Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

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

This site may earn commission on affiliate links.
What is the weather like in your area? My 2019 LR RWD with 49,890 miles shows 305 at 100% in the summer and 285 in the cold of winter.

You try the battery calibration thing, but personally I’d just drive it normally & cycle the battery 20%-90% over the next few months and see if things get back to normal. Who knows how the previous owner treated it.
 
I am in Florida so HOTT lol
The car name was "Krunk" so........ lol We bought it for an around town type vehicle as my other vehicle is a 2021 Ram Rebel which gets 12 mpg

We planned on using the truck for beach runs and the m3 for errands and kid school runs.

I will try the calibration and keep this updated for anyone else as it might help.
 
you mean true SOC or displayed SOC?

I took a reading few months ago at 0.46% SOC (true SOC 5 to 7.1%) and I had imbalance 16mv and cell voltage was around 3.060V
so sounds like you have perhaps a bit more juice than what the car suggests.
Here my SMT with mine at around 0%. It was a moment where I was deciding to stop OR to go lower sub 0 %. (in real I allowed to go -1,56% SOC and -1,56% SOC Expected)

This is the frame at -0,28% SOC (just used some buffer)
1646916318478.png


What I don't understand are the values SOC UI, SOC Min SOC max , considering that at the screen it was 0% since some km before.
 
  • Informative
Reactions: KenC
I tried asking elsewhere but only got guesses for answers... so here's my question:

In the energy app, why would the "instant range" prediction be different depending on whether the 5, 15 or 30 mile graph is being displayed? The manual says the instant prediction takes the last few consumption datapoints rather than an average. I would have expected to see the same instant range regardless of which chart was being displayed.

Small point I know and everyone will say dont use the predictions because they are way off.. so why does Tesla bother to show them? I cant help feel that understanding how they are calculated would help.
 
I bought a 2020 Tesla model 3, dual motor. My range at 100% is 274 miles.

If you used the slider in the app to find the range, it might not be the same as a real full charge in the car and can not be used to find degradation etc.

After full charge, let the car sleep for an hour or so. Then drive it down to low SOC, preferably lower than 10% and let it sleep at that SOC a couple of hours(ore more) before charging.
 
Here my SMT with mine at around 0%. It was a moment where I was deciding to stop OR to go lower sub 0 %. (in real I allowed to go -1,56% SOC and -1,56% SOC Expected)

This is the frame at -0,28% SOC (just used some buffer)
View attachment 779083

What I don't understand are the values SOC UI, SOC Min SOC max , considering that at the screen it was 0% since some km before.
Do you have any other pictures from that point? Battery pack voltage or cell voltages would be nice to see. Imbalance would be interresting too.

A sleep after this in that SOC would be nice for the BMS, to find the true SOC from the Open Circuit Voltage.

SOC = the SOC according to the screen.
As the screen SOC have a 4.5% true SOC hidden (buffer) you should have about 4.5% true SOC at 0% on the screen.
I think SOC Min looks like the true SOC value, as 4.5-0.28= 4.22.
Other from this I havent digged deep into the different values. I habent found the values important to me.

From logic, SOC min should be the SOC the lowest cell and Max the highest viltage cell…?

For calibration my ”new theory” is that if one charges to 100% and the charging stops there is really not a big need to let the car sleep. That end of the scale is quite safe to call ”known” direct after the charging is done. It stops at 4.20v/cell so the battery is at 100% SOC and the energy is counter already.

The real need is to find the other end and when driving the cell voltage is drooped due to the load. It takes some time for the battery to recover to the OVC-voltage. For this the car need to sleep for a while.
 
  • Like
Reactions: conv90
What I don't understand are the values SOC UI, SOC Min SOC max , considering that at the screen it was 0% since some km before.

From the github page by the people behind Scan My Tesla:
These are off compared to the displays in the car. Kept alive for legacy reasons, and theories about where the absolute empty of the battery is
Found here:
Model 3 CAN bus IDs and data

Basic concusion: Dont dig that much into that. Use the SOC number and renember that the BMS tries to keep 4.5% of the total capacity below 0% SOC.
 
  • Like
Reactions: conv90
Do you have any other pictures from that point? Battery pack voltage or cell voltages would be nice to see. Imbalance would be interresting too.

A sleep after this in that SOC would be nice for the BMS, to find the true SOC from the Open Circuit Voltage.

SOC = the SOC according to the screen.
As the screen SOC have a 4.5% true SOC hidden (buffer) you should have about 4.5% true SOC at 0% on the screen.
I think SOC Min looks like the true SOC value, as 4.5-0.28= 4.22.
Other from this I havent digged deep into the different values. I habent found the values important to me.

From logic, SOC min should be the SOC the lowest cell and Max the highest viltage cell…?

For calibration my ”new theory” is that if one charges to 100% and the charging stops there is really not a big need to let the car sleep. That end of the scale is quite safe to call ”known” direct after the charging is done. It stops at 4.20v/cell so the battery is at 100% SOC and the energy is counter already.

The real need is to find the other end and when driving the cell voltage is drooped due to the load. It takes some time for the battery to recover to the OVC-voltage. For this the car need to sleep for a while.
Yes you are saying things that make sense about SOc Min /max and buffer
I have no Scrren of SMT but like I said some post ago
YES, 3,052 (min) to 3,070 (max), 18 imbalance when car was stopped and me out of car with the car consuming 0,32 kW , so not sleeping [ALL of this at -1,56% SOC.)
For what I understand the SOC on the screen is the SMT "SOC Expected" , not the SMT "SOC".
example: SOC 53,1, SoC Expected 52,3 = SOC on Screen =53%
About 4,200v/cell:
It's a degraded battery the one showing LESS than 4,20? OR 4,20, it's always reached?
I mean: a degraded battery will shows less than 4,2?
 
  • Like
Reactions: AAKEE
Yes you are saying things that make sense about SOc Min /max and buffer
I have no Scrren of SMT but like I said some post ago
YES, 3,052 (min) to 3,070 (max), 18 imbalance when car was stopped and me out of car with the car consuming 0,32 kW , so not sleeping [ALL of this at -1,56% SOC.)
For what I understand the SOC on the screen is the SMT "SOC Expected" , not the SMT "SOC".
example: SOC 53,1, SoC Expected 52,3 = SOC on Screen =53%
About 4,200v/cell:
It's a degraded battery the one showing LESS than 4,20? OR 4,20, it's always reached?
I mean: a degraded battery will shows less than 4,2?
You need to have the battery completely not loaded(Open Circuit) to se 4.20v.
For other lithium batteries I have used they keep 4.20V after charging to 4.20v but if they are really tired they start to show perhaps 4.16 och 4.17V, and split the voltage between the cells even if balanced charged . The thing that seem to affect the possibility to keep 4.20v after charging seem to be if they have ben left at high SOC for long. I have batteries that are five years and still keep 4.20v/cell.

I don know for these cells but I would guess that a tesla battery would keep the ability to stay at 4.20 volt for quite long. I guess that leaving the car fully charged for long in a hot climate could have an negative effect.

Tesla burn of voltage from the highest cells to balance them. this means that the voltage might drop after a charge until the balance is finished. A 100% charge stays connected for a while at 100% when the last cells with low er voltage get their final charging, but I guess the balancing can take longer time and after the chareging is done, the voltage probably drop a little when the balancing burns of the highest cells.

When charging to 90%, the charging need to stop when the target is reached and the balancing will continue after the charging stops causeing the voltage to drop slightly.

[Edit]To be clear, as there is load when the car is on and nit sleeping the voltage most probably is not 4.200V when checking with SMT.
Also, as we see the voltage with three digits its possible/probable that a cell dont keep 4.200 if it was charged to this value, so maybe 4.195 to 4.200 at rest (OVC) and slightly lower at the load havibg the car not sleeping. AC, lights = off should reduce the drop.
 
Last edited:
For other lithium batteries I have used they keep 4.20V after charging to 4.20v but if they are really tired they start to show perhaps 4.16 och 4.17V, and split the voltage between the cells even if balanced charged . The thing that seem to affect the possibility to keep 4.20v after charging seem to be if they have ben left at high SOC for long. I have batteries that are five years and still keep 4.20v/cell.
That self-discharge was one of the characteristics of NCA cells left at high SOC for extended periods as mentioned in one of the papers covered recently in this thread if I remember correctly.
 
That self-discharge was one of the characteristics of NCA cells left at high SOC for extended periods as mentioned in one of the papers covered recently in this thread if I remember correctly.
Im not refering to self discharge but that the cell is charged to 4.20V/cell but when the charger is disconnected the voltage drop to 4.16-4.17V.
A fresh cell stays very close to 4.20v after a charge but older cells closing in to the end of life(in my hobby) develop a voltage drop.
 
Im not refering to self discharge but that the cell is charged to 4.20V/cell but when the charger is disconnected the voltage drop to 4.16-4.17V.
A fresh cell stays very close to 4.20v after a charge but older cells closing in to the end of life(in my hobby) develop a voltage drop.

Well, the paper describes them as "internal short circuits" which I assume would also mean self-discharge - at least to some degree:
Cells kept at 100 % SoC do not show the fastest capacity fade but develop internal short circuits for temperatures T ≥ 40 o C. Degradation is slowest for cells stored close to 0 % SoC at all temperatures. Rates of capacity fade and their temperature dependencies are distinctly different for SoC values below and above 60 %, respectively.
 
Check this out. A Kia Niro that has gone over 90,000 KM with 1% degradation. But you are OK to lose 11% of capacity after 14,000 miles for an expensive Tesla. A cheap KIA doesn't have that issue. Gee, I wonder why?
Well, I do not own a type rating in the Kia Niro but:

The video show that the full charge voltage is 4.14V/cell.
For NMC622(which it seem to have) the full charge standard is 4.20V/cell.

This means, in plain words that the Niro seem to have a top buffer.

As NMC will degrade (from calendar aging) about the same as Teslas NCA cells we can be sure that degradation has happend.
If it do mot look like any degradation it is very probable that Kia hides the degradation by reducing the top buffer.
When it is done this way it will look like no degradation until the top buffer is eaten up, and after this it will show.

Tesla do not really hide any degradation as the do not have any top buffer.

If you had a choise of either;
-Buying 70kWh but newer being able to use more than 65 of these( but not seeing any degradation initially) or
-Buying 70kWh and being able to use all 70 initially.

What road would you choose?

Below is a chart from a research report ( there is a lot of this kind of research and they mainly tell us the same thing).

Teslas ”classic” battery is NCA. Model 3 in europe get NMC and maybe NCMA now.
Nito has NMC622 cells.( most common type of NMC in the recent years research)
We see that there is not a big difference in the degradation.
Kia can not stop the mechanic of degradation so they most certainly will degrade. But we might not see it initially because they probably have hodden this in the top buffer.

1E31FD3A-F3F9-4962-A630-B1306EDBE315.jpeg
 
Check this out. A Kia Niro that has gone over 90,000 KM with 1% degradation. But you are OK to lose 11% of capacity after 14,000 miles for an expensive Tesla. A cheap KIA doesn't have that issue. Gee, I wonder why?


I'm not ok with this at all. Especially given that degradation for the amount of cycles doesnt seem that much different from i.e. a mobile phone. In my opinion from the user reports we are seeing Model S has less degradation than the 3.
Noone here has posted a teslafi average curve for the S though so we cant quite confirm.


Actually lets take that back, I am very happy for degradation, I just think advertising the EPA range is false advertisement. It should be advertised with average degradation in mind after 2-3 years. In fact, they probably should predegrade the batteries or add some chemical which reduces charge by i.e. 5-10% but then protects against further degradation.... But while its a race to the top of the highest EPA range this just wont happen and regulators need to step in when EVs are more common.
 
If you used the slider in the app to find the range, it might not be the same as a real full charge in the car and can not be used to find degradation etc.

After full charge, let the car sleep for an hour or so. Then drive it down to low SOC, preferably lower than 10% and let it sleep at that SOC a couple of hours(ore more) before charging.
Thanks. I just hit 100% charge. Took forever due to using a standard 110v outlet. My electrician comes tomorrow thankfully. I will now use the car without charging and get it under 10%... Lets see what happens :)
 
  • Like
Reactions: AAKEE
I'm not ok with this at all. Especially given that degradation for the amount of cycles doesnt seem that much different from i.e. a mobile phone. In my opinion from the user reports we are seeing Model S has less degradation than the 3.
Noone here has posted a teslafi average curve for the S though so we cant quite confirm.

First of all, there isnt the same chemistry in model 3 2170 as in early Model S.

When Increasing the capacity per kilo(weight) / per Liter(volume) I think the main principle is to increase the part of Nickel and reduce the cobalt part.
Cobalt is among other things acting stabilizing on the battery.

Tesla have revealed that they did reduce the cobalt with 59% during six years from early model S to the introduktion of the M3.
The 2170L did get firther reduction byt the number was not disclosed.
We probably should be ready to find that Tesla found out that the initial batteries did hold up much longer than needed and that
during the development of new battery cells didnt need to be as good as the batteries anyway will survive the cars.

Much lower part of cobalt but new battery technology - good but maybe not as good as before.

While you wait for the
Teslafi Model S data, go trough the Teslalogger data on Teslalogger.de. You can selet most
models/versions and check for the degradation curves. ( in some charts there is obviously versions with different range in the same graph but that can be judged for when viewing the graph.

What I see is about the same degradation for Model S and X as model 3. I do not think the tesla battery survey is showing us the exact thruth.
 
  • Informative
Reactions: Dave EV
First of all, there isnt the same chemistry in model 3 2170 as in early Model S.

When Increasing the capacity per kilo(weight) / per Liter(volume) I think the main principle is to increase the part of Nickel and reduce the cobalt part.
Cobalt is among other things acting stabilizing on the battery.

Tesla have revealed that they did reduce the cobalt with 59% during six years from early model S to the introduktion of the M3.
The 2170L did get firther reduction byt the number was not disclosed.
We probably should be ready to find that Tesla found out that the initial batteries did hold up much longer than needed and that
during the development of new battery cells didnt need to be as good as the batteries anyway will survive the cars.

Much lower part of cobalt but new battery technology - good but maybe not as good as before.

While you wait for the
Teslafi Model S data, go trough the Teslalogger data on Teslalogger.de. You can selet most
models/versions and check for the degradation curves. ( in some charts there is obviously versions with different range in the same graph but that can be judged for when viewing the graph.

What I see is about the same degradation for Model S and X as model 3. I do not think the tesla battery survey is showing us the exact thruth.

Yes, and we know that obviously the S chemistry has drawbacks i.e. perhaps little degradaton but supercharger throtteling etc.
Where is the teslalogger data collected from?

I have had a look at the Model 3 longrange data and that seems to be roughly consistent with Teslafis average (i.e. 459km average at 50k kms for my long range performance). >100k kms is obviously scewed because thats mainly high milage drivers who suffer from less time based degradation.

However, I can also clearly see that Model S has maybe around half of the degradation the 3 has.


I.e. at 60k kms Model 3 LR is 499 (520km) ->456km
at 60k Model S100D is 500km -> 483km
 
  • Like
Reactions: AAKEE