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

[kxts]

Your battery was too cold to get near 250kW. It'll get above 200 kW in the summer. It was acting normal for having been preheated in winter.

Not enough data to make that conclusion. Definitely don't use a single Supercharging experience to make that determination. The Supercharging profile of the Panasonic battery has not been adequately characterized.

This is a good point, though I was also basing this on the LG batteries supposedly being 74.5kWh usable (unless I’m misunderstanding that.) Both from the math that @AlanSubie4Life did above and looking at my 12->90% supercharging session that added 60.7kWh (== 77.8kWh extrapolated to 100%.)

One of my takeaways from the comparison graph is that maybe for the examples used there, Tesla just hadn’t unlocked the maximum supercharging rate for LG. Maybe more safety testing was needed on the LG batteries to charge faster and in reality the two types are not very different.

Hopefully frunk disassembly yields some concrete data to confirm all this guesswork and speculation.

 

[AlanSubie4Life]

As far as the road trip itself, I was quite satisfied with the range! I made it from Portland to Sisters (actually halfway between Sisters and Bend, so a few more miles) with 41%, which is exactly what the nav system predicted. Then I made it from Sisters to the supercharger in Salem with 12% left, after a starting prediction of 11%. I could’ve actually made it to the supercharger in Woodburn with 6% or so, but that was too risky and I wanted to try out 250kW charging. Total distance was 278 miles for that 88% of charge, so 316 miles of range extrapolated to 100%.

Great, glad it went well. That result is really very, very good. Not surprisingly one of the best results I've heard of. I think the heat pump helped you considerably (it's close to an optimal temperature for it). In addition the slow travel in the mountains helps. Those fast interstate runs really eat the range. As long as you're going a reasonable pace and it's not an extreme grade with too much regeneration going downhill, the fact that it was mountainous has very little impact on your range. Good to see you did so much better than the ABRP prediction (sounds like they are being conservative for now).

It was relatively (completely?) dry for you, too, and that helped you a lot, too. Wet roads make a considerable difference. Don't expect this sort of result if it's raining (for your planning purposes)!

Not enough data to make that conclusion.

Do we know whether US vehicles will receive something other than Panasonic? I trust your info on this...I haven't been following what is planned to go where.

What info are people using to make a determination (in the US) of which vendor's battery they have? In the EU it sounds like it's pretty clear when they have an LG battery as it has lower initial capacity (maybe - the data on this to me is very unclear which battery is which as I haven't seen it all summarized with data in one place). There is definitely a battery which is currently limited to a little over 74kWh in the EU.

I actually figured all batteries in the US would be Panasonic (not saying that as a fact...it's just a guess). They're going to have to be all the same capacity in the US eventually, for sure. Otherwise the current EPA test result is not valid. And it would seem weird to put different capacity batteries in the same trim level.

@kxts 's battery is definitely around 78kWh (currently usable...no idea if there is some capacity that is locked out). I have no idea which manufacturer it's from though. Again, I thought in the US they were all Panasonic?

 

[AlanSubie4Life]

looking at my 12->90% supercharging session that added 60.7kWh (== 77.8kWh extrapolated to 100%.)

Be careful about drawing this conclusion from the session. It is correct to do extrapolation as you used it (since the errors cancel), but the "dirty secret" (it's not a secret) is that you actually only added 0.955*60.7kWh = 58kWh of usable energy to your battery. Why? Because the claim (which is well supported) is that 4.5% of your battery is below 0%. So 0%-100% only contains 78kWh*0.955, 74.5kWh (this has no relation to the 74.5kWh from EU batteries, totally unrelated). So charging up 78% of that 74.5kWh is 58kWh (not 60.7kWh). I'm not sure whether the kW rate reported by the Supercharger is accurate or not (it would not be easy at all to see that miles are not being added at quite the rate it says it is charging at, so the kW reporting may well be correct). I don't know how Tesla charges for this, but the 4.5% discrepancy is small and they do have to pay for AC-DC conversion losses, so it is fair enough (in the end they're charging you for less energy than they draw from the utility, not more). This displayed number also doesn't count heat losses in the battery which Tesla also has to pay for - it's probably best to think of the number as a reflection of how many rated miles are added to the pack. Abstracts from the energy quantity a bit (though it can be calculated as above).

Total distance was 278 miles for that 88% of charge,

Case in point, I assume that the trip meter said 0.88*0.955*0.99*78kWh / 278mile = 233Wh/mi? (Which would imply 64.9kWh used...not 68.6kWh). The 0.99 is to account for small inaccuracy in the trip meter (seems to read very slightly low).

You would have had to be tracking the trip meter data to see this of course. And you would have had to lose no energy in Park at Sisters. (So I'd expect maybe a slightly lower result than 233Wh/mi - which is excellent, in any case.)

 

[TimothyHW3]

Not enough data to make that conclusion.

More data could be better, but since all charging sessions with the LG drop below 100kW in about the 35-40% SOC range, I think this indeed is a Panasonic battery. Also, the 555km range shown is another indication - the LG ones are missing 3kWh which is about 15 or so missing miles. Also, green on Twitter reported that the new upgrade 44.x.something brought a "new pack" that was limited to the 332 miles. I think that is enough evidence to conclude that he indeed has the Panasonic one. Maybe Tesla used the LG ones for export only, because EPA is more strict in the US.

 

[Zoomit]

Again, I thought in the US they were all Panasonic?

That’s what I expect as well. At least for current builds.

I think that is enough evidence to conclude that he indeed has the Panasonic one

Yes, sorry I got my LG and Panasonic batteries confused. It likely a Panasonic battery assembled in GF1, just like all the other NA Model 3/Ys.

 

[TimothyHW3]

my 12->90% supercharging session that added 60.7kWh (== 77.8kWh extrapolated to 100%.)

You can't do the math like that on a Tesla I am afraid. If you look at the screen - it is not accurate as it only counts added kms and transfers back to kWh, it doesn't measure kWh added.

If you look at your Tesla.com account history for each supercharging added, the value there includes charging loss and battery heating from stators which depending on the charge time is about 6%-8%. And is also rounded up.

Also, your 12%-90% is not 78% from 77.8 actually, but rather 78% from 77.8-3.5kWh buffer.

The only way to know is to read the Can Bus with scan my tesla.

 

[TimothyHW3]

Yes, sorry I got my LG and Panasonic batteries confused. It likely a Panasonic battery assembled in GF1, just like all the other NA Model 3/Ys.

Are they though? We can't be sure. Plus, which Panasonic one - the old one or the new one. Might be the old one, or a new one capped at 77.5kWh to reduce degradation.

 

[kxts]

Be careful about drawing this conclusion from the session. It is correct to do extrapolation as you used it (since the errors cancel), but the "dirty secret" (it's not a secret) is that you actually only added 0.955*60.7kWh = 58kWh of usable energy to your battery. Why? Because the claim (which is well supported) is that 4.5% of your battery is below 0%. So 0%-100% only contains 78kWh*0.955, 74.5kWh (this has no relation to the 74.5kWh from EU batteries, totally unrelated). So charging up 78% of that 74.5kWh is 58kWh (not 60.7kWh).

Case in point, I assume that the trip meter said 0.88*0.955*0.99*78kWh / 278mile = 233Wh/mi? (Which would imply 64.9kWh used...not 68.6kWh). The 0.99 is to account for small inaccuracy in the trip meter (seems to read very slightly low).)

You can't do the math like that on a Tesla I am afraid. If you look at the screen - it is not accurate as it only counts added kms and transfers back to kWh, it doesn't measure kWh added.

If you look at your Tesla.com account history for each supercharging added, the value there includes charging loss and battery heating from stators which depending on the charge time is about 6%-8%. And is also rounded up.

Also, your 12%-90% is not 78% from 77.8 actually, but rather 78% from 77.8-3.5kWh buffer.

Unfortunately I forgot to reset the trip meter, and the Teslamate data has a gap so I can’t calculate from that easily (looks like from where cell service was spotty in the mountains.)

Good to know on the charging data. Just started looking into all this a few days ago and there’s much to learn. This data comes from Teslamate. I’m not sure if it’s pulling that 60.7 value directly from the API or calculating it from other values...anyone know offhand? You’d expect it to match the number in my account history, yet that number is slightly higher at 62 kWh. Accounting for 1 significant figure of rounding in my account, that gives us a range of 1.3 - 2.7% less for the Teslamate number...so doesn’t match up with the 4.5% number or charging loss. I’ll dig into the source code if nobody knows and figure it out.

 

[AlanSubie4Life]

This data comes from Teslamate. I’m not sure if it’s pulling that 60.7 value directly from the API or calculating it from other values...anyone know offhand?

Not sure. Looks pretty comprehensive, so could be calculated from the input current and voltage, and integrated over time, but I have no idea at all what they are doing. If they did calculate that way, they could definitely come up with a roughly correct answer that is higher than the useful energy added (which is what seems to have happened).

They could not capture losses in the cable that Tesla is going to charge you for. So would be generally slightly lower than Tesla’s number even if done perfectly.

Bur depending on the method there could be some slop here, so have to only expect rough alignment.

Accounting for 1 significant figure of rounding in my account, that gives us a range of 1.3 - 2.7% less for the Teslamate number...so doesn’t match up with the 4.5% number or charging loss. I’ll dig into the source code if nobody knows and figure it out.

If I understand correctly you were charged for 62kWh. And above we calculated ~58kWh of useful energy added.

That seems about right after accounting for the stator heating, and DC charging losses. Stator heating is on the order of 7kW max (can look at YouTube or other SMT screen captures to see the actual max). So could account for ~2kWh there.

As @TimothyHW3 reiterated above, you can safely ignore the “kWh added” on the charging screen. As we said, it’s just directly calculated from miles added * charging constant. So it is not a measurement of energy added, and should not be expected to be correct.

Unfortunately I forgot to reset the trip meter, and the Teslamate data has a gap so I can’t calculate from that easily (looks like from where cell service was spotty in the mountains.)

It’s ok. You should be able to arrive at the same conclusion in future, even for shorter trips around the city. Just takes a decent length trip (25-50 miles) to start to see how the trip meter relates to rated miles use, and see the 4.5%/95.5% factor.

 

[jaqueh]

Be careful about drawing this conclusion from the session. It is correct to do extrapolation as you used it (since the errors cancel), but the "dirty secret" (it's not a secret) is that you actually only added 0.955*60.7kWh = 58kWh of usable energy to your battery. Why? Because the claim (which is well supported) is that 4.5% of your battery is below 0%. So 0%-100% only contains 78kWh*0.955, 74.5kWh (this has no relation to the 74.5kWh from EU batteries, totally unrelated). So charging up 78% of that 74.5kWh is 58kWh (not 60.7kWh). I'm not sure whether the kW rate reported by the Supercharger is accurate or not (it would not be easy at all to see that miles are not being added at quite the rate it says it is charging at, so the kW reporting may well be correct). I don't know how Tesla charges for this, but the 4.5% discrepancy is small and they do have to pay for AC-DC conversion losses, so it is fair enough (in the end they're charging you for less energy than they draw from the utility, not more). This displayed number also doesn't count heat losses in the battery which Tesla also has to pay for - it's probably best to think of the number as a reflection of how many rated miles are added to the pack. Abstracts from the energy quantity a bit (though it can be calculated as above).



Case in point, I assume that the trip meter said 0.88*0.955*0.99*78kWh / 278mile = 233Wh/mi? (Which would imply 64.9kWh used...not 68.6kWh). The 0.99 is to account for small inaccuracy in the trip meter (seems to read very slightly low).

You would have had to be tracking the trip meter data to see this of course. And you would have had to lose no energy in Park at Sisters. (So I'd expect maybe a slightly lower result than 233Wh/mi - which is excellent, in any case.)

What is the usable capacity of the pre-refresh Model 3? I've seen many say that it's 73.5 kWh? So, does it seem like Tesla added 1kWh more of usable space to the battery no matter the lottery if they are reporting that range to the EPA?

Also wondering, what kind of miles I should expect to see on delivery day at 90% (if this is how much they charge to).

 

[AlanSubie4Life]

What is the usable capacity of the pre-refresh Model 3? I've seen many say that it's 73.5 kWh?

73.5kWh is very very wrong for pre-refresh Model 3. (Unless you go all the way back to the 2017/2018 RWD before capacity unlock from 310rmi to 325rmi.). However, "usable" capacity is a debatable term - usable (above 0%) was right around that value, as measured on the trip meter for new 2020 AWD vehicles. I prefer not to use that definition of usable though it is a perfectly reasonable one.

The usable (my preferred definition - how much you can get out of it before shutdown) capacity of the pre-refresh Model 3 was 79.5kWh. As per SMT, it was somewhere around 78kWh - you can't directly compare these numbers as the kWh may be "different" (measurement differences). (Showed degradation below 77.6kWh for 2020, and below 76kWh for 2018/2019.)

The new 2021 US AWD Model 3 had 78.6kWh in the EPA test (but we have just one data point and there is variability so this could have been an outlier on the low side). Per SMT I would expect it to show about 78kWh when new (I have not seen such info yet). It seems the degradation threshold is still around 77.6 or 77.7kWh; need more data to be sure. (347rmi*139Wh/rkm*1.6093rkm/mi = 77.6kWh)

So to first order, for now, it doesn't look like there is any capacity change on the AWD from 2020 to 2021.

The new 2021 US Performance Model 3 had about 80.8kWh in the EPA test. I expect it to show about 79.5kWh in SMT when new. No data yet. This is way above anything we have seen in the EPA test before, so it's very likely the new denser cells are included in packs in the Performance vehicles. And there may be more to unlock still; I imagine Tesla is being conservative.

Also wondering, what kind of miles I should expect to see on delivery day at 90% (if this is how much they charge to).

Looks like right now (with limited information) they are charging to 347. I expect this to change to 353 at some point soon (after a software update to change the constant - not increase capacity - but we'll see), for new vehicles. Vehicles already delivered may never show 353 even after the update because of capacity loss between now and whenever this hypothetical software update occurs. We'll see what happens. It's normal for (eventually) the constant to be set such that when the vehicle battery is relatively new, it shows the EPA range (in this case 353 rated miles). But we also had a similar situation a year ago where the 2020 vehicles briefly showed 310 rated miles at full charge rather than 322. The software update changed the constant and aligned the numbers (no capacity change).

A lot of moving pieces here so it's also possible there could be a small capacity unlock which gives the 353 rated miles in addition to the constant change. We're currently at a ~2% discrepancy.

 

[TimothyHW3]

I've seen many say that it's 73.5 kWh? So, does it seem like Tesla added 1kWh more of usable space to the battery no matter the lottery if they are reporting that range to the EPA?

It was 77.5/77.8+kWh when brand new. Now we have two variants - one reporting 77.5 (probably capped) and one (probably LG) at 74.5kWh (which also sometimes delivered with 1+% degradation out of the box at around 73-73.5)

 

[jaqueh]

73.5kWh is very very wrong for pre-refresh Model 3. (Unless you go all the way back to the 2017/2018 RWD before capacity unlock from 310rmi to 325rmi.). However, usable capacity is a debatable term - usable (above 0%) was right around that value, as measured on the trip meter for new 2020 AWD vehicles.

The usable capacity of the pre-refresh Model 3 was 79.5kWh. As per SMT, it was somewhere around 78kWh - you can't directly compare these numbers as the kWh may be "different" (measurement differences). (Showed degradation below 77.6kWh for 2020, and below 76kWh for 2018/2019.)

The new 2021 US AWD Model 3 had 78.6kWh in the EPA test (but we have just one data point and there is variability so this could have been an outlier on the low side). Per SMT I would expect it to show about 78kWh when new (I have not seen such info yet). It seems the degradation threshold is still around 77.6 or 77.7kWh; need more data to be sure. (347rmi*139Wh/rkm*1.6093rkm/mi = 77.6kWh)

So to first order, for now, it doesn't look like there is any capacity change on the AWD from 2020 to 2021.

The new 2021 US Performance Model 3 had about 80.8kWh in the EPA test. I expect it to show about 79.5kWh in SMT when new. No data yet. This is way above anything we have seen in the EPA test before, so it's very likely the new denser cells are included in packs in the Performance vehicles. And there may be more to unlock still; I imagine Tesla is being conservative.



Looks like right now (with limited information) they are charging to 347. I expect this to change to 353 at some point soon (after a software update to change the constant - not increase capacity - but we'll see), for new vehicles. Vehicles already delivered may never show 353 even after the update because of capacity loss between now and whenever this hypothetical software update occurs. We'll see what happens. It's normal for (eventually) the constant to be set such that when the vehicle battery is relatively new, it shows the EPA range (in this case 353 rated miles). But we also had a similar situation a year ago where the 2020 vehicles briefly showed 310 rated miles at full charge rather than 322. The software update changed the constant and aligned the numbers (no capacity change).

A lot of moving pieces here so it's also possible there could be a small capacity unlock which gives the 353 rated miles in addition to the constant change. We're currently at a ~2% discrepancy.

Ah I see then that this confirms the smaller battery pack, but heat pump and maybe larger rear motor efficiencies to the new model 3. I'm really wondering if I should get the new LR AWD or a used LR. Being in CA, I drive around 80mph on the freeway, so real world range doesn't seem like it would be that greatly improved with the new model 3.

 

[AlanSubie4Life]

Now we have two variants - one reporting 77.5 (probably capped) and one (probably LG) at 74.5kWh (which also sometimes delivered with 1+% degradation out of the box at around 73-73.5)

To clarify: In the EU. This is not happening in the US (so far).

Ah I see then that this confirms the smaller battery pack,

No, there's little to no evidence the pack is smaller for the US AWD 2021 vs. the US AWD 2020. See above. The degradation threshold appearing to be the same suggests that the target new pack capacity is the same. Even if the EPA test ended up getting 0.9kWh less out of the pack (it varies!).

being in CA, I drive around 80mph on the freeway, so real world range doesn't seem like it would be that greatly improved with the new model 3.

Yeah, not much, but it's a little hard to say right now. If there is some improvement in the rear motor efficiency (we don't know) that would help the AWD a little. At 80mph the heat pump will be less helpful but it will still help you a lot on the rare cooler (40-50F) days (and it appears on warmer days as well - it appears AC is a little more efficient from the EPA results but I haven't really looked at that in detail).

It definitely doesn't seem like it's worth an upgrade for the changes, in California. For freeway runs in California you'll only using 60-70% of the battery capacity anyway so it seems like it's just going to save you a few minutes per trip at best.

 

[jaqueh]

It was 77.5/77.8+kWh when brand new. Now we have two variants - one reporting 77.5 (probably capped) and one (probably LG) at 74.5kWh (which also sometimes delivered with 1+% degradation out of the box at around 73-73.5)

Yeah I've heard the woes about the LG car, but those don't seem to be (at least I hope) to ever hit the US market.

 

[TimothyHW3]

rear motor efficiencies to the new model 3

There is no motor efficiency so far from what I have seen. Early tests indicate that motor efficiency is a no-no at least vs 2019/20 models (I will try to run my own test)
Heat pump has some effect, but only if you expect to do short commutes in -5C to 5C weather, otherwise no or a small no.

 

[AlanSubie4Life]

There is no motor efficiency so far from what I have seen. Early tests indicate that motor efficiency is a no-no at least vs 2019/20 models (I will try to run my own test)

Probably best to identify whether there has been a change in the rear motor (to something like 980, or something other than 990 at least). Has anyone actually tried this yet? It's easy to determine!

If it's a 990, I wouldn't expect much significant change and any efficiency improvements (which did seem apparent in the EPA raw results - see the EPA summary post from me on that 2020 AWD is 181Wh/mi on the dyno test, 2021 AWD is 176Wh/mi) are due to other more subtle changes/drivetrain improvements, tire changes, etc. Again: the highway result for the 2021 AWD nearly exactly matches the 2020 Performance 18" result (in Wh/mi - 2020 Performance was 175Wh/mi). But the reasons for that are unclear.

Running tests on the road has so many variables it would be near impossible to identify a 3% efficiency change (highway). You'd basically have to take two models with identical new tires, identical climate control settings (off), and run them literally side by side on the same highway at the same time for a decent distance, to have a chance. And you couldn't do it with any traffic on the road at all since that would affect aero losses for the two vehicles differently. And you'd have to be sure that the highway you're testing has an identically worn road surface in both lanes (rare). Anyway, in the city the improvement was more like 6% in EPA testing. Anyway, much better to just look at the EPA dyno results. Of course this is also affected by things like the coefficients they set, etc. But at least it is easier to identify what the variables are.

 

[EV Promoter]

Hi @AlanSubie4Life , thanks for the explanations.

Yesterday i've got by chance the two lines of the average and rated consumption perfectly superimposed @142 Wh/km.
The constant is absolutely confirmed @139 Wh/km, see graph.
The % of charge were 64%, just down from 65%, unfortunately it gives no decimals, should have been something around 64,5% giving the 535 km as previusly ascertained.

Question: where i can find some label/code of the battery on the car? Ready to do some removal.

I've succefully sent a token through Tesla.info, but the codes are all over the place and incorrect.

 

[AlanSubie4Life]

where i can find some label/code of the battery on the car?

Check the post linked here (MASTER THREAD: 2021 Model 3 and differences from 2020), and the post above it. You need a bright flashlight to place on the top of the aero shield, and a decent phone camera. Turn the front wheels to the LEFT.

Place the light on the top of the lower front aero shield, illuminating the middle front edge of the battery clearly.

From behind the front right wheel, poke an iPhone 12 mini into the bowels of the vehicle (go high, above the suspension members, reach in pretty far). Set 5x zoom. Reach way in, angle the camera toward the front of the battery and focus it just above the coolant pipe, on the label. If you look carefully, you'll be able to just see the label on your phone screen (would be easy if the wheel were off but you'll be at an extreme angle). Take a picture (be sure to click to focus).

Note: my vehicle is a 2018, and the coolant pipes may be routed differently on the 2021. I assume they will not block the label. Note the label is different (part number high up) on the 2021 label. So you should be able to capture the part number (below the serial number).

Assuming the pipe is not blocking the part number (go high with that phone - this won't work on prior year vehicles but looks like it will on 2021), I don't think you need to take anything apart.


My part number ends in "M," lol (probably all I need to know). You won't have this problem; this is no longer where the part number is located on the label, AFAIK. This is the "low angle" shot - you don't want this one - it's taken with the front wheels turned to the RIGHT, from in front of the wheel. Not useful for you.

 

[TimothyHW3]

and rated consumption perfectly superimposed @142 Wh/km.
The constant is absolutely confirmed @139 Wh/km, see graph.

You see how the 337km and 344km diviate where usually they should match like on the previous models? Do you wanna know why? This is because you got the LG 74kWh battery. Tesla seems to overreport the rated range on the battery to make it look like you have more range(above the battery icon), but the algorithm for the remaining range uses the accurate calculation (hence 10km difference)

On the 2020 models and the 2021 Panasonic 82 models, these two km ratings match, there is no deviation in battery vs energy rated km left.

I have seen 3 other screenshots like that from these 74 batteries and they all diviate by about 10km.

Don't know how Tesla wants to get away with that honestly, they can either report true range of 520km instead of 535km on these LG batteries or they can lower the graph to below 140Wh/km. A real mess, they never should've
used these bulls batteries from LG in the first place...