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

[AAKEE]

Damage information? With respect AAKEE, surely the data does show (irreversible?) degradation. Let's use the publication's Figure 2 with NMC stored @ 50°C. When 70-80% SoC is used, at the end of the author's 9-10 months storage, the relative capacity appears to be around 83% of the original starting value.

I think youre missing the point. Go back to the post with the figure that’s discussed:
Bouba #4720

The upper three figures, a, b and c do not show damage. They show the anode potential.
Its very clear that the graphs of degradation do show “damage” so that is not a matter of discussion, I hope

 

[Steve446]

Hi AAKEE, the graphs A and B show damage to the batteries but they are almost identical...yet the NMC battery in graph E falls off a cliff...almost reaching the Tesla replacement guarantee in only 10 months!
In summer here a car can easily get to fifty degrees (if you didn’t put cabin overheat protection on)...of course only for a few hours a day and only at the height of summer...but it would add up over the eight year warranty period.
Just out of interest, I built a car porch to protect my Tesla from the sun

Bouba, based on the following, even on very warm days it's unlikely that as an average, your battery would spend much time if any at 50°C. 1) it will cool down overnight, 2) it does have a thermal control system which I assume will keep it cooler than continuous ambient daytime exposure would where you are based and 3) it should be cooler in winter. As a comparison, we do have some warm days here in W France (foot of Brittany). So when it's 28°C like yesterday and after a 120 Km drive, my battery was around 31°C. That's less than supercharging preconditioning; eventually it drops back to around 18 to 21°C. I put a carport in place too and it certainly helps keep the car cooler.

 

[Bouba]

Bouba, based on the following, even on very warm days it's unlikely that as an average, your battery would spend much time if any at 50°C. 1) it will cool down overnight, 2) it does have a thermal control system which I assume will keep it cooler than continuous ambient daytime exposure would where you are based and 3) it should be cooler in winter. As a comparison, we do have some warm days here in W France (foot of Brittany). So when it's 28°C like yesterday and after a 120 Km drive, my battery was around 31°C. That's less than supercharging preconditioning; eventually it drops back to around 18 to 21°C. I put a carport in place too and it certainly helps keep the car cooler.

The cabin overheat protection has made me think...it will stop the interior of the car exceeding 40 degrees C...but does that mean it will heat up the battery because it powers the air conditioner...or does the air conditioner keep the battery cool ?

 

[Steve446]

I think youre missing the point. Go back to the post with the figure that’s discussed:
Bouba #4720

The upper three figures, a, b and c do not show damage. They show the anode potential.
Its very clear that the graphs of degradation do show “damage” so that is not a matter of discussion, I hope

No problem, crossed wires. I should have mentioned that I referred to graph series d, e and f as you say.

 

[Steve446]

The cabin overheat protection has made me think...it will stop the interior of the car exceeding 40 degrees C...but does that mean it will heat up the battery because it powers the air conditioner...or does the air conditioner keep the battery cool ?

The battery has a fluid cooling system so should condition the battery as per Tesla programming (but I don't know what that might be...).

 

[GtiMart]

The AC doesn't consume enough energy to make the batteries heat. If necessary, the car will cool the batteries using the same AC.

 

[AAKEE]

The cabin overheat protection has made me think...it will stop the interior of the car exceeding 40 degrees C...but does that mean it will heat up the battery because it powers the air conditioner...or does the air conditioner keep the battery cool ?

If you use the cabin overheat protection when the car is standing outside ypu probably lower tge battery temp a few degrees during the time that the overheat protection is active.
For the average battery temp durijg tge whole year, it probably do not do a big difference.
If you consider leaving the AC on, at 20C onboard, you probably could lower the temperature more. Still, average over one year, probably not much and probably an expensive way to try to keep the degradation down.

I would use shading from direct sun when possible and park in a garage if possible.
(I do both, as I have a garage and also a shading carport.)

 

[conv90]

The constant is the same, it is the estimated capacity that changes.

Sure, but the point is:
if the capacity is not really changed and I'm able to respect the constant Wh/km while driving , IF the "raised by Software" kWh are FAKE i'll end the juice before to be at 0% ....

 

[Steve446]

Sure, but the point is:
if the capacity is not really changed and I'm able to respect the constant Wh/km while driving , IF the "raised by Software" kWh are FAKE i'll end the juice before to be at 0% ....

But, I thought that the displayed value is dynamic ie it will self correct at some point before it gets to 0 SoC on the screen or have I misunderstood something?

 

[AlanSubie4Life]

Sure, but the point is:
if the capacity is not really changed and I'm able to respect the constant Wh/km while driving , IF the "raised by Software" kWh are FAKE i'll end the juice before to be at 0% ....

I suspect Tesla tries to bias any capacity estimates on the pessimistic side, if they can, so I think it is more likely that the recovered energy is real and the initial estimate prior to software update was too low. Only way to really know is to do a long drive without stopping to fairly low SoC, and look at the trip meter and find out whether it seems fairly well dialed in. And of course crosscheck with SMT since you have it (definitely do both concurrently!). Trip meter values should be about 1% lower than the delta in SMT nominal remaining values.

 

[conv90]

I suspect Tesla tries to bias any capacity estimates on the pessimistic side, if they can, so I think it is more likely that the recovered energy is real and the initial estimate prior to software update was too low. Only way to really know is to do a long drive without stopping to fairly low SoC, and look at the trip meter and find out whether it seems fairly well dialed in. And of course crosscheck with SMT since you have it (definitely do both concurrently!). Trip meter values should be about 1% lower than the delta in SMT nominal remaining values.

When I was at "only" 73,9 kWh I performed a sub zero% trip to a Tesla destination charger just to see if any additional millivolts were left. The 0km left appeared just crossed the 0% SoC Expected
The maesurement (with SMT) was in line with the fact that 73,9 was the real capacity and that no additional charge was left (I consumed over 2,3% past the buffer).
So the "pessimistic side" it's for sure a real thing but not up to the point of the difference between 73,9 kWh and the 78,1 kWh of today (it raised again )
My car has 27k km it's a march 2021
An update from 4 of my italian friend with a Model 3 Perf with E3DL 82 kWh battery pack:
1 with a september 2021 and about 20k km from 76,7 , recovered to 79,5 expecially in the last 5 days (he can't say the initial value)
1 with a march 2021 /same as mine but with 25k km more than mine recovered from 74 to 75 only
1 with a march 2021 /same as mine (same km as mine) recovered from about 75 to about 76 only
1 with a march 2021 /same as mine but with 10k km less lost about 1 kWh (75 to 74kWh)

 

[AlanSubie4Life]

The 0km left appeared just crossed the 0% SoC Expected
The maesurement (with SMT) was in line with the fact that 73,9 was the real capacity and that no additional charge was left (I consumed over 2,3% past the buffer).

It’s very hard to say with any certainty what is going on in your case. So what was the kWh delta (startkWh - endkWh) in SMT Nominal Remaining for this test case where you went 2.3% into the buffer? Sounds like you did not monitor trip meter consumption (this would be key as well; it's a different measurement than the kWh delta (or just rated miles delta) measurement, so it may well yield a different result aside from the standard 1% discrepancy due to "heat loss")? Definitely do both if you try this again (but it unfortunately has to be done continuously with no stops).

One possibility is that on some packs Tesla loses/lost track of the true “zero” point and was pessimistic about it. That would mean there would be more energy than the standard 4.5% in the buffer. I have no idea how likely this would be (voltages, battery balance, etc should give BMS a clue - but a software bug can manifest in nearly any conceivable way with the right set of inputs). And it is possible that software updates have rectified the limitation and it has been gradually reassessing and updating the CAC.

Complete and total speculation. I have no idea. Just randomly speculating with some hypothesis to fit the set of data and facts.

 

[AAKEE]

But, I thought that the displayed value is dynamic ie it will self correct at some point before it gets to 0 SoC on the screen or have I misunderstood something?

Yes, slightly missunderstanding.

SOC is determined by measuring the cell voltage when the battery is at rest( Open Contactor Voltage).

100% SOC is 4.20V.
0% (by the manufacturer/bransch standard) is 2.50V.

As soon as the battery is delivering power the voltage drops and the SOC can not be measured. But it can be estimated.
To estimate the SOC the BMS estimates the capacity( the amount of available energy between 100% and 0%.

When the capacity (Nominal Full Pack) is determined the SOC can be estimated by calculating the SOC by reducing the initial SOC with the used energy.
A 80kWh battery that started with 100% SOC should have about 50% SOC if 40kWh is used. If this car is parked and sleeps the OCV can be measured. Lets say 50% (true) SOC is at 3.78V/cell. If the battery voltage is measured to 3.76V, there is less than 50% SOC, and then the real SOC is known from the Voltage. The car adjust the SOC after/during the sleep.

This is an exampke from last week:
Arriving with 45%, the car is parked over the working day and when departing the SOC is 46%. The later value is more riable as it is “true SOC” measured by the OCV.

If it is really obvius I guess tesla can adjust “on the fly” but it will be hard to judge the real SOC when driving.

 

[AAKEE]

Yeah, if that’s true (and I don’t know), it makes TeslaFi useless for this. But I suspect it can’t be true, because I thought @AAKEE ’s TeslaFi plots look pretty steady and he lives in a very cold place.

Yes, my teslafi range have not shown any seasonal variation. Teslafi is erratic in the range graph (this comes from limitations in the tesla API I guess).
I did write down the NFP almost daily for the first 9 months, then I got teslalogger but I have some glitches in that data so i do not have full coverage.

If anything, I had the lowest NFP during the the middle of the summer( a two week period with slightly below 80kWh).
Besides that, I had around 80.5-81 from the car was new and for three months. Then one month with 81.4 stable. All fall, autumn and winter was around 80.5-81.

I might be able to superimpose the NFP on this graph, to show the stable NFP.

We can see my latest discharge to find the real battery capacity as the first clear drop below 500km range. In reality the NFP dropped to 79.1 as the lowest NFP and this correlates to 498km so teslafi underestimates the range with quite much.
The following climb is what usually happens.

 

[AAKEE]

For SOC determination, the steeper the curve the easier it is for the BMS to find the correct SOC.
There is a good reason to go below 3.2Volt when doing a BMS calibration. This is the OCV so its the voltage during sleep, not driving.
To be sure about being on the steep side, aim for 3-3.1V when in Park, no full AC, lights off.

This is a Panasonic NCA battery (NCR18650B)

Test here: OCV test

 

[conv90]

It’s very hard to say with any certainty what is going on in your case. So what was the kWh delta (startkWh - endkWh) in SMT Nominal Remaining for this test case where you went 2.3% into the buffer? Sounds like you did not monitor trip meter consumption (this would be key as well; it's a different measurement than the kWh delta (or just rated miles delta) measurement, so it may well yield a different result aside from the standard 1% discrepancy due to "heat loss")? Definitely do both if you try this again (but it unfortunately has to be done continuously with no stops).

One possibility is that on some packs Tesla loses/lost track of the true “zero” point and was pessimistic about it. That would mean there would be more energy than the standard 4.5% in the buffer. I have no idea how likely this would be (voltages, battery balance, etc should give BMS a clue - but a software bug can manifest in nearly any conceivable way with the right set of inputs). And it is possible that software updates have rectified the limitation and it has been gradually reassessing and updating the CAC.

Complete and total speculation. I have no idea. Just randomly speculating with some hypothesis to fit the set of data and facts.

unfortunately during this test i've only a screen of SMT when at -1,56% SoC and same -1,56 SoC expected.
In this precise moment and car in P and displayed 3,3 kWh buffer (at 73,9kWh NFP the buffer is less than usual 3,6kWh) the Nominal remainig was 2.2 kWh.
So perfect in line with caculations. There were some more kWh hidden over the 2,2 kWh? Who knows?
Probably (..maybe) if continuing to drive to go furter in the buffer(and over), the Nominal Remaining could have been UNDER 0kWh if the real capacity was more.
But I doubt (and I'd not try it...) it would go under (-4kWh). (4 kWh is what I reaganed in the last month).

 

[AAKEE]

unfortunately during this test i've only a screen of SMT when at -1,56% SoC and same -1,56 SoC expected.
In this precise moment and car in P and displayed 3,3 kWh buffer (at 73,9kWh NFP the buffer is less than usual 3,6kWh) the Nominal remainig was 2.2 kWh.
So perfect in line with caculations. There were some more kWh hidden over the 2,2 kWh? Who knows?

Did you let the car sleep at that low SOC?
I think the key point is to let it sleep, to show the OCV to the BMS.
If the OCV is higher than expected, the BMS ”finds” the lost energy.

If the OCV is below 3.2V, the span of uncertiness of the SOC is reduced quite much. ( The voltage climbs when the battery is at rest so its needed to aim below when the car is awake, 3.0 to 3.1V or so.)

By the way, what cell voltage did you have at -1.56% ?

 

[AlanSubie4Life]

Probably (..maybe) if continuing to drive to go furter in the buffer(and over), the Nominal Remaining could have been UNDER 0kWh if the real capacity was more.

Probably the battery would have shut itself down, though it may have been premature. Hard to know.

Anyway, if you get another opportunity, monitor both the start and end on SMT and ALSO monitor trip consumption (also zero the trip meter just in case you do have to stop (briefly!), which messes up the experiment, but if it is real quick with minimal time spent in park it might not be too large an error).

 

[gabrielschulho]

Hey, all!

I tried to make heads or tails of all the discussion regarding battery degradation, but I think I need a helping hand. I live in Southern California. I have a Tesla Model 3 Standard Range I bought at the end of September 2021. It is now May 2022. I charged it using the ~1kW wall charger that plugs into a regular 110V wall socket. I took one road trip during which I charged it at a supercharger around 5-6 times. Once or twice I let the charge dip below 40 miles. One time it was as low as 15 miles left. The total accrued mileage since I bought the car is 7297 miles. Aside from that one road trip, I charge the car to 90 percent during off-peak hours (9 PM - 4 PM the next day).

I tend to press the accelerator fairly aggressively, if that makes a difference.

When I first bought the car, charging to 90% meant I could get 231 miles. Now, it means I can only get 218 miles. I'm a bit alarmed by the fact that there's already a 13 mile decrease, and I've owned the car for less than 7 1/2 months.

Is this normal? Is it worth making a service request?

Thank you in advance for your advice!

 

[gabrielschulho]

P.S.: I don't know if it has LFP batteries or not.