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

[AlanSubie4Life]

(there is some part of the loss ”hidden” but we can disregard that for this discussion).

Yep, for 2020 Performance there was basically none of the loss hidden. 77.8kWh is where loss started to show and that is about where batteries started.

Perhaps 0.5kWh at most was hidden. There is relatively little data on SMT reads from those packs but I don’t recall seeing anything above 78.x kWh.

 

[nurusz]

1. The battery will survive the car in many cases (~10-15 years) by just following the very few tips/rules Tesla gives.
2. Read 1 again and make sure you understand it. Dont worry too much.
————————
Artificial pause
————————

3. If you follow the Tesla simple rules, like only use max 90% charge for daily use this will not result in the minimum degradation.
4. It is possible to reduce the degradation by about 50% (without any big sacrifice) by applying the known facts from research:
- Use low SOC when possible. Below 55% cut the degradation from time in half.
- Charging late, like in the morning before the drive instead of when arriving in the evening before reduce the time at higher SOC.
-Do not charge more than you need (including a margin for range anxiety).
- Charge often - the smaller each cycle the smaller the degradation from cycles.

5. Low SOC is NOT dangerous. Below 20% is safe and do not cause increased wear. It actually reduce the calendar aging.
100% SOC is not as bad as the forum myth says. You do not need to drive asap. The degradation is about the same at 100% as at 80-90%. Sometimes slightly more, sometime slightly less.
90-100% increase the cyclic aging, this is the main reason to stay below 90% for daily driving.
High ambient temperature increase the cyclic aging. The combination of high SOC and high temperature causes more degradation by time. Avoid leaving the car with high SOC very warm days.

Thank you. That's clear enough for anyone

 

[FirstInTown]

GREAT thread about calendar aging. I charge our 2022 MX at 50% nightly and use about 15-20% daily in the winter, so this should minimize calendar aging in a practical way. (Can't set charge to less than 50 and don't want to go 2 days or 50-->10 and then charge, so this is practical way of keeping charge under the 55% point.

I have looked at the charts in this thread and all show high charge AND high temp is a bad thing for capacity loss. My question is the Preheating for Super Charging that takes the battery to the dreaded 50degC before even start charge. We live in the cold for winter where it is -20C in the mornings lots of time. I remember reading how fast charging under 10C is BAD and the car won't really do it. But warming to 50C seems like "damn the battery life, we just want you to be able to take a 250kwH rate and go about your merry way"

So last time I Super Charged, I tried to delay nav to the SC until about 30 minutes out. Battery was warmed from about 12C to 30C and took at 120kwH on a 150kwh charger. Charged for 6 minutes and left as we didn't need much to really get home. Battery was STILL being warmed by the car during the charge session according to the phone app and hit 42C by the time we left. Car SLOWLY used the heat into the cabin (Heat Pump car)and battery came down to 25C 1.5 hours later when got home as it was about 0C out by this time.

Is what I did better? Or should I have let the car take the battery to 50C before got to charger if nav started earlier? (and if the des was a 250kw charger, would have charged faster but I knew it was an older 150.)

I don't trust Tesla to not be just taking the battery as warm as possible (50C+) to minimize time at the charger, at the expense of my battery life. Less time at charger makes for good public relations (I can brag how fast it charged) and doesn't tie up a station for other Teslas, so is to their advantage. But not my batteries life???

Or: What is the ideal temp to start Super Charging at for battery life, NOT fastest charge rate?

 

[AAKEE]

GREAT thread about calendar aging. I charge our 2022 MX at 50% nightly and use about 15-20% daily in the winter, so this should minimize calendar aging in a practical way. (Can't set charge to less than 50 and don't want to go 2 days or 50-->10 and then charge, so this is practical way of keeping charge under the 55% point.

Good plan!

I have looked at the charts in this thread and all show high charge AND high temp is a bad thing for capacity loss. My question is the Preheating for Super Charging that takes the battery to the dreaded 50degC before even start charge. We live in the cold for winter where it is -20C in the mornings lots of time. I remember reading how fast charging under 10C is BAD and the car won't really do it. But warming to 50C seems like "damn the battery life, we just want you to be able to take a 250kwH rate and go about your merry way"

Research about fast charging of lithium batteries did find that preheating to 40C more or less eliminated the lithium plating that is the degrading part from supercharging.
Tesla reduces the charging speed if the battery is not fully precondituioned, this is more or less safe to say that it is made to counter lithium plating.
But somewhere there is probably a compromise as non preheated batteries will charge slower and be like a brake pad in the charging que at superscharging stations.

I always precondition completely if it is possible, if not I precondition as much as possible.

So last time I Super Charged, I tried to delay nav to the SC until about 30 minutes out. Battery was warmed from about 12C to 30C and took at 120kwH on a 150kwh charger. Charged for 6 minutes and left as we didn't need much to really get home. Battery was STILL being warmed by the car during the charge session according to the phone app and hit 42C by the time we left. Car SLOWLY used the heat into the cabin (Heat Pump car)and battery came down to 25C 1.5 hours later when got home as it was about 0C out by this time.

Is what I did better? Or should I have let the car take the battery to 50C before got to charger if nav started earlier? (and if the des was a 250kw charger, would have charged faster but I knew it was an older 150.)

Lets say the battery is at elevated temperature for two hours for a supercharging session. The end temperature will not differ much even if you do not precondition as the battery will be heated longer during the charging session.

So the real difference is about one hour before the charging session, when preconditioning before the charging. This will in general be at lower SOC, as most people supercharge from a lower level, at least not above 50%.

So by preheating you perhaps follow the read line for one hour, instrad of the blue line. The difference between the read line and the blue line is the difference in calendar aging, for a new battery (first year) we see about 5% for 10 months.
10 months is about 7440 hours.

That 1 hour of elevated battery temp will cause you 1/7440 of 5%, or 0.0007%.

Another way to see it is that 10 months of preheating would cause 5% degradation.
These 10 months of preheating would be enough for 7440 supercharging sessions.

A slightly leading question: what do you think is worse, 7440 supercharging sessions or 5% of increased calendar aging

I don't trust Tesla to not be just taking the battery as warm as possible (50C+) to minimize time at the charger, at the expense of my battery life. Less time at charger makes for good public relations (I can brag how fast it charged) and doesn't tie up a station for other Teslas, so is to their advantage. But not my batteries life???

Or: What is the ideal temp to start Super Charging at for battery life, NOT fastest charge rate?

This is a picture from a research report about lithium plating from fast charhing.
At 1C (about 75 kW charging power for a model 3/Y) we need 25C to avoid lithium plating.
Its obvious from this chart that 40C or above is a good idea if we do 150-250 kW.

I once had to supercharge in winter time with a cold battery, had no options.
The battery was some 15-20C and the charging speed rocketed to 137-138 kW more or less directly (t’was a supercharger V2 and the other 12 kW was heating).
I did’nt feel happy over that situation but I had no choise. Just took the minimum SOC to make it home.

[Edit]Maybe this picture needs an explanation. Cold be misinterpreted without any more information. The batteries cecled above the lithium plating temperature perhasps did 4000 cycles before reaching 20% capacity loss as per the picture, but at -10 or 0 or +5C the battery perhaps only lasted 40 or 400 cycles before reaching the 20% capacity loss. [End Edit]


From my guess, without having the data from Teslas compromise of short charging time vs low degradation, I guess that preheating is not a bad thing.
I am quite sure that the loss is bigger by not preconditioning.

This is one of the few things we can not read research and be sure how to handle it because we do not know how Tesla did draw the line.

Everything points to the ”should precondition” if the battery care is in the mind.

 

[AAKEE]

Here is capacity loss due to SEI put in charts for the same C-rate and temperature.
Calendar aging is mainly a SEI growth issue. The SEI growth in these charts is mainly because of the cycles(cyclic aging), and not calendar aging.
We can see in the upper row (for example) that the SEI growth is not very different during cycles between different temps. It has a marginal effect, but the lithium plating has a big temperature dependent effect, and at lower temperature it’s bad. For 25C and 45C, there is no reduced capacity from lithium plating at 1C. At 2C, there is none at 45C but barely visible at 25C.
At 3C, there is some, not much at 25C and Still non at 45C.
I think this research was performed on NMC-cells, if I recall it Correct.

 

[FirstInTown]

WOW!!! Lots of good stuff. Have to digest the Lithium plating and SEI and the effects. This thread since page 270 has had great knowledge about storage temp and the 55% thingy. Need to learn about the Lithium plating and SEI and lots of this data looks like shown in the last 2 posts.

Thanks!

 

[conv90]

Just installed in the past night the 2023.6.8 update.
Something went wrong (sure for many Italian Owners).
It seems they changed the constant and/or mixed some value (km with miles calculations).
Something in the equation is wrong.
The energy screen is displaying 1,5x to 2.0x consumption.
If usualy I have a 160 Wh/km , today I had about 280-300 Wh/km . It seems it sucks like when you go to a supercharger.
Actually the real consumption seems OK (I made 52 km out of a 11%).
The little box at the left bottom of the screen is out of range too ( not as the Consumption Tab graph).
So 52 km , it seems I used 11% only.... so near to the constant of about 155 Wh/km (normal for me with no AC in a 18-20°C day ).
180 Wh/km on the little box at the left bottom (slightly out)
Energy screen Tab Consumption with the graph almost always over 300 Wh/km.
In the graph showing if you are driving over or less the constant, it seems TOO GREEN to be real (almost all green line and very little orange parts).
Im the only one?

 

[AAKEE]

60.000km / 2 years 3 months update:

78.8 kWh capacity / 495-496* km range left (out of 507 km).

The Low SOC strategy works fine!

*Theres a bug in Scan my Tesla so full rated range is dependant on the SOC, at hogh SOC 495-496 is shown, and when SOC decreases, that full rated rabge decreases.

 

[AlanSubie4Life]

The Low SOC strategy works fine!

And keeping the pack cool (or frozen!) a large % of the time. I don’t remember how much that matters at low SOC but I suspect you will tell me.

Hmm…seems like temp definitely still matters a lot based on above plots.

 

[AAKEE]

And keeping the pack cool (or frozen!) a large % of the time. I don’t remember how much that matters at low SOC but I suspect you will tell me.

Hmm…seems like temp definitely still matters a lot based on above plots.

Its of course a temp x SOC x time effect as I
use to refer to.

Low temperature alone wont do it. Its a good thing, or rather much heat is not.

You can see that while there is a difference between 25C and 10C, it is not very big.
Its about 15-20% lower at 10C than 25C.

This was the forst five months so:
80% at 25C for 2 years would ”cost” 7.7%
80% at 10C for 2 years would cost 5.9%
30-40% at 25C for 2yrs = 3.3%
30-40% at 10C for 2 yrs = 2.2%

The big difference is keeping the SOC below 55%, that cut it in half.
From 40C to 10C cut it in half but 40C average over a year is not that common I guess.
Last time I checked, my one year average was 13.46C if my memory dont fail me.
I had 35% avg SOC, but thats has increased to about 40-45% ( new job, car often home for weeks charged daily to 55% and only short drives between working weeks).

That said, the average degradation here and what we see in TMC from USA is not huge.
Most people here in Sweden see 5% the first year and 10% after four years.
There is companies testing cars before selling second hand and 10-14% seems normal according to those tests.
( I suspect they use to high load and get a heat loss not accounted for).

So the other cars around here see more than the double degradation compared to my car.

 

[AAKEE]

@KenC

I’d like to try to put your cars values in my degradation formulas (as it is lower than average). Im doing a automated degradation calculation and like to try it on your car. (My car fit perfect in the calc.)

I need these values:
- Age in months of the car or purchase/manufacturing date.
- Average temperature for the car during a year
- Regular charging SOC.
-End of the day SOC, before next charge.
-Time of day when the charging is set to start.

 

[PianoAl]

- Age in months of the car or purchase/manufacturing date.
42 months

- Average temperature for the car during a year
55° F

- Regular charging SOC.
80%.

-End of the day SOC, before next charge.
50%

-Time of day when the charging is set to start
I use departure charging, but on average it starts around 5 am.

64K miles.

 

[KenC]

@KenC

I’d like to try to put your cars values in my degradation formulas (as it is lower than average). Im doing a automated degradation calculation and like to try it on your car. (My car fit perfect in the calc.)

I need these values:
- Age in months of the car or purchase/manufacturing date.
- Average temperature for the car during a year
- Regular charging SOC.
-End of the day SOC, before next charge.
-Time of day when the charging is set to start.

Mfr'd Oct 2018, delivered Dec 13, 2018
Avg ambient temps?

Because of my fat fingers, it's usually 58%.
Typically, end of day is 50%.
It starts when I get home, around 5pm.

 

[DrChaos]

@KenC

I’d like to try to put your cars values in my degradation formulas (as it is lower than average). Im doing a automated degradation calculation and like to try it on your car. (My car fit perfect in the calc.)

I need these values:
- Age in months of the car or purchase/manufacturing date.
- Average temperature for the car during a year
- Regular charging SOC.
-End of the day SOC, before next charge.
-Time of day when the charging is set to start.

Can I participate?

1. June 2022 (m3lr)
2. San Diego California, guess about 68F, 20C
3. 50%
4. approx 38% most days
5. scheduled departure, usually charges 4:00-430 am -> 5:45 am.

 

[AAKEE]

Mfr'd Oct 2018, delivered Dec 13, 2018
Avg ambient temps?

Because of my fat fingers, it's usually 58%.
Typically, end of day is 50%.
It starts when I get home, around 5pm.

My formulas put you at 5.68% degradation, remainin cap 73.4 kWh.
I know your car is doing better than that...

 

[AAKEE]

Can I participate?

1. June 2022 (m3lr)
2. San Diego California, guess about 68F, 20C
3. 50%
4. approx 38% most days
5. scheduled departure, usually charges 4:00-430 am -> 5:45 am.

Its a 82.1kWh Panasonic, right? The initial shown capacity with this battery varies (much compared to the full pack when new), long term calculations should be ok anyway I think but after nine months, not sure.

So it says 2.78% degradation -> 79.8kWh
The LR is still capped at 79kWh, I guess so you should still have full range shown.

 

[DrChaos]

Its a 82.1kWh Panasonic, right? The initial shown capacity with this battery varies (much compared to the full pack when new), long term calculations should be ok anyway I think but after nine months, not sure.

So it says 2.78% degradation -> 79.8kWh
The LR is still capped at 79kWh, I guess so you should still have full range shown.

My full range at delivery was 358 mi, which is the USA EPA range value. It stayed fixed there for 3 months or so.

I'm now showing about 346-350 (extrapolating from SoC and displayed range at that SoC, with an occasional 100% charge showing the same). 2.78% down from 358 is right on though. I presumably didn't have all of the 82.1 initially.

All the drop was after the warm months (which are July through November in San Diego), and has stayed virtually the same since then. I presume there will be another drop every summer. There are usually 2-4 weeks of strong heat for about a week at a time anytime from July through October.

A theoretical question: I've read how the calendar degradation is modeled as proportional to sqrt(T) with T being time. Suppose the baseline degradation for an average person who charges to 80-90% is A * sqrt(T/T0). If I charge only up to 50% and reduce degradation by 2x, does this mean the degradation is

A/2 * sqrt(T/T0)

or

A * sqrt(0.5 T/T0) meaning the argument of the sqrt takes twice as long to hit 1

The first is better than the second, meaning the slowdown in degradation happens at the same age (i.e. derivative of degradation w.r.t. time), but I suspect the second is more likely to be true, because it involves the deposition of lithium on the electrodes which doesn't come back out and that depends on their physical state. So with less already deposited, it deposits faster.

 

[inigoml]

@KenC

I’d like to try to put your cars values in my degradation formulas (as it is lower than average). Im doing a automated degradation calculation and like to try it on your car. (My car fit perfect in the calc.)

- Age in months of the car or purchase/manufacturing date.
20 months (well, 12 months since battery was replaced in April 22)

- Average temperature for the car during a year
18 degrees

- Regular charging SOC.
55% last six months. Previously 70%

-End of the day SOC, before next charge.
30%

-Time of day when the charging is set to start.
01:00 at home (25%), 10:00 at work (garage, 75%)

 

[AAKEE]

My full range at delivery was 358 mi, which is t I've read how the calendar degradation is modeled as proportional to sqrt(T) with T being time. Suppose the baseline degradation for an average person who charges to 80-90% is A * sqrt(T/T0). If I charge only up to 50% and reduce degradation by 2x, does this mean the degradation is

A/2 * sqrt(T/T0)

or

A * sqrt(0.5 T/T0) meaning the argument of the sqrt takes twice as long to hit 1

Calendar aging is plain [SOC x Temp x square root (time)]

From this chart, find the degradation from a specific SOC and temperature, for example 40% and 25C, looks like 2.8% or so.
To calculate degradation D at the specific time T (in months):

D = 2.8 x square root ( T/10)

I have used data from several charts like this, this is good but do not have lower temperatures, so I did find that in s couple of other charts and incorporated lower temperature. The degradation curve at 25C is converted into mathematics and the change in degradation due to colder or warmer is made into a correction factor from the 25C curve.
Lover temperature reduce the degradation less than higher temps increase it.

My own logged data put the average cell temp 10C above the average of the cars outer air temp probe.
My battery has been about 12C warmer than the annual average outside temperature. (Car is garaged at about +10C when we have winter).
For calculating the long term degradation the average cell temp should be used.

 

[AAKEE]

- Age in months of the car or purchase/manufacturing date.
20 months (well, 12 months since battery was replaced in April 22)

- Average temperature for the car during a year
18 degrees

- Regular charging SOC.
55% last six months. Previously 70%

-End of the day SOC, before next charge.
30%

-Time of day when the charging is set to start.
01:00 at home (25%), 10:00 at work (garage, 75%)

Was the battery remanufactured, or brand new?

For a brand new battery i get about 4.1% degradation.
first six months was 70 to 45% daily?
I guess the 75% times charging at work reset the nightly average?