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

[inigoml]

to 0.9*52.9kWh - but I write it this way to show the separate factors.) This is because the car does not display the actual energy added to the vehicle battery pack - it displays the number of miles added (in your case 90% of ~253mi (407km), which is 228mi (367km) ), multiplied by the charging constant which is (roughly) 55kWh/263mi = 209Wh/mi (130Wh/km).

Hmmm... Makes sense. If kWh is calculated form range then of course it won't match.

I can assure that I've wasted 50 kWh that is what charger has shown at the end of session. But net energy pushed into battery packet is what Teslafi has told me. Next time I will check it with SMT in order to check for discrepancies.

Thank you, mate!

 

[AlanSubie4Life]

Hmmm... Makes sense. If kWh is calculated form range then of course it won't match.

I can assure that I've wasted 50 kWh that is what charger has shown at the end of session. But net energy pushed into battery packet is what Teslafi has told me. Next time I will check it with SMT in order to check for discrepancies.

Thank you, mate!

Yep, check that; I think you'll find the discrepancy. TeslaFi is just using the API I think for that information so it won't be correct either and should just match the screen in the car (without the rounding of course).

kWh is calculated from rated mile/km delta. If you swap back and forth between display modes of Energy & Distance you'll find the two numbers are always related by the charging constant. And rated miles are directly related to the BMS estimates. So the "energy" number displayed is just a conversion, and it uses the wrong value since displayed rated miles don't have the same energy content as the charging constant (they have 4.5% less energy).

50kWh is about right for energy from the charging station; it works out to 91% efficiency. You'd have to check the EPA documents but it is possible that they have slightly better charging efficiency than prior models. Or it could just be rounding error. Was it 50.0kWh or 50kWh displayed by the charging station?

I checked them and I actually can't find the document for the smaller 55kWh LFP, I might have missed it. But anyway for the 60kWh LFP it still shows 89% efficiency (at 208V, unknown current, 62.047kWh with recharge event energy of 69.598kWh), though they don't seem to have attached their customary spreadsheet, but in any case 90% is close enough. https://dis.epa.gov/otaqpub/display_file.jsp?docid=54389&flag=1

(Here's one for the NCA which shows the calculations more explicitly, though both documents have the info: https://dis.epa.gov/otaqpub/display_file.jsp?docid=54290&flag=1)

 

[AlanSubie4Life]

(For completeness I'll just post the screen captures from above links. Tesla does a great job documenting the actual efficiency of 89%, though it would be nice if they would fully specify the charging conditions...):
RWD (LFP 60kWh), 89.2%:

AWD (NCA, not LFP), 89%:

 

[inigoml]

(For completeness I'll just post the screen captures from above links. Tesla does a great job documenting the actual efficiency of 89%, though it would be nice if they would fully specify the charging conditions...):
RWD (LFP 60kWh), 89.2%:

Just impressive. Impossible to refute your arguments! LFP 55kWh must be quite similar if not equal to LFP 60kWh.
Charging station pushed 50.something kWh. No roundtrips.

 

[AAKEE]

(For completeness I'll just post the screen captures from above links. Tesla does a great job documenting the actual efficiency of 89%, though it would be nice if they would fully specify the charging conditions...):

Yes, I am quite sure that the charging conditions affect the efficiency.

I guess it is AC charging. For me it seems that the higher charging power the less losses. The AC on-board charger probably have lesser losses with 11kW charging power.
I did mount a energy meter( approved for billing during sales of electricity) directly in line before the WC and it seems 11kW can show around 5-6% losses when compared to the both used energy during the cycle before charging and the filled energy in the rated km’s or difference in nolinal remaining.
Lowering the charge power to a few kW will more or less double the ”measured” losses.

 

[AlanSubie4Life]

Yes, I am quite sure that the charging conditions affect the efficiency.

I guess it is AC charging. For me it seems that the higher charging power the less losses. The AC on-board charger probably have lesser losses with 11kW charging power.
I did mount a energy meter( approved for billing during sales of electricity) directly in line before the WC and it seems 11kW can show around 5-6% losses when compared to the both used energy during the cycle before charging and the filled energy in the rated km’s or difference in nolinal remaining.
Lowering the charge power to a few kW will more or less double the ”measured” losses.

I think it is at least 32A/208V charging being used. At some point they had a document which gave the duration of the charge event but I can’t remember where I saw it.

Anyway I haven’t seen any data from any owner showing more than about 91% efficient charging. So I think that’s about as good as it gets. And certainly that optimum would occur at 11.5kW charge rate. Whether or not they are using 48A or 32A during their EPA testing is the question. You’d think they’d use 48A to get the best efficiency possible as it improves the MPGe they can claim.

 

[AAKEE]

I think it is at least 32A/208V charging being used. At some point they had a document which gave the duration of the charge event but I can’t remember where I saw it.

Anyway I haven’t seen any data from any owner showing more than about 91% efficient charging. So I think that’s about as good as it gets. And certainly that optimum would occur at 11.5kW charge rate. Whether or not they are using 48A or 32A during their EPA testing is the question. You’d think they’d use 48A to get the best efficiency possible as it improves the MPGe they can claim.

I have a lot of examples better than 91%, I think I did write about this before here ( I understand of course if you didnt see my posts about this )

Today, I had a charge about 0300 or so this night(from around 30-35% SOC).

I drove to work, the car was parked for about six hours (sentry off, around 15-20C outside) and then I drove home.
The speed was about 55-60mph mostly for the highway part and some slower driving in both ends.

When i arrived at home I did set the car to charge asap(for the sake of this thread).
I had driven 93 km witjh an average of 151Wh/km = 14.043 kWh used.

I took a picture of the for billing approved electrical meter(mounted directly before the WC) the seconds before the charge was started and another when the charging was finished.
The difference was 14.37kWh so this points at about 2.3% loss in all.

I used the Tesla Wall Charger V2 with 11 kW power (three phase 230V x 16A).

I guess that the SOC might not be precisely the same(both said 56% but maybe not the same down to the points), and I probably have the data to look this up later.





I guess that a full charge whreee you let the car sit over the night (EPA protocol if I dokt miss-remember?) will cause a full bsttery balancing, which is done by burning the over voltage cells down will cost some energy. I slso guess that a full battery will have much more self drain over the night than a 55% charge will have at short term basis.

My other measurements have pointed at 10% or more with a cold battery needing heating, or a slow low power charge with 2-3kW also about 10-12%
A regular 11kW charge where not heating pf the battery was needed have been around 5% or slightly less.

My energy meter is new, bought and mounted when the WC was installed.
The energy meter should deliver very precise values per the specs and the approval for charging for electricity.

 

[AlanSubie4Life]

I have a lot of examples better than 91%, I think I did write about this before here ( I understand of course if you didnt see my posts about this )

Today, I had a charge about 0300 or so this night(from around 30-35% SOC).

I drove to work, the car was parked for about six hours (sentry off, around 15-20C outside) and then I drove home.
The speed was about 55-60mph mostly for the highway part and some slower driving in both ends.

When i arrived at home I did set the car to charge asap(for the sake of this thread).
I had driven 93 km witjh an average of 151Wh/km = 14.043 kWh used.

I took a picture of the for billing approved electrical meter(mounted directly before the WC) the seconds before the charge was started and another when the charging was finished.
The difference was 14.37kWh so this points at about 2.3% loss in all.

I used the Tesla Wall Charger V2 with 11 kW power (three phase 230V x 16A).

I guess that the SOC might not be precisely the same(both said 56% but maybe not the same down to the points), and I probably have the data to look this up later.

View attachment 819174
View attachment 819175
View attachment 819176

I guess that a full charge whreee you let the car sit over the night (EPA protocol if I dokt miss-remember?) will cause a full bsttery balancing, which is done by burning the over voltage cells down will cost some energy. I slso guess that a full battery will have much more self drain over the night than a 55% charge will have at short term basis.

My other measurements have pointed at 10% or more with a cold battery needing heating, or a slow low power charge with 2-3kW also about 10-12%
A regular 11kW charge where not heating pf the battery was needed have been around 5% or slightly less.

My energy meter is new, bought and mounted when the WC was installed.
The energy meter should deliver very precise values per the specs and the approval for charging for electricity.

Interesting. I’ve done similar myself and I end up at 90%-92%. (40A 240V)

Anyway, can do a bit further investigation on a longer charge and since you have SMT you should be able to make sure you know the exact endpoints. I guess you have done this - if you have the posts you can point me to them; I have not seen them.



It just does not seem reasonable to expect more than a percent or two difference from something Tesla measures and is critical to their specs (wall to wheels efficiency in this case).

In any case I guess we can just see what the data says.

Good point on the overnight sitting with the EPA protocol. I guess they don’t count energy use during that time (which is silly, they could and should track it!) and that could make efficiencies a couple % lower (I can’t see more than that) than what the actual number is. But that would put us at 91-92% rather than 89%. That potentially explains that discrepancy (people do seem to measure low 90s) which kind of bothered me…

2.3% loss seems way too low. AC-DC is not that efficient (a few % loss), and we know at 11.5kW we are using a couple hundred watts to keep the car awake which is already a couple %. And we aren’t even counting charging losses in the battery itself (isn’t that a couple % - you know that better than I do).

 

[AlanSubie4Life]

Anyway in the interests of this discussion it would be cool if @inigoml retook his test case of 10-100% with SMT data this time. Obviously we would have to trust that the public charger was accurately metering the energy.

It’s certainly true that efficiencies are somewhat higher if you don’t charge to 100%, since the charging slows a bit at the end, and there may be rebalancing too. Could take an interim datapoint at 90% and continue the charge if you have the ability/time.

 

[AAKEE]

Interesting. I’ve done similar myself and I end up at 90%-92%. (40A 240V)

Anyway, can do a bit further investigation on a longer charge and since you have SMT you should be able to make sure you know the exact endpoints. I guess you have done this - if you have the posts you can point me to them; I have not seen them.



It just does not seem reasonable to expect more than a percent or two difference from something Tesla measures and is critical to their specs (wall to wheels efficiency in this case).

In any case I guess we can just see what the data says.

Good point on the overnight sitting with the EPA protocol. I guess they don’t count energy use during that time (which is silly, they could and should track it!) and that could make efficiencies a couple % lower (I can’t see more than that) than what the actual number is. But that would put us at 91-92% rather than 89%. That potentially explains that discrepancy (people do seem to measure low 90s) which kind of bothered me…

2.3% loss seems way too low. AC-DC is not that efficient (a few % loss), and we know at 11.5kW we are using a couple hundred watts to keep the car awake which is already a couple %. And we aren’t even counting charging losses in the battery itself (isn’t that a couple % - you know that better than I do).

While I havent seen 2.3% before, 4-5% is the lowest I think, lithium batteries have very good energy efficiency.
I newer charged asap, and checked values like this before. I always did use a normal night cycle.

This time i did not take any readings direct like before. It was just a brief test. But teslalogger and trslafi perhaps can give more clues. I’ll check when I have the time.

Lithium batteries normally have about 99% energy efficiency.

Delivered energy / charged energy = energy efficiency. I think you could google that and see that the battery cells themself should not be in the 90% ball park.

 

[AlanSubie4Life]

I think you could google that and see that the battery cells themself should not be in the 90% ball park.

Certainly not. But just another % or two. Between that, running the car in idle mode, and AC-DC conversion losses, it seems best case 95% and that assumes excellent AC-DC efficiency of 98% (seems high; I’d guess 96-97%).

Yet Tesla reports 89% consistently. You pointed out one possible reason for the large 6% gap (overnight sitting could lose a % or two).

I’d guess real world for charges to 100% we’d see best case 92% or so. But anyway we’ll see what the data say. I’ve seen claims of 96% but I believe for those people were looking at the screen in the car which obviously would not be correct.

 

[AAKEE]

I havent looked at any numbers for the charge today.

I made a 0-100% Supercharge about one month ago. Arrived with 0.13% which the car adjusted to 0.40% before I had hooked the SuC V2 cable.
Tesla charged me 82kWh and the added energy was 78.39 kWh according to teslafi. But this is the rated miles, including the hidden buffer so it should be 78.39 x 0.955= 74.9kWh added I think.

SMT values:
Nominal remaining before = 3.70kWh
Nominal remaining after = 79.0 kWh
Estimated charged amounth = 75.3 kWh.
This is 8% loss but a supercharging session with high C-charging will cause higher loss than a slower charge. Also, Im quite sure the losses in the SuC is covered by Tesla so we are not charged directly for that.

 

[AlanSubie4Life]

I havent looked at any numbers for the charge today.

I made a 0-100% Supercharge about one month ago. Arrived with 0.13% which the car adjusted to 0.40% before I had hooked the SuC V2 cable.
Tesla charged me 82kWh and the added energy was 78.39 kWh according to teslafi. But this is the rated miles, including the hidden buffer so it should be 78.39 x 0.955= 74.9kWh added I think.

SMT values:
Nominal remaining before = 3.70kWh
Nominal remaining after = 79.0 kWh
Estimated charged amounth = 75.3 kWh.
This is 8% loss but a supercharging session with high C-charging will cause higher loss than a slower charge. Also, Im quite sure the losses in the SuC is covered by Tesla so we are not charged directly for that.

Good data. Totally different than AC charging of course.
Certainly more I^2*R losses and other fan and cooling losses. But no conversion losses in the car and impact of idle mode losses much less significant. So different loss sources.

Tesla not charging explicitly for AC-DC conversion which is interesting to me. Though not a surprise. Obviously they are charging for it implicitly.

 

[Franzie3]

New here. have a 2020 M3LR that I just picked up roughly a month ago with only 8000 miles. From driving nearly 1000 miles so far (few road trips) it appears my 100% range is likely 300 Miles on the nose. it seems. every 10% of battery is equivalent to 30 miles. now if im not mistaken the EPA rnage on this model was 325 and i know that its never accurate or anythign, but im curious where everyone else is battery wise with this similar config (M3LR-aero wheels). are you seeing the 30miles per 10% battery? better or worse?

 

[AlanSubie4Life]

New here. have a 2020 M3LR that I just picked up roughly a month ago with only 8000 miles. From driving nearly 1000 miles so far (few road trips) it appears my 100% range is likely 300 Miles on the nose. it seems. every 10% of battery is equivalent to 30 miles. now if im not mistaken the EPA rnage on this model was 325 and i know that its never accurate or anythign, but im curious where everyone else is battery wise with this similar config (M3LR-aero wheels). are you seeing the 30miles per 10% battery? better or worse?

Yep, your estimated battery capacity is 72.5kWh (69kWh usable), vs 77.8kWh (74.3kWh usable) when it was new.

All is well.

I’m assuming here the “distances” you are talking about here are what the car displays, not what you have actually traveled. 300 actual miles for 100% would be really good!

 

[Franzie3]

Yep, your estimated battery capacity is 72.5kWh (69kWh usable), vs 77.8kWh (74.3kWh usable) when it was new.

All is well.

I’m assuming here the “distances” you are talking about here are what the car displays, not what you have actually traveled. 300 actual miles for 100% would be really good!

Y
I haven't went to a full 100 percent yet. But yea these are just rough math estimates as every 10% battery is 30 miles on the nose. At 95% im at 285 miles so that wxtra 5% will get me id guess 15 miles. I can usually get ~220whm on my commute (75 miles one way) if I don't go crazy and stay in lane and use autopilot. I've gone as low as 215whm crawling in some traffic but hey still low.

 

[095179005]

Tried calculating my NFP value based on previous posts - 336km range at 120Wh/km at 51% SOC - comes to ~79kWh.
Does that sound right?

I'm guessing the calculation is off by 200Wh just cause it's one measurement?

 

[Wolfshy81]

Hi everyone,

I tried reading this thread but it’s big and long!

I’ve recently had my battery for my APRIL 2021 MIC M3P replaced. I’m curious did I get the same, better, or worse to what was in it originally (which I don’t have any info on, hopefully the above is enough to go by though seems I had ~ 79 kWH according to a third party app when I started using it).

Battery replacement details as below
HVBAT,REMAN,LONG 1.00 RANGE,AWD-
RWD,3PH,M3(1137377-
01-P)

Thanks!

Hi guys, just wanted to re-up my post. It’s been a while, and I was just wondering with the benefit of time, if the above battery pack might better be identifiable now. Tesla won’t tell me anything in particular about it.

Thanks!

 

[KenC]

Tried calculating my NFP value based on previous posts - 336km range at 120Wh/km at 51% SOC - comes to ~79kWh.
Does that sound right?

View attachment 821271
I'm guessing the calculation is off by 200Wh just cause it's one measurement?

Rounding. Most of the rounding error will come from the SOC. Whatever it says is +/- 0.49%, right?

 

[AAKEE]

Rounding. Most of the rounding error will come from the SOC. Whatever it says is +/- 0.49%, right?

Not to be picky, but to strengthen the understanding for the need to do the calculation with relatively high SOC, I’d say the rounding error is actually [ +/- 0.49 divided by the actuall SOC in absolute number].

This means, that at 100% SOC the rounding error is 0.49%. (In “theory”, as if charged until it stops, the true SOC is 100% without rounding.)
But at 4.9% SOC the error is 10%, as the true SOC can be 10% off. (+/- 0.49 / 0.1 = 4.9%)
At 1% SOC, 0.49 / 0.01 = 49%.

At 51% SOC we get about 1% error.