TeslaFi - Battery Degradation Reports (upload your data)

[pdx_m3s]

Estimated miles at 100% tracks with temperature, so I expect this thread to look a little better come summer time.

 

[AlanSubie4Life]

I just hooked up ScanMyTesla App and will report later this week on the data I pull.

287 at 100% full charge.

Assuming you're still at 287.5rmi at 100%, the fullkWh from ScanMyTesla should be about 70.3-70.5kWh for you, as opposed to about 76kWh (or possibly a bit higher) when new.

I'm really very interested in seeing SMT data from someone with a brand new car. I think some videos exist on YouTube showing these new vehicles with higher values than 76kWh (for older model year vehicles) but my current understanding of how things work can't really explain that. 2020 3P vehicles I would expect would show 77-78kWh when new, right now. But would not be surprised if a brand new 2020 3P vehicle showed 79kWh or more.

 

[scottf200]

That’s an interesting one. I’ll take your word for it is is not temperature related.
Seems like quite a different look than typical degradation, and at 15%, you are probably less than 1st percentile. Though your mileage is obviously higher than most. Keep us posted.

Estimated miles at 100% tracks with temperature, so I expect this thread to look a little better come summer time.

This definitely does not look temperature related. Please note that this is the second Dec-Jan (orange arrows) that this car has gone thru.
The blue estimated 100% number dropped and then dropped more.
The green line is the firmware to see if it appears to be related to that. Look above the 2nd orange arrow and you can see the two major FW updates of 2019.36.* and 2019.40.*.

 

[scottf200]

You are doing a lot better than I am. Charging to 90% daily only nets me 272 miles at 24k miles. Started out at 290 with 4k miles. My 18 LR is screwed.

Would say your chart looks like mine. Yet mine has 20k more in miles. More recent cliff jumping has ensued for both of us.

If you guys are TeslaFI users and if want to then you sent me a cut-n-paste of the data from your tables below. You can send/PM a CSV or whatever file (google sheet view only link is fine too).
I would then graph all three sets of data together to see if there is any correlation by firmware or odometer. Probably just a novelty thing. Just throwing it out there.

 

[AlanSubie4Life]

The green line is the firmware to see if it appears to be related to that

However, the "minor" firmware revisions aren't taken into account. For all we know they might matter. (In fact, for the 2019.40.50.1 firmware, which is "minor," we know that's when the 2020 Performance vehicles received their "constant" changes based on wheel size.)

Please note that this is the second Dec-Jan (orange arrows) that this car has gone thru.

Yes. As I said, I do wonder what the true CAN bus capacity is of a brand new Model 3. I've heard it can be 77-78kWh but can't currently find a screen shot of that. That doesn't really work with the 245Wh/rmi (for Dual Motor) constant, because that would mean 318 rated miles (it never shows above 310). So maybe there is a little funny business with the first 1-2kWh of degradation where it just pulls from a bottom buffer and everything shifts down. But when that's exhausted then loss of capacity starts to show. That could explain why nothing showed up the first winter, maybe. Or, there is something wrong with the whole constant formulation (though I have never seen a discrepancy with older vehicles).

To me it does look like there might be some correlation with temperature but I see why you think there is not. In any case, you should be able to remove that from the equation if you have an opportunity to really warm up the battery and then look at the result. I don't know exactly when the TeslaFi pulls its data and what the exact temp of the battery might be at that point.

Definitely a bit weird though. Really no idea.

 

[scottf200]

However, the "minor" firmware revisions aren't taken into account. For all we know they might matter. (In fact, for the 2019.40.50.1 firmware, which is "minor," we know that's when the 2020 Performance vehicles received their "constant" changes based on wheel size.)

I would like to graph down to the 'minor' firmware number after the week-of-the-year ... however, I don't know how to put that in a MSexcel / Gsheets formula for the 5 formats below. If someone wants to take this as a side conversation then I'm more than happy to. I'm currently do the below to fractionalize the week-of-the-year.

0123456789012345
+---------+-----
2017.50
2018.6
2018.6.1
2018.10.4
2018.48.12.1

To me it does look like there might be some correlation with temperature but I see why you think there is not. In any case, you should be able to remove that from the equation if you have an opportunity to really warm up the battery and then look at the result.

It will be a couple months before things get decently warmer there. One test that is being considered is charging to 100% SOC on a weekend day and see via ScanMyTesal and TeslaFI every minute captures when the charging stops. That is not when it gets to 100% but when it continues to 'charge' past that point and does cell balancing. ScanMyTesla does not have module/cell level voltages for TM3s (like the X/S) but you should be able to see the delta between the min and max voltage get smaller as the cell balancing is going on (again ... after you hit 100% and it continues to take amps).

I don't know exactly when the TeslaFi pulls its data and what the exact temp of the battery might be at that point.

Well I think TeslaFI would pull the data when the charging stopped. So the battery is warmed up form being charged for multiple hours. It is also in a garage. It is only 16 amps but I figured that would still be warming up the battery!

 

[AlanSubie4Life]

It is only 16 amps but I figured that would still be warming up the battery!

Might not be enough to keep up if it is fairly chilly. Make sense to use most of the energy to charge the battery, so if it's not below freezing, and you're charging fairly slowly (meaning as long as it's above freezing, you probably don't have to worry about lithium plating (which would necessitate warming) because the charge rate is slow), it's going to just use most of your 3.8kW of power to add energy to the battery. We know efficiencies are worst case something like 90% (if it's not having to warm the battery), so that would only leave 400W that would show up as heat (and we know something like 200W of that is for the MCU which wouldn't be directly warming the battery). That's a really weak hair dryer trying to warm up 1000 pounds of liquid and battery & metal exposed to cold air. Remember the battery heaters (motor/inverter) are something like 3.5kW EACH (in a Dual Motor).

For efficiency reasons, it is really bad to generate significant heat when charging. It's only for necessity that any heating is done, and of course inherent waste heat generated during the charging process also warms things up. But we know from conservation of energy that there can't be that much of that. The faster you charge, of course, the warmer the battery will get (and will have to be, in order to accept the charge rate without permanent damage).

 

[HostileHarry]

Assuming you're still at 287.5rmi at 100%, the fullkWh from ScanMyTesla should be about 70.3-70.5kWh for you, as opposed to about 76kWh (or possibly a bit higher) when new.

I'm really very interested in seeing SMT data from someone with a brand new car. I think some videos exist on YouTube showing these new vehicles with higher values than 76kWh (for older model year vehicles) but my current understanding of how things work can't really explain that. 2020 3P vehicles I would expect would show 77-78kWh when new, right now. But would not be surprised if a brand new 2020 3P vehicle showed 79kWh or more.

You are spot on. Nominal remaining is 70.6kWh according to SMT

 

[AlanSubie4Life]

You are spot on. Nominal remaining is 70.6kWh according to SMT

Thanks. Not spot on though; the middle of my guess window was 0.3% off. Curious - did that end up displaying as 288 @ 100%? It's possible the constant I'm using is very slightly low.

 

[vitesse]

There is no range mode on the model 3

I don't think you're missing much!

 

[HostileHarry]

Thanks. Not spot on though; the middle of my guess window was 0.3% off. Curious - did that end up displaying as 288 @ 100%? It's possible the constant I'm using is very slightly low.

I haven't done a full charge to 100% for at least 3 or 4 months. My 80% charge last night (according to TeslaFi) shows an estimated 287.53 at 100% charge.

I'm a bit disappointed in the Model 3 battery degradation. My 2016 Model X 90 has only seen about 6% degradation after the first year or so, but has remained very constant ever since. Hopefully the Model 3 battery settles down and maintains that 8% loss.

 

[AlanSubie4Life]

Hopefully the Model 3 battery settles down and maintains that 8% loss.

I think all Model 3 owners are hoping that at this point.

 

[SomeJoe7777]

I have two Model 3s -- one is a RWD LR, the other is a P3D.

Check out the difference:

3 LR:

P3D:

The 3LR has 2% loss in 19,500 miles.

The P3D has nearly 12% loss in 30,000 miles.

Both vehicles are charged every night to 90% at 24A (1/2 capacity) or 48A (full capacity) if only one car needs charging.

The P3D has 518 kWh of supercharging (5.3% of all kWh charged), and 9 charges to 100%.

The 3LR has 505 kWh of supercharging (11.1% of all kWh charged), and 7 charges to 100%. (We use the 3LR for road trips).

[Note: The 3LR's charging numbers are approximate because there was no logging for the first 4300 miles of ownership).


Riddle me that one, Batman.

[Edit: Numbers for 3LR were off].

 

[KenC]

This definitely does not look temperature related. Please note that this is the second Dec-Jan (orange arrows) that this car has gone thru.
The blue estimated 100% number dropped and then dropped more.
The green line is the firmware to see if it appears to be related to that. Look above the 2nd orange arrow and you can see the two major FW updates of 2019.36.* and 2019.40.*.

Busy chart with lots of good info. However, I think you might see more if you focus on just two variables, Rated Range and Temp. Adjust the scales to see more clearly, and see what you get when you just focus on the two, cause I see some correlation with temperature drop.

 

[scottf200]

Busy chart with lots of good info. However, I think you might see more if you focus on just two variables, Rated Range and Temp. Adjust the scales to see more clearly, and see what you get when you just focus on the two, cause I see some correlation with temperature drop.

Thanks! There are really just 2 to focus on.
1) Dark blue is Rated Range.
2) Purple is Temp.
Orange arrows just show the winter months at the same temp and you can see vast differences where they are at the blue lines.

 

[AlanSubie4Life]

The 3LR has 2% loss in 19,500 miles.

I would argue that you actually have 5.5% capacity loss on this, as the full rated range is 325 miles.

If you had a CAN bus reader, it would read about 307rmi*234Wh/rmi = 71.8kWh

And the 3P+ would read about 277rmi*245Wh/rmi = 67.9kWh.

They both started around 76kWh or so.

That 3P+ is not looking good. I'm starting to get a bad feeling about my future with my car.

 

[scottf200]

If you had a CAN bus reader, it would read about 307rmi*234Wh/rmi = 71.8kWh

FYI, ABRP log show 72.8 kWh -- Tesla Battery Charging Data from 801 Cars

kWh Wh/mi Range
311.11 = (72.8 * 1000) / 234 = ( kWh*1000 ) / Wh_per_mile

 

[AlanSubie4Life]

FYI, ABRP log show 72.8 kWh -- Tesla Battery Charging Data from 801 Cars

kWh Wh/mi Range
311.11 = (72.8 * 1000) / 234 = ( kWh*1000 ) / Wh_per_mile

That's an old post which predates the unlock of visible capacity on the Model 3 LR RWD. The constant has never changed, but in 2018 it only had 310 miles of range. Then they changed it to 325 rated miles in March 2019 by making more capacity visible. As near as I can tell, that's what happened, anyway. It wasn't an actual unlock of capacity (it was always there, just hidden below zero miles), since in the EPA test they were always able to make that distance (I think it was close to 330 miles actually) and the test shows they discharged 78kWh from the LR RWD battery (very similar to the Dual Motor battery which was 79.2kWh).

Read backs of nearly 78kWh have been reported on new cars. (I don't know how to align that with the maximum number of 76kWh which seems to be applicable to the LR vehicles, but in any case the number is at least 76kWh on a new vehicle.)

(That being said, the vehicle above didn't appear to receive the "full" unlock - but it probably had significant mileage on it at that point, so hard to say what should have been expected. We know it has the same number of cells as a Dual Motor battery, so it "should" have the same capacity, and the EPA test backs that up.)

 

[SomeJoe7777]

I would argue that you actually have 5.5% capacity loss on this, as the full rated range is 325 miles.

If you had a CAN bus reader, it would read about 307rmi*234Wh/rmi = 71.8kWh

(That being said, the vehicle above didn't appear to receive the "full" unlock - but it probably had significant mileage on it at that point, so hard to say what should have been expected. We know it has the same number of cells as a Dual Motor battery, so it "should" have the same capacity, and the EPA test backs that up.)

The 3LR is a very early released vehicle (VIN 0047XX). Perhaps it's possible that it's battery is a slightly different construction/part#, and it did not receive the 325 mile range upgrade?

Is the CAN bus on the 3 that would read out the battery capacity accessible from the OBD-2 port, and is the data unencrypted? If so, I may be able to read it with an oscillosope that has serial protocol decoding.

 

[scottf200]

Is the CAN bus on the 3 that would read out the battery capacity accessible from the OBD-2 port, and is the data unencrypted? If so, I may be able to read it with an oscillosope that has serial protocol decoding.

You need an adapter to go from a unique one (of several) that Tesla is using to an OBDII. We got one for my sons car and he is using ScanMyTesla to display the data. Most details on the web that I know of are here: Diagnostic Port and Data Access - Tesla Owners Online