charge would get me 294 range in miles, at least on my screen....

A year later and this is now 267 for a100% charge. As advised I have been letting my battery drop to sound 50% less when it's been possible.....


Is there anything else I can do...? It's very frustrating....

 

[LikesToys]

In order to advise you would have to define what type of battery you have LFP vs NCA - model and year. Then follow best practices for that specific battery chemistry

 

[Snos Rap75]

Model 3, long range dual motor, 2019. Not sure what LFP or NCA is?

 

[tashtibet]

https://www.reddit.com/r/TeslaLounge/comments/1d08d5j/my_new_model_3_lrdm

 

[EVRider-FL]

You used to be able to switch the range display between estimated and ideal values or something like that. Check Controls > Display > Energy Display and see if those options are there. Also, I think Tesla recently changed the estimated range to account for the age if your battery, so that could be a factor.

 

[rpiotro]

Are you charging it past 80% for daily use? A daily charge to 100% should be avoided unless you are immediately heading out on a road trip.

In a perfect world you should keep the SOC between 40% and 70% with that battery. We don't live in a perfect world but, between having a home level II charger and my typical daily use I can keep it in that range quite often. According to Tesla, 80/20 is fine too.

My SR+ will turn three years old this summer. Only about 18K miles though. According to my Tessie app the SOH (state of health) of the battery is 94.2%. They say the first 5% goes pretty quick. In then plateaus for a good while. That's what I'm seeing here. I guess coddling the battery pays off.

Also, it may just be a calibration issue. They say you should drop it to 10% maybe twice a year to recalibrate. If that is important to you.

 

[Snos Rap75]

Generally never go over 80 unless going on a long trip. What is SOC?

I do struggle to do it to 10% to be honest.

Interesting actually to see what people have taken theirs down to. I was in a trip yesterday and had the choice to wait until I reached the charger I wanted to but battery was expected to be on 13% on arrival. I bottled it and stopped at another one before.... Had anyone rushed taking it so low and does it work until across hitting zero?

 

[tstafford]

Generally never go over 80 unless going on a long trip. What is SOC?

I do struggle to do it to 10% to be honest.

Interesting actually to see what people have taken theirs down to. I was in a trip yesterday and had the choice to wait until I reached the charger I wanted to but battery was expected to be on 13% on arrival. I bottled it and stopped at another one before.... Had anyone rushed taking it so low and does it work until across hitting zero?

SOC is "state of charge"

 

[M3BlueGeorgia]

Generally never go over 80 unless going on a long trip. What is SOC?

I do struggle to do it to 10% to be honest.

Interesting actually to see what people have taken theirs down to. I was in a trip yesterday and had the choice to wait until I reached the charger I wanted to but battery was expected to be on 13% on arrival. I bottled it and stopped at another one before.... Had anyone rushed taking it so low and does it work until across hitting zero?

13% is generally safe for an arrival percentage providing you are reasonably close to the supercharger (say less than 75 miles) and aren't expecting a significant change in environment before getting to the Supercharger.

Closer you are, lower risk.

Examples of "significant change"
- Driving into torrential rain
- Driving into severely dropping temps

 

[ucmndd]

13% arrival in good weather is no problem at all. I’m comfortable with half that.

 

[AlanSubie4Life]

12 months ago, 100% charge would get me 294 range in miles, at least on my screen....

A year later and this is now 267 for a100% charge

These data points don't make a lot of sense for a 2019.

267 is low for a 5-year-old LR Dual Motor (though certainly possible). That's 65.4kWh, which is down 16%.

What doesn't make sense is that after four years the vehicle was down 7.5%, and now after 5 years it's down 16%. That's abnormal.

So could be a bad datapoint or there may be something that's about to go south with the pack. Most likely it's a bad datapoint.

Remember that the rated miles are just units of energy. They reflect how much energy your pack contains.