Hearing super conflicting advice about battery charging levels

[jjrandorin]

Roger that, @jjrandorin. It's a dilemma, for sure. You see it on every single Internet forum, of every conceivable interest. Certain topics just naturally beg to be talked about... and so they pop up daily, if not hourly. If you pull them all into a single, coherent thread, that thread soon becomes massive. Pointing someone to it is like pointing them to an encyclopedia. I read through the entire "How I Recovered Half of my Battery's Lost Capacity" thread a few months ago and it took many hours. Diving into one of those is kind of like diving into a book... it requires a modicum of commitment.

Most people don't want to spend that much time and energy.

The flip side is you just shrug and move on, knowing that the same subjects are going to keep coming up in countless, very similar, threads.

I'm not sure there's a good answer. Maybe sticky those threads you think hold particular value. Otherwise, just accept that trying to bring coherence to the countless threads people create is kinda like pissing in the wind.

Thanks for all you do here. Both for your steady hand as moderator as well as your cogent, thoughtful replies in trying to help people.

I agree with you on what you are saying here in this post on the fact that there isnt a really great answer, other than stickying threads of particular interest, and / or "just accepting it" (which is what I am trying to do right now, lol.

Thank you for the compliment, I appreciate it.

 

[SageBrush]

The navel gazing and constant dissection of this topic is so very over the top.

Indeed, although I am not cavalier with my batteries. I avoid the 10% worse behavior to garner 90% of the possible benefit of navel gazing.
That means avoiding letting the battery cook at 100% SoC by only charging to 100% shortly before use. In my day to day I don't charge to 100%.

End of drama

 

[Steve446]

As you should. The navel gazing and constant dissection of this topic is so very over the top.

As a “bad” Tesla owner that has constantly flogged his battery by doing the things people in this thread think are catastrophic like:

* Charging to more than 0% (90% even, or - gasp - 100% over a hundred times!)
* Letting my car exist where it is hot
* Liberally using superchargers

… my observed degradation over 5.5 years and 150,000 miles of actual driving (more than almost anyone participating in this thread will actually drive in their cars) is quite in-line with other cars of my vintage.

Just plug it in and drive.

That's a fair comment and if that approach works for you and probably a fair few other users then that's great. Plug in, forget the details and enjoy, nobody will criticize!

This part of the forum does attract many questions on battery topics and I guess a reasonable aim would be to allow those wanting more information to be able to find it. The question remains how to do that? The moderator and a few knowledgeable members do that regularly but I fear some of the more banal but well meaning questions may wear them out before their batteries degrade thanks to all of them!

 

[AAKEE]

Indeed, although I am not cavalier with my batteries. I avoid the 10% worse behavior to garner 90% of the possible benefit of navel gazing.
That means avoiding letting the battery cook at 100% SoC by only charging to 100% shortly before use. In my day to day I don't charge to 100%.

End of drama

I see regularly that myths about battery degradation keeps people having hard understanding why they lost quite much of the battery capacity. Some refer to the battery lottery, and others to the myth that keeping 80% is babysitting the battery.

At the same time we have people that doesnt care about the lost capacity.
Theres both the ones that really do not care, and thats a fine thing.
Buth we also have the people that doesnt seem to care today and that got cought by some of the myths, that probably will male them wake up some day abd find that they actually do care.

One myth is the ”cook at 100%”.
While I actually reduce the time at high SOC I know that for resonable temperatures 100% doesnt cook the battery much more than 80% SOC.
Tesla tell us not to use 90-100% daily, but they never talk about not leaving the car at 100% for longer times.

It is very probable that the stay below 90% for daily is because cycles with high SOC wear quite much. Using 90% instead of 100% about doubles the cycles a battery can do.

All research shows that there is a not very big difference between 80% and 100%. In some cases ( some 2170 NCA) the eorst degradation actually happend at 80% SOC).
I wouldnt recomment charging to 100% and leave the car like that for long, but it is not as bad as people often write.

 

[Apprunner]

I see regularly that myths about battery degradation keeps people having hard understanding why they lost quite much of the battery capacity. Some refer to the battery lottery, and others to the myth that keeping 80% is babysitting the battery.

At the same time we have people that doesnt care about the lost capacity.
Theres both the ones that really do not care, and thats a fine thing.
Buth we also have the people that doesnt seem to care today and that got cought by some of the myths, that probably will male them wake up some day abd find that they actually do care.

One myth is the ”cook at 100%”.
While I actually reduce the time at high SOC I know that for resonable temperatures 100% doesnt cook the battery much more than 80% SOC.
Tesla tell us not to use 90-100% daily, but they never talk about not leaving the car at 100% for longer times.

It is very probable that the stay below 90% for daily is because cycles with high SOC wear quite much. Using 90% instead of 100% about doubles the cycles a battery can do.

All research shows that there is a not very big difference between 80% and 100%. In some cases ( some 2170 NCA) the eorst degradation actually happend at 80% SOC).
I wouldnt recomment charging to 100% and leave the car like that for long, but it is not as bad as people often write.

You are saying that 90% doubles the cycles vs. 100% but you are saying that there is little difference between 80% and 100%? A little confusing.

 

[AAKEE]

You are saying that 90% doubles the cycles vs. 100% but you are saying that there is little difference between 80% and 100%? A little confusing.

Perhaps I wasnt clear enough.

We need to separate between calendar aging and cyclic aging.

If we look at Cyclic aging will have the most wear for cycles 100-0%. In this case there is not a big difference between 100 to 0% and 100 to 10%, because low SOC cause very little wear.
Between 100 to 0% and 90 to 0% the battery can hold up for much more cycles.

A limit at 90% for daily charging will have a good effect on longevity (compared to 100% daily).
(But limiting the charge further down increase the life many times more)

This picture from a research report show that low SOC is not dangerous. The lower the SOC, the lower the cyclic aging.

High SOC causes much more cyclic aging than low SOC.

There is a term often used, deep cycles that could give one the idea that deep = low SOC = bad because of the deep discharge but this is not really true.
Big cycles is worse and not because of going down to low SOC but because they start with a high SOC value, and high SOC cycles causes worse wear.

Small cycles is better, low SOC diring the cycles is better.

If we look at calendar aging, the basic principle is ”the higher the SOC the higher the degradation”. There is a step which increase degradation if SOC is above, and reduce is SOC is below that step. For NCA that step in between 55-60% SOC.
slightly higher values is valid for NMC and LFP.

There is a lot of research showing the same thing so this is confirmed.

Another research report, for calendar aging.
The worst degradation happens at around 80%

Here’s yet another report, 2170 NCA cells of a brand that could not be revealed. ( there is not that many manufacturers of 2170 NCA cells).

We can clearly see that 100% is not the worst case scenario for degradation.

The lower the SOC, the lower the calendar aging. The lower the temperature, the lower the calendar aging.

 

[Steve446]

Perhaps I wasnt clear enough.

We need to separate between calendar aging and cyclic aging.

If we look at Cyclic aging will have the most wear for cycles 100-0%. In this case there is not a big difference between 100 to 0% and 100 to 10%, because low SOC cause very little wear.
Between 100 to 0% and 90 to 0% the battery can hold up for much more cycles.

A limit at 90% for daily charging will have a good effect on longevity (compared to 100% daily).
(But limiting the charge further down increase the life many times more)

View attachment 830695
This picture from a research report show that low SOC is not dangerous. The lower the SOC, the lower the cyclic aging.

High SOC causes much more cyclic aging than low SOC.

There is a term often used, deep cycles that could give one the idea that deep = low SOC = bad because of the deep discharge but this is not really true.
Big cycles is worse and not because of going down to low SOC but because they start with a high SOC value, and high SOC cycles causes worse wear.

Small cycles is better, low SOC diring the cycles is better.

If we look at calendar aging, the basic principle is ”the higher the SOC the higher the degradation”. There is a step which increase degradation if SOC is above, and reduce is SOC is below that step. For NCA that step in between 55-60% SOC.
slightly higher values is valid for NMC and LFP.

View attachment 830691

There is a lot of research showing the same thing so this is confirmed.

Another research report, for calendar aging.
The worst degradation happens at around 80%

View attachment 830692

Here’s yet another report, 2170 NCA cells of a brand that could not be revealed. ( there is not that many manufacturers of 2170 NCA cells).
View attachment 830694
We can clearly see that 100% is not the worst case scenario for degradation.

The lower the SOC, the lower the calendar aging. The lower the temperature, the lower the calendar aging.

Can you publish the reference for the "Figure 1 SoC dependant capacity fade after 18 months..." ? Maybe you already did? I assume that the three operating window are the authors recommendations?. Thank you for any help.

 

[AAKEE]

Can you publish the reference for the "Figure 1 SoC dependant capacity fade after 18 months..." ? Maybe you already did?

I have, but not in this thread.
here you go:
https://www.researchgate.net/profil...erative-Braking.pdf?origin=publication_detail

I assume that the three operating window are the authors recommendations?. Thank you for any help.

The researchers did choose three different SOC ranges, with the same depth of discharge, 25% ( ie the same miles driven).

These are not chosen as recommended SOC at all but as three examples of possible SOC ranges that can be used for the same driving habits.

The research involves finding the benifit / impact of regenerative braking, but tell us readers a lot more than that.

One important note: while this is a quite good report each research had its limitation from the test report. Sometimes the conclusions is faulty due to the test setup.
This report is good, but it can be easy to read a report wrong and thereby draw own faulty conclusions.

The low SOC range causes the least degradation. But this is when compared to the medium and high SOC range.
It is important to understand that this do not mean that the low SOC range they did choose is the best of any SOC range. Its better than medium and high in this test but to 99% certainty a even lower SOC range would be better.
It would also be better to use even smaller dept of discharge.

There is a lot of misinformation and misunderstandings (”myths”) about batteries.
Batteryuniversity.com is a good idea. Unfortunately they have some information that is not correct, and some information that is correct but too hard for the average person to understand correct, causing people to understand it wrong which then leads to battery myths.

This is a cut out of batteryuniversity.

I did read the test report from which they took the picture. First fault, the graph show DST, each single cycle. You can not see directly wich SOC that will give you the most miles driven for a certain degradation.
You need to multiply the cycles with the dept of discharge to see the degradstion ”per mile driven”.
Due to this the normal graph shows FCE, Full equivalent cycles. One 100-0% is one FCE, but you need 10 cycles(DST) to reach one FCE if you use 75-65%, 10% DoD. That means that you directly can relate to how many miles the car will go for a certain degradation.

Next ”fault” is that there is no low SOC cycles. The test mainly focus on different dept of discharge and not where the SOC range is placed.
Reading that information and looking at the picture have made many think that 75-65% is the best SOC range with the lowest degradation. I have seen many people throwing references to that in forums with quite clear ”I know this because I did read it at batteryuni…”
For that picture, 75-65% looks best because the way to present it fools the reader. Also doesnt it say that the research test did not even try the low SOC we know is very good for longevity.

 

[Steve446]

I have, but not in this thread.
here you go:
https://www.researchgate.net/profil...erative-Braking.pdf?origin=publication_detail


The researchers did choose three different SOC ranges, with the same depth of discharge, 25% ( ie the same miles driven).

These are not chosen as recommended SOC at all but as three examples of possible SOC ranges that can be used for the same driving habits.

The research involves finding the benifit / impact of regenerative braking, but tell us readers a lot more than that.

One important note: while this is a quite good report each research had its limitation from the test report. Sometimes the conclusions is faulty due to the test setup.
This report is good, but it can be easy to read a report wrong and thereby draw own faulty conclusions.

The low SOC range causes the least degradation. But this is when compared to the medium and high SOC range.
It is important to understand that this do not mean that the low SOC range they did choose is the best of any SOC range. Its better than medium and high in this test but to 99% certainty a even lower SOC range would be better.
It would also be better to use even smaller dept of discharge.

There is a lot of misinformation and misunderstandings (”myths”) about batteries.
Batteryuniversity.com is a good idea. Unfortunately they have some information that is not correct, and some information that is correct but too hard for the average person to understand correct, causing people to understand it wrong which then leads to battery myths.



View attachment 830907

This is a cut out of batteryuniversity.

I did read the test report from which they took the picture. First fault, the graph show DST, each single cycle. You can not see directly wich SOC that will give you the most miles driven for a certain degradation.
You need to multiply the cycles with the dept of discharge to see the degradstion ”per mile driven”.
Due to this the normal graph shows FCE, Full equivalent cycles. One 100-0% is one FCE, but you need 10 cycles(DST) to reach one FCE if you use 75-65%, 10% DoD. That means that you directly can relate to how many miles the car will go for a certain degradation.

Next ”fault” is that there is no low SOC cycles. The test mainly focus on different dept of discharge and not where the SOC range is placed.
Reading that information and looking at the picture have made many think that 75-65% is the best SOC range with the lowest degradation. I have seen many people throwing references to that in forums with quite clear ”I know this because I did read it at batteryuni…”
For that picture, 75-65% looks best because the way to present it fools the reader. Also doesnt it say that the research test did not even try the low SOC we know is very good for longevity.

Thank you, that's very clear for me and nice arguments for low SoC. I look forward to reading the article too.

 

[AAKEE]

This one is good: http://mediatum.ub.tum.de/doc/1355829/document.pdf

Almost 200 pages though. But it covers most of the important parts so a good read.

Another calendar aging view. The ”grouping” att 55% and below or 60% and above is easy to see.

SOC vs calendar aging and temperature.
3.45V is about 20-25% SOC or so.
3.70V is about 45-50% SOC
4.10V is about 90% SOC.

 

[AAKEE]

This is one one many very good reports from Evelina Wikner, now I think she is post doc in lithium batteries at Chalmers in Sweden:
https://research.chalmers.se/publication/512004/file/512004_Fulltext.pdf

We can see that the lower SOC during the cycles, the longer the life. Cells cycled between 0-10% had the longest life. To compare, use the same C-number(it is the rate of discharege and charge) between different SOC-ranges.
Cell number #20 was cycled at 2C and did 10000 FCE (thats 100.000 0-10% cycles!!!) and still kept within 95% of the capacity: less than 5% degradation.
Cells cycled between 40-50% lost more than 5% in 1000FCE, and did loose about 20% after 5000 cycles.

The ”regular” 60-70% or 70-80% used by many Teslas showed much more degradation than the ones mentioned above.

0-10% SOC during cycles almost give the cells infinite life.

The cells cycled 80-90% lost a lot, despite ”shallow” cycles with only 10% DoD.
The cell cycled 0-90% initially followed the 80-90% cell closely, despita 0-90% beeing larege cycles.
We clearly see that the part of each cycle ”deep down” at low SOC do not cause much wear.
It is the high SOC part of the cycle that takes the big bite of the battery.

One way to think about the lithium ion battery is lika a ballon. No stress when deflated, and the more we inflate, the more the stress.

 

[AAKEE]

Here is a myth that I have come across a lot of times, and each time I have a hard time to overcome that fact that batteryuniversity did write that on their page. ”Because it is written there, it must be true”. But it is very much not true.

The statment seem to origin from NASA, and their selected optimum voltage( = SOC) in a satellite or something.
It is, of course a compromise of different things. 3.92V/cell is not the optimum SOC for any EV.
It might be the optimum SOC for some odd EV owner needing more SOC daily.

It is a higher SOC than about 55% (3.92V should be around 65% or so in a Tesla with a NCA battery so it will be above the step that increases calendar aging(that is located between 55-60% SOC).

”Optimum SOC” would be to charge to that level that gives the needed range between charging sessions, with a margin that do not cause the stomach tro go bad. Not more, if we look at battery life.
(As always, if one do not care about trying to keep the battery degradation low= 90% will be fine).

The lower the SOC, the lower the calendar aging.
The lower the temp, the lower the calendar aging.

The lower the SOC, the lower the cyclic aging.
The smaller Depth of Discharge, the smaller the cyclic aging.

Litium batteries do not suffer from going below 20% SOC, in fact they like it there very much all the way down to 0% SOC (which is NOT overdischarged by the way).

Comparison for cycles should be done with [FCE, Full equivalent cycles] as we then can see directly which cycling option will deliver the most amph hours or miles driven in a car.

 

[Steve446]

Here is a myth that I have come across a lot of times, and each time I have a hard time to overcome that fact that batteryuniversity did write that on their page. ”Because it is written there, it must be true”. But it is very much not true.


View attachment 831020

The statment seem to origin from NASA, and their selected optimum voltage( = SOC) in a satellite or something.
It is, of course a compromise of different things. 3.92V/cell is not the optimum SOC for any EV.
It might be the optimum SOC for some odd EV owner needing more SOC daily.

It is a higher SOC than about 55% (3.92V should be around 65% or so in a Tesla with a NCA battery so it will be above the step that increases calendar aging(that is located between 55-60% SOC).

”Optimum SOC” would be to charge to that level that gives the needed range between charging sessions, with a margin that do not cause the stomach tro go bad. Not more, if we look at battery life.
(As always, if one do not care about trying to keep the battery degradation low= 90% will be fine).

The lower the SOC, the lower the calendar aging.
The lower the temp, the lower the calendar aging.

The lower the SOC, the lower the cyclic aging.
The smaller Depth of Discharge, the smaller the cyclic aging.

Litium batteries do not suffer from going below 20% SOC, in fact they like it there very much all the way down to 0% SOC (which is NOT overdischarged by the way).

Comparison for cycles should be done with [FCE, Full equivalent cycles] as we then can see directly which cycling option will deliver the most amph hours or miles driven in a car.

Wow this is great stuff, really! Good reading for the coming weekend Maybe questions later. Thank you.

 

[Unclepaul]

Practical rule of thumb is to charge your battery between 20 and 80%. This will be healthy for your battery, but still give you a sufficient charge for daily travels.

If you are planning a long trip, then charge your batter to 100% before leaving, and run it down as far as possible (within your comfort level) and get a quick charge when necessary as your battery will charge really fast when only about 5-10% remaining. Filling your battery is fastest at a low state of charge. Do not fill to 100% as that takes forever on the road.

Don't look for absolute numbers, and obsess over that last 1%. Life is to short, and anxiety will only make it shorter.

 

[Physicslawyer]

I can't find the post, but there was a bunch of data posted by the developer of the Stats app. Basically it said that all the cars they monitored that there didn't seem to be any difference on the battery degradation based on the charge level. For example cars that were frequently charged to 90% or had high supercharger use didn't seem to have any difference than those like me who keep it about 50% other than when needed. This is counter to everything I have read on the subject. I'll see if I can find the post and reference it here.

I have been testing this. I got a MYP with 4 miles to start. It now has close to 13k miles and has been supercharged exclusively, mostly to 90%, sometimes to 100% (once a week). The range went from 303 to 285-287 depending on the day, about a 6% drop in 3.5 months.

 

[DrChaos]

I can't find the post, but there was a bunch of data posted by the developer of the Stats app. Basically it said that all the cars they monitored that there didn't seem to be any difference on the battery degradation based on the charge level. For example cars that were frequently charged to 90% or had high supercharger use didn't seem to have any difference than those like me who keep it about 50% other than when needed. This is counter to everything I have read on the subject. I'll see if I can find the post and reference it here.

You can see on that report that in that dataset nearly all users set their ending SOC above 70% and most were between 80 and 95%.

The size of the dataset for < 60% SOC max is unfortunately minuscule. Even though the Tesla UI actually shows the 'Daily' limit to be between 50% and 90%, too many people think it is recommended to be at 90%. I don't think the effects of lower SoC max could be seen easily. I

I also don't like the analysis of the 'Min and Max' SOC (presumably lifetime) per car vs degradation. That's going to be very noisy and uninformative, other than saying occasional 'min' and 'max' excursions don't make much of a difference (which we know already). The analysis isn't well tuned to the physical parameters which appear to be important from the chemistry research literature. Instead I would try to get a proxy of average state of charge through time, and length of time, and correlate that with capacity loss. Given the statistics of the dataset, there may be few cars with an average SOC low enough to see clearly the effects (lower calendar and cyclic degradation) implied by the battery research.

In a nutshell, I still believe the battery research literature more and will limit my max SOC limit to 50% for normal daily use.

 

[DayTrippin]

In a nutshell, I still believe the battery research literature more and will limit my max SOC limit to 50% for normal daily use.

That is exactly what I believe and the approach I am following as well.

 

[KenC]

You can see on that report that in that dataset nearly all users set their ending SOC above 70% and most were between 80 and 95%.

The size of the dataset for < 60% SOC max is unfortunately minuscule. Even though the Tesla UI actually shows the 'Daily' limit to be between 50% and 90%, too many people think it is recommended to be at 90%. I don't think the effects of lower SoC max could be seen easily. I

I also don't like the analysis of the 'Min and Max' SOC (presumably lifetime) per car vs degradation. That's going to be very noisy and uninformative, other than saying occasional 'min' and 'max' excursions don't make much of a difference (which we know already). The analysis isn't well tuned to the physical parameters which appear to be important from the chemistry research literature. Instead I would try to get a proxy of average state of charge through time, and length of time, and correlate that with capacity loss. Given the statistics of the dataset, there may be few cars with an average SOC low enough to see clearly the effects (lower calendar and cyclic degradation) implied by the battery research.

In a nutshell, I still believe the battery research literature more and will limit my max SOC limit to 50% for normal daily use.

I was surprised by the data. Anyhow, the Stats developer, in the twitter thread, decided to adopt my charging strategy since his daily commute is similar to mine. I drive about 30 miles, and charge between 50% and 60% SOC.

 

[roblab]

I never stress about charging, and I'll probably live longer because of it. I plug my car in when it's in the garage, and it charges to 90%, so it's usually sitting at 90% charge unless I'm driving somewhere.

What's the big deal? Are people worried that their battery might die sooner? If so, so what? There are new and refurbished batteries available for a decent price if it comes to that, but with the battery warranty set at 8 years, most of us will have sold our cars and moved up to a newer model. That's my plan, anyway. Have you seen the new Model S?

When the vast majority of owners pay zero attention to charging and figure the car will do what it's programmed to do, in a few years there will be a lot of used Teslas for sale which have never had any special charging care and will operate just as well as those that have. In other words, some people just want to worry about nothing. Just drive your car and enjoy it and charge it in the most convenient way you can arrange.

 

[DayTrippin]

I am a performance addict and I try to keep my addiction in check. One of my original concern was the loss of performance at lower SoC. I've found my '22 MS LR to still be very quick even at less than 50% SoC. Case in point I ran a Challenger Hellcat Redeye with about 47% charge level. It wasn't planned, just picked up some food for my wife and they guy was basically stalking me to run. It wasn't even close even with the fairly low SoC from either a stop or a roll.

Even at 50% SoC on my S, I can get an easy 160 miles at normal speeds I drive with the AC on. In my 3, it is definitely enough less that it can be concerning. At least with such a low SoC on either car, a quick 5 min stop can pump in enough range to normally get me home with either car.

I am watching the SoC at lot more closely than someone like @roblab because I also have to deal with pretty high temps as well. I can't see having the SoC @ 90% and have 105-110F temps in the shade as well. I can't do anything about the high temps but I can control the SoC. That also helps with cyclic aging as well and typically recharge even when I've just put 20-30 miles on the car unless it is super hot out.

I plan on keeping my car for a long time unless something happens to it which is always a possibility.