MASTER THREAD: 2021 Model 3 - Charge data, battery discussion etc

[AlanSubie4Life]

but when we charged it to 100% on a V3 SuC it reached 80,5kWh straight away. Ive posted these results here, maybe two months back.

Yes I remember. And I remember it happened very quickly. I just worry about those early results showing quick recovery being impacted by temperature issues (looked low capacity because the pack was very chilly, etc.).

I just see a lot of reports recently from LR Y/3, etc, with 82.1kWh FPWN, but with energies in the mid-79kWh range.

And we have a number of examples with owners who see no evidence of recovery (much more rare with Peformance owners but @conv90 definitely had issues). It isn’t just one off, I don’t think, anyway.

I feel like if it doesn’t recover to something in the ballpark in the first couple full charges (still below 80kWh), there might be little hope for getting to 81kWh (all temps being equal, etc.).

Could be wrong though.

Then back to 60% daily...In order to disprove your theory (@AlanSubie4Life) I need to get to 80,9kWh, I guess. Way to go

It will be interesting. I guess yeah if you get to 80.6kWh or 80.9kWh or so from where you are now, I would believe it is maybe just a balance/calibration issue and actually even the low initial capacity packs can recover.

In the case of your Performance I guess I don’t think an artificially limited pack makes sense. But you could record voltages at low SoC I guess (I assume it maxes out at 403V at 100% as expected). Might be useful to compare to someone with the 77.8kWh pack at some point. Or to someone with an 82.1kWh with high capacity.

 

[FredMt]

The first LR E3LD 82kWH I could my hands on started with a low 79,X Nominal Full Pack, but when we charged it to 100% on a V3 SuC it reached 80,5kWh straight away. Ive posted these results here, maybe two months back.

@ArnaudC06 case - TM3P 2021 / end Q4 2020 - was an exception 79,5kWh max after supercharge.
Unfortunately @ArnaudC06, has sold his car and this case can no longer be analyzed.

For information some screens before supercharge :

Batterie de 82 kWh sur la Model 3 Grande Autonomie version 2021

forums.automobile-propre.com

And the indications after balancing at the supercharger:
491km EPA max.

Batterie de 82 kWh sur la Model 3 Grande Autonomie version 2021

forums.automobile-propre.com

 

[AAKEE]

My guess is still, that this is mostly a BMS calibration topic.

Yes, thats my guess as well.

I also do not think Tesla/Panasonic has a lot of different 2170L cells.
Most probable, theres only one version ?

Differences in the battery pack version numbers might come from different hardware. Probably they get different versions of BMS:es due to different chip versions, and maybe other differences in hardware on the exact same bases we see different hardware on other car brands models in the same model/model year.

So, these differences shouldnt affect the real batterycapacity, as the battery cell itself set the limit.

I can check my battery sticker if it is to any value for the discussion( didnt find the guide for that by a brief search, I did se and read it some months ago…).

My M3P iscurrently at about 81.0 kWh NFP.
I had a few higher charges and after this it isnt 81.4(as I had steady for a mounth or more). So far I havent been charging at the normal 56-60% for more than a few days so I do not know if it will climb back to the good old numbers by itself.
I’m positive that this is only a BMS calibration thing and that there is not a big change in the batterys real capacity, so Im not worryed about this at all.

 

[eivissa]

I just worry about those early results showing quick recovery being impacted by temperature issues (looked low capacity because the pack was very chilly, etc.).

When I did the 60% SoC limit test I can safely say that the temperatures weren't changing much. It was in the range between 15°C to 25°C during all those days. Therefore I dont beleive the temperature played any role, but it was all the charging technique.

But you could record voltages at low SoC I guess (I assume it maxes out at 403V at 100% as expected)

My TeslaLogger is recording the Voltage vs. SoC on every charge, basically 24/7. So I can share this data when I have done more driving and charging.

didnt find the guide for that

https://youtu.be/_MawITCEg6U 12

• First I filmed between the bottom cover (lower aero cover) and the lower control arm. Can see the sticker but the wiring harness is in the way. (6-20 seconds in the attached video)
• Next I go just above the control arm, and I try to swivel and reposition the camera several times. Spend 5+ seconds for each angle to let the autofocus do its magic.
• Wheel turned all the way to the left, go in behind the RIGHT front wheel.
• I’m holding the phone in portrait mode, barely need to tilt it to move around.
• You can see on the screen when you get above the arm, but I don’t think you can see the screen while the camera is pointed at the sticker.
• Start with the phone vertical, screen turned directly toward the tire/right side of the car.
• I would say tilt the phone about 15-20 degrees to the left, swivel the screen slight toward the front of the car, and angle the camera a bit backwards (away from you)

I didnt bother checking my sticker on the Q4/2020 Performance as I was just happy with the way this pack was performing overall and there was no doubt about its cells and capacity to me. Whatever @AAKEE finds on his sticker, must be true to my old one as well. After all our VIN's and production dates were pretty close together I belive (845### / 19.11.2020).

 

[AlanSubie4Life]

I also do not think Tesla/Panasonic has a lot of different 2170L cells.

The evidence is circumstantial (variability in how quickly capacity loss shows, etc.), but even with the old cells, I think there was about 1-2kWh of variability in the maximum nominal full pack.

Maybe we are seeing something like that here. I don’t know.

I guess we’ll see what the data says. Lots of reports now of packs with numbers in the 79.x kWh range. Will see how they do, if people report. If it is just BMS, SOME of these should go to ~81kWh if they are not limited, or selected for low capacity in some way.

didnt find the guide for that by a brief search, I did se and read it some months ago…

MASTER THREAD: 2021 Model 3 - Charge data, battery discussion etc

My TeslaLogger is recording the Voltage vs. SoC on every charge, basically 24/7.

Cool. As long as it has plenty of open-circuit voltage points that seems like it should be useful…if we can find some data to compare to at 5% SOC or so (hopefully can get one digit after decimal precision on the SOC too…)

 

[AAKEE]

I have some new interresting data.

Remember my car more or less always sit in the garage, except when at work. The garage is well isolated to withstand -30 to -45 C. I keep some 12C ( + … ) during winter and its warmer now but it is much colder than outside the hot days.
I had 81.4kWh NFP for a long time, it have been down 81 the latest 7 days or so after I did some charges to higher SOCs a couple of times.
Thursday evening I left the car outside the garage, south side/in the sun. I drove about 10km yesterday, and left it outside again. It was charged to 60% and had about that during the night. We have descent weather with about +25C, and not that cold night either.
When I took the car for a drive today the NFP was 79.8 kWh, all time low for me.
It was a 20km drive and then parked in the shadow for a couple of hours, then I read NFP 80.4 kWh before driving home.
Didnt check the NFP at home after the drive.

I would guess the low NFP is connected to the two days out in the heat.

 

[AlanSubie4Life]

I have some new interresting data.

Remember my car more or less always sit in the garage, except when at work. The garage is well isolated to withstand -30 to -45 C. I keep some 12C ( + … ) during winter and its warmer now but it is much colder than outside the hot days.
I had 81.4kWh NFP for a long time, it have been down 81 the latest 7 days or so after I did some charges to higher SOCs a couple of times.
Thursday evening I left the car outside the garage, south side/in the sun. I drove about 10km yesterday, and left it outside again. It was charged to 60% and had about that during the night. We have descent weather with about +25C, and not that cold night either.
When I took the car for a drive today the NFP was 79.8 kWh, all time low for me.
It was a 20km drive and then parked in the shadow for a couple of hours, then I read NFP 80.4 kWh before driving home.
Didnt check the NFP at home after the drive.

I would guess the low NFP is connected to the two days out in the heat.

I guess that is about a 1.2kWh swing.
Not sure how much randomness in the BMS to expect. Seems like heat might have something to do with it. Kind of surprisingly large I guess. Maybe this is mostly BMS! Haha, we will see.

I’m headed to 46C temperatures (in Oregon and Washington! - insane) in a few days so it will be interesting to see how the Superchargers do and what the battery says (my old 2018 P pack is 10% degraded so it should be easy to see pack capacity variability from the rated miles at 100%). Curious how much it varies. Seems pretty rock solid right around 287 rated miles at the moment.

 

[Jejunjm]

The evidence is circumstantial (variability in how quickly capacity loss shows, etc.), but even with the old cells, I think there was about 1-2kWh of variability in the maximum nominal full pack.

Maybe we are seeing something like that here. I don’t know.

I guess we’ll see what the data says. Lots of reports now of packs with numbers in the 79.x kWh range. Will see how they do, if people report. If it is just BMS, SOME of these should go to ~81kWh if they are not limited, or selected for low capacity in some way.


MASTER THREAD: 2021 Model 3 - Charge data, battery discussion etc


Cool. As long as it has plenty of open-circuit voltage points that seems like it should be useful…if we can find some data to compare to at 5% SOC or so (hopefully can get one digit after decimal precision on the SOC too…)

So my battery nominal started out at 79.x. (well when I gave you the data at about 2k miles) would charging it to 99% give an interesting/useful data point. Perhaps same questions over: how is nominal battery calculated? Is this what determines true battery degradation?

 

[AlanSubie4Life]

So my battery nominal started out at 79.x. (well when I gave you the data at about 2k miles) would charging it to 99% give an interesting/useful data point. Perhaps same questions over: how is nominal battery calculated? Is this what determines true battery degradation?

If you happen to be able to, charging to 100% would be interesting (just drive it right after or use the heat to bring it down).

I personally believe there is a strong correlation between reported NFP and the capacity loss of the pack.

Tesla does a lot of work to make the calculation of the NFP accurate (it needs to be!). I don’t know the details of exactly how that is done, but they need the CAC and they need the voltage curve for that constant current draw.

 

[AAKEE]

So my battery nominal started out at 79.x. (well when I gave you the data at about 2k miles) would charging it to 99% give an interesting/useful data point. Perhaps same questions over: how is nominal battery calculated? Is this what determines true battery degradation?

If anything you should charge to 100%, then the BMS has a true datapoint for the high end.
The BMS tries to estimate what the battery capacity is but it will be a guesstimate.
True capacity/degradation can only be measured by a full charge and the measure the delivered capacity until the low limit( = car stops).

Alansubie bet me on speed

 

[Jejunjm]

If anything you should charge to 100%, then the BMS has a true datapoint for the high end.
The BMS tries to estimate what the battery capacity is but it will be a guesstimate.
True capacity/degradation can only be measured by a full charge and the measure the delivered capacity until the low limit( = car stops).

Alansubie bet me on speed

Ok charging to 100....I'll screenshot. Anything besides nfp?

 

[conv90]

I have some new interresting data.

Remember my car more or less always sit in the garage, except when at work. The garage is well isolated to withstand -30 to -45 C. I keep some 12C ( + … ) during winter and its warmer now but it is much colder than outside the hot days.
I had 81.4kWh NFP for a long time, it have been down 81 the latest 7 days or so after I did some charges to higher SOCs a couple of times.
Thursday evening I left the car outside the garage, south side/in the sun. I drove about 10km yesterday, and left it outside again. It was charged to 60% and had about that during the night. We have descent weather with about +25C, and not that cold night either.
When I took the car for a drive today the NFP was 79.8 kWh, all time low for me.
It was a 20km drive and then parked in the shadow for a couple of hours, then I read NFP 80.4 kWh before driving home.
Didnt check the NFP at home after the drive.

I would guess the low NFP is connected to the two days out in the heat.

Do you know if your P is 00-T o 00-P?
It's really curious that all of you, you see high variations on NFP and me I see always 79,5 79,6 79,4 79,7.

 

[AAKEE]

Ok charging to 100....I'll screenshot. Anything besides nfp?

Yes, Nominal remaining will show how much capacity the BMS think is actually stored in the battery.
NFP = nominal full pack = BMS thinks ”probably have capacity about this size”

Nominal remaining= BMS calculation of what actually is stored at that point. That is more i terresting at a 100% charge, but note both.

Look at my last full charge, NFP 81.6 but nominal remaining 82.0.

 

[AAKEE]

Do you know if your P is 00-T o 00-P?
It's really curious that all of you, you see high variations on NFP and me I see always 79,5 79,6 79,4 79,7.

I did never care to check the sticker. Did get SMT right away and after I connected it, I saw the 82.1kWh full pack when new. That did it for me.

I have seen very steady numbers. Usually connects SMT when I arrive at home or sometimes before leaving work. I did have 80.1 for a long time during the coldest days and then it i creased about 0.1 per day to 81.4…that one was rock solid for at least a month.( except for temporary increase to 8-.5 and 81.6 at two 100% charges).
It decreased to 81.0 after some higher than usual 56-60% charges, and was rock solid at 81.0 for maybe a week before I put it outside for two nights in warm weather.
Now parked inside and will continue to see where the NFP goes.
Its so nice weather that I have about 20C in the garage now, so the datk and cold hideaway is of until the good part of the summer ends( which isnt that long up here )
No, I’d say my NFP usually is rock solid.

But for the (unhappy) persons seeing below 80kWh, I think my below 80 value today should be seen as a proof that ambient temps might be the thing causing part of the issue. I am not one single percent worries that my battery only can hold 79.8kWh as I read earlier.
BMS imperfection in calcsshould be the thing.

 

[AlanSubie4Life]

was rock solid at 81.0 for maybe a week before I put it outside for two nights in warm weather.

It's really a bit odd that at a Supercharger where the battery is substantially warmed, usually capacity increases are observed, yet you saw your car sit outside in the heat and the capacity went down. It's not exactly the same since in one case the battery is being charged and it's a different estimation, while in the other the contactors are open and the BMS is doing its estimations, but still it's strange. In general warmer temps do seem to mean a pack with higher capacity...but oddly not in your case.

Just to be clear: 1-2 kWh of error/variation isn't that surprising to me (that's completely tolerable with the way Tesla includes a buffer, continuously readjusts estimates, etc.) It's the mean value of the estimates that we're seeing on these new vehicles when hooked up to SMT which puzzles me, vs. prior pack readings (like @AAKEE , the original @eivissa vehicle, and a few other reports linked to by @FredMt , and other people I may have missed who have contributed).

I do expect values to bounce around, but around some value that is probably the rough "true" capacity (no such thing as "true" capacity really exists since the pack capacity changes a bit all the time based on environmental conditions).

But these errors and environmental factors are generally pretty small. But they are on the same order as the difference in apparent average pack capacities 81kWh vs. 79kWh, so it does make it difficult to tell which factor we are seeing here. Hence the whole debate about whether it's a pack difference or an estimation issue.

The answer is always "more data," of course.

Getting SMT data from US owners (along with a reminder about vehicle type and purchase date of course) would be nice, as well as from European owners.

 

[AAKEE]

It's really a bit odd that at a Supercharger where the battery is substantially warmed, usually capacity increases are observed, yet you saw your car sit outside in the heat and the capacity went down. It's not exactly the same since in one case the battery is being charged and it's a different estimation, while in the other the contactors are open and the BMS is doing its estimations, but still it's strange. In general warmer temps do seem to mean a pack with higher capacity...but oddly not in your case.

Just to be clear: 1-2 kWh of error/variation isn't that surprising to me (that's completely tolerable with the way Tesla includes a buffer, continuously readjusts estimates, etc.) It's the mean value of the estimates that we're seeing on these new vehicles when hooked up to SMT which puzzles me, vs. prior pack readings (like @AAKEE , the original @eivissa vehicle, and a few other reports linked to by @FredMt , and other people I may have missed who have contributed).

I do expect values to bounce around, but around some value that is probably the rough "true" capacity (no such thing as "true" capacity really exists since the pack capacity changes a bit all the time based on environmental conditions).

But these errors and environmental factors are generally pretty small. But they are on the same order as the difference in apparent average pack capacities 81kWh vs. 79kWh, so it does make it difficult to tell which factor we are seeing here. Hence the whole debate about whether it's a pack difference or an estimation issue.

The answer is always "more data," of course.

Getting SMT data from US owners (along with a reminder about vehicle type and purchase date of course) would be nice, as well as from European owners.

We know from research that high temperatures degrade batterys in long term.
For shorter periods and at lower SOC the batteries doesnt degrade that much.

But this behaviour with the NFP shouldnt be about the battery cells but how the BMS calculates the NFP. Theres probably a formula taking the ambient temperature in to consideration. I think eivissa showed data for the NFP going up and down. I havent seen that behaviour, yet. Or st least before today. Maybe I havent checked the NFP frequent enough, bit I mainly checked it at least once a day and only did see very steady results. Today was the first time there was two different values with more than 0.1 kWh change. Maybe a coincidence, but the sole new factor was the ambient temperature.

Will be nice to se what happens to my NFP after a few days(& nights) in the garage at stable lower temps than in the hot sun.

 

[Jejunjm]

Had to leave and it still said 30 minutes left at 100%(7kw charger).

 

[AlanSubie4Life]

Had to leave and it still said 30 minutes left at 100%(7kw charger).

Thanks. And you have an LR, not a Performance. Your second screen shot unfortunately got covered by the volume, but anyway it's probably high 79s for the NFP and nominal remaining is 80.1kWh.

What were your rated miles displayed at 100%? Still 353 or maybe 354 miles I assume? (That's what SMT says but not sure if it pulls that from the car or calculates it somehow).

Seems like a decent result, in any case. Definitely way better than the old-style packs (by about 2kWh).

 

[Jejunjm]

Thanks. And you have an LR, not a Performance. Your second screen shot unfortunately got covered by the volume, but anyway it's probably high 79s for the NFP and nominal remaining is 80.1kWh.

What were your rated miles displayed at 100%? Still 353 or maybe 354 miles I assume? (That's what SMT says but not sure if it pulls that from the car or calculates it somehow).

Seems like a decent result, in any case. Definitely way better than the old-style packs (by about 2kWh).

Sorry was on a rush. Nominal remaining was 80.3 and nfp 79.9. I assume nominal remaining might have gone up if I had let it keep charging...

Rated miles on the car display was 354.

 

[AlanSubie4Life]

Sorry was on a rush. Nominal remaining was 80.w and nfp 79.9. I assume nominal remaining might have gone up if I had let it keep charging...

Rated miles on the car display was 354.

No problem. Anyway, at face value it looks like a somewhat lower pack capacity than a few (but not all!) of the Performance vehicles we've seen. But a very healthy result for the LR, based on what we've seen so far.

We can speculate about what might have happened if you had left it charging...but we'll never know, and that's ok. We can also speculate about what might happen if you modify charging regimen, etc. (as others have done). But no one will know until you do it of course. You can see what happens and report here, if you wish. Anyway, you have a baseline and you can just see what happens.

BTW, out of curiosity, was the destination you navigated to downhill for a bit? Trying to make sure I understand why that projection screen was pinned at 100%. Historically I didn't think they just pinned the SOC like that, but that projection looks like they are - or the destination was downhill for the first 5-7 miles or so. Trying to figure out which it was. You're in Tucson, I thought, so from what I know about that and what I think are palo verde in the background (but seems like too many trees for Tucson - so confused!!!), it probably was not downhill. So that would imply some pinning. Which confuses me. I can actually kind of understand why the energy screen might LOOK that way - but I am curious whether the actual results would match that projection.