What should my ideal charge percentage be?

[AAKEE]

Hi, sorry to bump this up, but after reading many pages I am not clear -- so many different recommendations. for an original P85 (battery replaced with a new one 5-6 years ago), what is the best method to prolong the battery?

Same as for most other EV’s.

For Panasonic NCA cells, keep 55% displayed SOC or any lower SOC.

You can see that by staying at or below 55% the cslendar shing is cut in half.
No problem charging to 100% when needed as long as you keep the average SOC low ( = reduce time above 55%)

I see references to 50-55%, but that means charge to that as max and how low is too low?

As long as you do not need more, charge to 50-55%. When ypu need more, charge what you need but charge ”late” to shorten the time above 55%.

In practical, your battery has already got 5-6 years of calendar aging and the rate reduces with time so probably quite low degradation rate now.
I would still use low SOC when possible to reduce the rate.

If anyone can summarize the current thinking I would greatly appreciate it.

 

[aerodyne]

I have a 2014 Model S P85D. I have always set the charge limit to 90%. A few times per year I will do longer trips and charge to 100%. I don't drive a lot - my car only has about 110,000 km in that time. My battery has lost about 5% if I can believe the car's range reporting. Initially I had 400km of range when the battery was at 100%, now I have about 380km. I don't baby the battery.

I baby the crap out of my 18 month old MS. Pack capacity was better than most, but just started to take a nosedive, now lost 5% since new and dropping fast.

I babied the last car too, 2015 85D. Highest range of any car on Teslafi. After I sold it, pack bricked with no warning at 64k miles while showing 260 miles of RR out of 270.

All the advice you see does not apply to a pack with 7 thousand cells, any one of which can fail and eventually brick the pack.

Just follow the recommendations from Tesla, 20-80 SoC and don't think there is some magic you can do to significantly affect pack health. That's the job of the BMS.

 

[NeuroRad]

I believe my battery is warrantied for 8 years. There’s near zero chance I’ll own the car outside the 4 year warranty period. Haven’t checked for any battery degradation nor will I. I stay charged at home 80%. I did lower from 85% after about a year of ownership after Tesla dropped the upper range to 80%.

 

[AAKEE]

I baby the crap out of my 18 month old MS. Pack capacity was better than most, but just started to take a nosedive, now lost 5% since new and dropping fast.

How did you do to babyvthe battery?

80% is about as same as 100% when it comes to degradation from time (calendar aging). In some cases 80% is even worse than 100%.

This is dsta from a 2016 or 2017 or so model S where the cells was taken from a almost new MS.

(That is not the only research data showing about the same thing. Instead do more or less all data show this).

Just follow the recommendations from Tesla, 20-80 SoC

I wouls say that there is no 80-20% advice from tesla.

They recently lowered the “daily” from “below 90” to 80% or below for most cars.

and don't think there is some magic you can do to significantly affect pack health. That's the job of the BMS.

The BMS can not stop or even manage the degradation from time (calendar aging). It is dependent on [time x temp x SOC].
The BMS can not stop the time, or change the temperature. And the owner is in charge or the SOC level.
The same is valid for cycles, the size of the cycle and between what numbers the cycle is placed is decided by the owner/driver.

And we can absolutely affect the degradation. My 2.5 year old M3P had about 1/3 of the degradation of other M3P 2021 at the same miles when I sold it.
Rangewise my car had 492 of 507km, thats 3%. The teslafi average at 66K km was about 460-465km, 465 = 8% degradation rangewise.

My MSP is at 98kWh after 8 months, where many other cars is down to ~95kWh, from 99.4 new value.

 

[AAKEE]

I believe my battery is warrantied for 8 years. There’s near zero chance I’ll own the car outside the 4 year warranty period. Haven’t checked for any battery degradation nor will I. I stay charged at home 80%. I did lower from 85% after about a year of ownership after Tesla dropped the upper range to 80%.

Still you use about the worst SOC when it comes to calendar aging.

I’ll post a few graphs from different research relort, that tested the calendar sging at different SOC.
Panasonic NCA (virtually model S/X cells), the two first pictures.

Actual model 3 cells taken out of a 2018 model 3.

Not Panasonic cells but a easy to read graph:

Again Panasonic NCA.

Calendar aging takes the very largest bite of our batteries the first fice years.
Cuclic aging is almost negligible for the most owners. 100K miles might cost 5% degradation but the first year only might cost 5% by being long time at around 70-90%.

Cyclic aging also wear less at lower SOC’s.
This is actual model 3 cells tested in 10% depth of discharge in different SOC regions:

The wear is slightly more between 5-25%, but as we still get 1000 Equivalent Full cycles for 15% loss we get about 400K km (250K mi) for 15% loss, thats about 0.75% cyclic loss each year.

Calendar aging, if the battery is most often at 70% or more, we loose about 5% the first year.

 

[12Pack]

I lobbied 80% with @AAKEE earlier in the thread, contacted the author of the papers they reference, and also contacted a peer reviewer of the material for the Journal of Electrochemical Society in which it was published (where I am a frequent author).

In conclusion, I now have created an automation to set the charge limit to 50% on Friday night (I have weekend cars) and back to 55% on Sunday night for the commuting week.

The joy of having the Plus version H battery on my MS is that this still gives me ~200 miles range. I am down about 70 bhp and have to make do with “only” 700, but driving in the UK is not about drag racing.

 

[aerodyne]

I generally agree with @AAKEE but I will suggest there are many possible pack level failure modes we don't know enough about yet, esp on the 2021+ Palladium Model S and Model X.

We also don't know what actions Tesla might take to mitigate those potential failures.

Case in point:. Air conditioning drain emptying on top of the pack pre 2015, causing corrosion and water in the pack. Also:

The infamous China fires, resulting we think, in the 2019 update causing the "Gates". Reduction in range, SuC charging speed, Regen, pumps running at 100%, etc.

For the old MS 85D car, stored in the middle of the range 55 to 62% suggested by the Elon Tweet circa 2018. Operated mainly 30-70% as mentioned in above tweet.

Never below 20%. 100% twice. 90% and above maybe 10 times. Avoided over 77% because the coolant pumps ran at 100%.

For the new car, never above 85, or below 22%. Store at 50% or lower. Per SMT, down from 99.4 KwH new to 91ish NFPafter only 18 months and 10k miles.

Both cars charged at home 7kw 90% of the time. Both cars in chill mode 99.998% percent of the time. Rarely full throttle, avoid max Regen.

Always garaged, AND covered, almost never outside in rain or cold temps. Highest storage temp 32 deg C, lowest 10 deg C

The guidance I get from Tesla is the battery bar turns yellow at 20%, and the charging graph says 80% or below for daily use. This is for the MS, the M3 may be different.

 

[AAKEE]

For the new car, never above 85, or below 22%. Store at 50% or lower. Per SMT, down from 99.4 KwH new to 91ish NFPafter only 18 months and 10k miles.

As degradation is rather predictable, I would guess that if the time above 55% is limited and you so not live on a very hot place, the calendar aging the first year would be around 3%, and 18 months would set you at about 3.7%, so lets say 4% then.

Your new S should have about 95kWh capacity.
I guess your BMS was not given the opportunity to see the battery capacity.
A few full charges, if you anyway are to drive a bit, and a few occasions with single number SOC might be good for the BMS calc. (I have not performed any BMS calibs myself, Im not worried about the degradation rate at low SOC).
So I would say that it is a steong possibility that the BMS is wrong.

Do you do any longer trips?
Mooking at the charging and resulting difference in SOC, or the re-estimation after a longer drive could tell quite much sbout the real capacity, without even doing a BMS calibration.

[Edit]I forgot: The different failure modes.
No we do not know anything about these.

Probably the battery manufacturing process produce cells with more even quality today than before.
And probably tesla did learn a lot about do’s and don’ts.

I guess the failure rate will be lower but newly introduced failure modes we can not protect us from in the most cases.
Water intrusion or broken BMB’s or contactor can happen to anybody.
Having less degradation before the failure should ensure I get a replacement battery with the same degradation or less.
I can not see a reason not to take care of the battery.

The main reason for me to write as much about batteries as I do is that I see a loth of myths or misinformation about lithium batteries. The mainstream has been more wrong than wright.
From all research about lithium batteries its not hard to draw the conclusion that as most chemistries behave similar and in modt cases the same chemistry behave very similar, and this had been going on for more than 10 years, the batteries will continue to behsve about the same. Solid Electrolyte Interphase will continue to grow, lithium plating will still happen etc.
Things will change slowly except for new inventions like Solid state batteries etc.

So it will be very safe to say that our Palladium batteries will have very similar degradation behaviour as the Panasonic NCA tested acfewcyears back. We can hope that the rate for SEI build up has been slightly reduced, and that the cells are less sensitive to lithium plating etc. But it is not wise to expect the things that are based on how the chemistry work will suddenly disappear.

One example is the LFP introduction into 3/Y. From my perspective it was clear that these batteries did not like 100% more than our Panasonic NCA in the model S/Y does.
One could never be 100% sure about it, but at least 97% sure that these batteries would degrade about the same as the LR/P Chemistries, and that the absolute main part would come from calendar aging.
This proved right, they mainly follow the expected degradation rate (from calendar aging).

 

[aerodyne]

As degradation is rather predictable, I would guess that if the time above 55% is limited and you so not live on a very hot place, the calendar aging the first year would be around 3%, and 18 months would set you at about 3.7%, so lets say 4% then.

Your new S should have about 95kWh capacity.
I guess your BMS was not given the opportunity to see the battery capacity.
A few full charges, if you anyway are to drive a bit, and a few occasions with single number SOC might be good for the BMS calc. (I have not performed any BMS calibs myself, Im not worried about the degradation rate at low SOC).
So I would say that it is a steong possibility that the BMS is wrong.

Do you do any longer trips?
Mooking at the charging and resulting difference in SOC, or the re-estimation after a longer drive could tell quite much sbout the real capacity, without even doing a BMS calibration.

I have done a few maybe 5, long trips using about 50-60%of the pack. Things were looking good until recently with 94ish KwH NFP. Currently 92.3 KwH NFP, 389 miles RR, per SMT.

It is what it is.

History from Teslafi:

 

[AAKEE]

I have done a few maybe 5, long trips using about 50-60%of the pack. Things were looking good until recently with 94ish KwH NFP. Currently 92.3 KwH NFP, 389 miles RR, per SMT.

It is what it is.

History from Teslafi:

I’d say the BMS is off.

My M3P that ended up with 492 out of 507km on thr last full charge (the average among the other cars was 460-465km at the same time) had a real hi kup by the BMS.

You can see the progressive large range drop leading to the lowest range about 49-50k km? The NFP that had been almost never below 80kWh dropped do about 75kWh (74.7 lowest point).

At the same time I made a 100-0% drive (actually divided to two drives), and the delivered energy was 75kWh, plus the Bms estimated 0.4kWh left until 0% displayed, plus the buffer. In the end, the capacity was proven to be about 79 kWh at that point.
It later climbed and stayed around 79 for quite long.

There is a few ways to use the recalculstion after a drive to calculate the true Capacity.
I already have detailed it in a few posts earlier so I’ll find one of these and post a link.
The short description: when driving the (displayed) SOC is calculated as it can not be measured. When parking, the BMS get to measure the cell voltage at less load, and when the car has slept for a while the true SOC can be measured via tge OCV.

After a longer drive, if the SOC goes up, the BMS has an underestimation, and vice versa.
After a charge, if the SOC goes down the car has an underestimation.
(In both cases we need to know that the SOC was correct before the drive (by letting the car measure the OCV).

I’ll be back later.
My MSP started the life with NFP = 95.3 kWh a few days after delivery.
I was not realky worried as the car had ben on low SOC since buikd date (not allowed to do offshore shippings with high SOC anymore).
I was able to calculate the capacity to 98-98.2 kWh after the first longer drive since I got the SMT in place. Of course, after a few weeks the BMS had also found the capacity to be 98.2kWh.

You can see the “recovery” of SOC, a sign of the BMS underestimating the capacity.
This was just a grab from teslafi, we get the best data from SMT about this.

 

[AAKEE]

Here is one of the posts explaining how the calc is done:
Range Loss Over Time, What Can Be Expected, Efficiency, How to Maintain Battery Health

 

[tesluv108]

I baby the crap out of my 18 month old MS. Pack capacity was better than most, but just started to take a nosedive, now lost 5% since new and dropping fast.

I babied the last car too, 2015 85D. Highest range of any car on Teslafi. After I sold it, pack bricked with no warning at 64k miles while showing 260 miles of RR out of 270.

All the advice you see does not apply to a pack with 7 thousand cells, any one of which can fail and eventually brick the pack.

Just follow the recommendations from Tesla, 20-80 SoC and don't think there is some magic you can do to significantly affect pack health. That's the job of the BMS.

I’m guessing the new owner contacted you , not so happily about the bricked pack? How did you know the old battery was ruined? If you were in the contact loop what did Tesla charge to fix it?

 

[tesluv108]

Here is one of the posts explaining how the calc is done:
Range Loss Over Time, What Can Be Expected, Efficiency, How to Maintain Battery Health

Have any studies been published yet on Model 3 LFP packs? We have one with under 200 miles on it. The “guidance” is to charge to 100% once a week. This still seems strange to me but looking for insights into this new way of thinking with new lfp technology

 

[AAKEE]

Have any studies been published yet on Model 3 LFP packs? We have one with under 200 miles on it. The “guidance” is to charge to 100% once a week. This still seems strange to me but looking for insights into this new way of thinking with new lfp technology

The 100% once a week is only to reset the BMS so it knows it is 100%.

LFP Batteries have very flat voltage curve (the state of charge is defines as the voltage in the cell at rest with no load).

It is so flat that the BMS can not safely measure the voltage and say X % so the state of charge between the 100% charges is depending on the BMS to calculate energy in- and out. In the end this might differ from the true SOC, so we have a risk of overestimating the SOC and getting stranded.
The car imcreases the buffer below 0% on the screen to have an extra margin, but this reduces the energy above 0% thus reducing the range.

A full charge resets the energy counting to a known correct level (100%).

For the sake of the battery there is no big difference from before.
As already known by researchers 100% is not really worse than 80%. In some cases 80% is even worse than 100%.

This is a chart from ~5 year back.
Problably still valid, they might have ben able to reduce the rate slightly but the main curvature is probably about the same.

If we use logged data, the degradation from LFP cars seems quite alike the LR and P cars. This is because the calendar aging is responsible for the most part of the degradation and they are about thecsame.
Cyclic aging on a 60K km LFP should be close nil.

LFP 21, about 6% at 60K km.
Most of it have to be calendar aging.

This is model 3 P 21 with the 82kWh NCA battery. Its slightly more sensitive then most other batteries but still about 7.5% loss at 60K

 

[zivac87]

I am extremely happy with my battery life so far. Battery is built on Oct 2022.
After following few simple advices from @AAKEE this is health report so far. As you can see avg consumption car is driven pretty hard. Summer was extremely warm. My only rule was stay between 5%-55% when ending my drives for the day, but when i want to drive charge it as late as possible to whatever SoC I want to get max power from car (usually 75-95% SoC)

 

[aerodyne]

I’m guessing the new owner contacted you , not so happily about the bricked pack? How did you know the old battery was ruined? If you were in the contact loop what did Tesla charge to fix it?

Yes
Yes
Covered under warranty, but a reman pack installed.

 

[peakay]

Lower SoC is better overall for degradation. Set your charge limit to as low as you can while still meeting your usage needs. Below ~55% is best if that works for you.

There is no “too low” SoC unless it’s fully dead, which is damaging to the battery cells. But the car will turn off before it gets to that point. There is no harm in running to below 20%, 10%, or even 5% if that’s what you’re asking.

Ok, I thought keeping it between 80 and 20% was the rule as running it down too far is bad?

So 55% max, daily charging and just keeping it off 0% is better?

 

[geordi]

Ok, I thought keeping it between 80 and 20% was the rule as running it down too far is bad?

So 55% max, daily charging and just keeping it off 0% is better?

I'd keep the lower limit a bit higher, 10% arriving at home for a charge is a good thing. Whatever your daily driving needs, aim for the max percentage to give you that, and as low as feasible.

 

[E90alex]

Ok, I thought keeping it between 80 and 20% was the rule as running it down too far is bad?

So 55% max, daily charging and just keeping it off 0% is better?

There is no harm in running it down unless you run it down until the car turns off and leave it there for long periods. There is also no harm in charging higher as long as you’re not leaving it there for long periods.

Exactly what you should charge to daily heavily depends on your own usage needs and comfort level. But generally the lower the SoC is when the car is sitting the better, with the biggest benefit if you can keep the average daily charge under 55%. That doesn’t necessarily mean do not charge above 55% daily.

For example if you had a long commute that uses 20% of your battery each way, you could charge up to 70-75% right before you leave. That way you arrive at work with 50-55% where it sits for 8 hours. Then you drive back home where it sits at 20-25% for 8-10 hours before being charged again to 70-75% right before you leave the next day. The car spends vast majority of the day below 55%.

Contrast this to someone that charges to 80% daily because it was “recommended” but only uses 5-10% a day and immediately charges back to 80% when getting home. The car spends most of its life at 70-80% and will likely have a more degraded battery over time compared to the previous example.

Personally, a 50% charge limit works more than enough for me and would even last me 2-3 days if I forgot to charge, but like I said your YMMV.

First and foremost charge to what lets you comfortably use the car without inconvenience or anxiety. If it happens to be 50% then that’s great. If it needs to be 80% then that’s fine too. But take into consideration your usage pattern and lower the limit if possible if you want to maximize battery health.

 

[asterothe]

How about being in a short band? From 55% to 40% work days? Never going too low?