Healthiest State of Charge at Rest?

In theory, immediate charge to 50% and later charge to a higher percent... but.. in practice, simply limiting top level of charge to somewhere below 90% gets you the vast majority of the benefit. The "50 now, more later" is quite a bit of scheduling and worry, for an extremely small gain. And I mean REALLY small. Very theoretical. As compared to 90% charge.

The storage mode in the Roadster maintains the SOC around 20%. It's fine to sit at that level.

Tesla says to keep it plugged in when possible. No need to obsess over this and not routinely charge to at least 80% if not 90%, unless you're letting it sit for weeks. Let the battery management system manage the battery.

A battery is perfectly happy sitting at a lower SOC than 50%. The reason why it's not recommended is because it reduces the utility of the car since there is lower available range and power (see the P85 threads complaining about performance at lower SOCs).

The only probable negative is if you purposefully try to reach a lower SOC, you may actually end up with a deeper DOD (which is worse for the battery). For example, instead of charging from 50-75%SOC (DOD = 25%) for two trips, you let the car reach 25% SOC before charging, making your cycle depth 25-75%SOC (50% DOD). The former may actually be better for the battery, despite higher average SOC, because of the lower DOD.

This factor is probably also why Tesla recommends just leaving the car plugged whenever it's possible.

There is also higher risk of over-discharge (easier to run the battery empty and permanently kill it). The latter concern was real in the Roadster (which can be "bricked"), but the Model S supposedly can't be "bricked".

Long story short: Lithium ion batteries don't like sitting at high temperatures at high SOC for long periods of time, and they don't like to be deeply cycled (high variance between maximum and minimum SOC). Those are really the main things to keep in mind.

A caveat is if you never hit high SOCs or deeply cycle, the battery might more easily get out of balance or the capacity meter inaccurate, which will make it appear to lose significant capacity (although it will come back after being balanced/calibrated).

Low state of charge seems to be perfectly fine for Lithium batteries. The recommendation for the Roadster to store at 20% shows that. Lowe state of charge is only bad for the battery when you draw lots of power from it. The voltage drops when power is drawn and since the voltage is already lower when the state of charge is low there is a risk that it drops below a certain limit where it becomes bad for the battery. But just sitting at low state of charge is fine.

If your daily usage is 30% (from 50 down to 20) I would probably charge up to 60% and end at 30% instead of 50 to 20. Quite honestly, if you are driving at those levels your batteries will age very slow, so don't worry about trying to optimize even further and thus making it too complicated.

0-20%.

Id actually recommend completely dead for a Model S.
3.3V / cell is about ideal; and the model S will read 0 rated range at that level.
I have module sitting like that for weeks/months, and no loss of voltage despite power to the bms board.

Precisely. The reason for 50% is Tesla would rather have a more user friendly, consumer friendly battery (ie, ready to drive at all times) that lasts 8-12 years, then a battery that requires baby-sitting and lasts 20+ years.

okashira said:

Precisely. The reason for 50% is Tesla would rather have a more user friendly, consumer friendly battery (ie, ready to drive at all times) that lasts 8-12 years, then a battery that requires baby-sitting and lasts 20+ years.

Click to expand...

More like: Friendly and 3 to 5 years, vs baby sitting and 3 to 5 years. Seriously. No matter what state of charge, Li batteries degrade just sitting. Storage temperature will make much MUCH more difference than 50% state of charge vs. 30% state of charge. No published 18650 cell has a "shelf life" over 3 years, even with 0 cycles and stored at an ideal state (which, BTW, is 40%, per manufacturers data sheets, not 30%). Of course, Tesla/Panasonic are not publishing the EXACT chemistry, we don't know if "100%" according to the car is 4.1 V/Cell, 4.15. 4.05, and more... so don't take ANY numbers too literally... nonetheless, physical age life of more than 5 years would be totally unexpected.

Bottom line: Don't worry too much. Set charge to 90% except just before travel, and let the Tesla manage its batteries.

Danal said:

More like: Friendly and 3 to 5 years, vs baby sitting and 3 to 5 years. Seriously. No matter what state of charge, Li batteries degrade just sitting. Storage temperature will make much MUCH more difference than 50% state of charge vs. 30% state of charge. No published 18650 cell has a "shelf life" over 3 years, even with 0 cycles and stored at an ideal state (which, BTW, is 40%, per manufacturers data sheets, not 30%). Of course, Tesla/Panasonic are not publishing the EXACT chemistry, we don't know if "100%" according to the car is 4.1 V/Cell, 4.15. 4.05, and more... so don't take ANY numbers too literally... nonetheless, physical age life of more than 5 years would be totally unexpected.

Bottom line: Don't worry too much. Set charge to 90% except just before travel, and let the Tesla manage its batteries.

Click to expand...

I'm sorry but you are simply misinformed. There is as much as an order of magnitude difference (and more) in degradation rate at 90% SOC vs 0-20% SOC.
In fact, a Model S cell stored at 10% SOC at 60°C will retain more capacity then one stored at 90% SOC and 25°C.

That said, the cells are designed for excellent durability in all conditions, so even at full charge, they will outlast almost any other chemistry.

My understanding is that degradation is both due to calendar life and charge/discharge cycles. If you baby the battery with regards to number of cycles, deep discharges, high voltages etc. then it will degrade anyway due to ageing (calendar life). These two effect do not add up, i.e. which ever effect your battery will degrade in a predicatable fashion due to ageing and this happens no matter how much you baby it.

okashira said:

I'm sorry but you are simply misinformed. There is as much as an order of magnitude difference (and more) in degradation rate at 90% SOC vs 0-20% SOC.
In fact, a Model S cell stored at 10% SOC at 60°C will retain more capacity then one stored at 90% SOC and 25°C.

That said, the cells are designed for excellent durability in all conditions, so even at full charge, they will outlast almost any other chemistry.

Click to expand...

Link please? I'd like to see a data sheet or white paper from either Tesla or Panasonic that states long term storage numbers. In fact, I'd like to see the exact cell variant/chemistry in use.

Thanks!

- - - Updated - - -

Johan said:

My understanding is that degradation is both due to calendar life and charge/discharge cycles. If you baby the battery with regards to number of cycles, deep discharges, high voltages etc. then it will degrade anyway due to ageing (calendar life). These two effect do not add up, i.e. which ever effect your battery will degrade in a predicatable fashion due to ageing and this happens no matter how much you baby it.

Click to expand...

Completely agreed. And temperature at storage has as much to do with it as anything else. If a person is going to expend effort, focus on keeping the battery out of extreme heat, whenever possible, will have the highest rate of return.

I Pm'd you the data, as you requested via PM. It's over 1000% difference in degradation rate comparing 30% SOC and 90% SOC for NCA cathode cells, data FROM PANASONIC.

650% difference for LiCo cathode comparing 0% to 100%.

The difference is more significant then the difference between 25C and 60C at the same SOC.

- - - Updated - - -

I don't know where you are getting this 3-5 year or 3 year shelf life? Maybe for some older cells, true for lead acid or chinese LiFePO4 for sure.
Panasonic NCA cell would probably last 30+ years sitting on the shelf at 10% or less SOC.

Let's get a NCA battery graph in here:

Dry conditions in fridge-like temperatures at low SOC seems to be ideal.



Update: Found the Source :http://www.nrel.gov/docs/fy12osti/53817.pdf

These are pure gold:

http://repository.osakafu-u.ac.jp/dspace/bitstream/10466/14150/1/2014900048.pdf
^all the info you'd want


http://www.mdpi.com/2304-6740/2/1/132/pdf

Post by some guy:
johnchamplinhall1 | MARCH 28, 2013
As a Ph.D. physical chemist with 40 years of lithium battery development experience and Model S owner I think I am in a pretty good position to reassure you. The Panasonic cell used in the MS is probably one of the best in the world. I am very familiar with its electrochemical design (LNCA) and have developed physics based models for predicting LNCA life. Providing that your daily driving pattern is reasonable (<100 miles) and its not too hot where you live I predict the battery will last (70% of new range) for > 20 years.
If you want further reassurance get a copy of the paper below by some Panasonic researchers (you have to buy it for $31). Their results equate to 400,000 miles of life even with 160 mile per day driving cycle
“Prevention of the Micro cracks Generation in LiNiCoAlO2 Cathode by the Restriction of DOD” Transactions, S. Watanabe, T. Hosokawa, K. Morigaki, M. Kinoshita, and K. Nakura, ECS Transactions 41 (41) 65-74 (2012).

I stand corrected!

Excellent data from Okashira, and lots to absorb. Those are indeed pure gold! I've been seeking this data ever since I heard about the Model S. THANK YOU, THANK YOU, THANK YOU!

Not only that, I'm going to lower my SOC limit, at least for my daily driving.


Oh, and did I say: THANK YOU

Look at the phd thesis in my last post

10%-70% SOC looks best

0-100% they got ~1200 cycles

10%-70% looks good for 4000+ cycles! That over 700k miles, almost pointless! lol

Very interesting, thanks.