2020, 2019, 2018 Model 3 Battery Capacities & Charging Constants

[WillyThePooh]

Great! Your calculations are correct. Thanks for posting here (eventually, after these numbers settle out, we will change the thread title to include 2021).

Degradation Threshold/Max Capacity: 53.5kWh. (I believe this is higher than 2020 by about 1kWh. There are pictures around here of the 2020 showing it was about 52.5kWh)

Other insights:

Charging constant is currently about the same as the 2020 SR+ (210-211Wh/rmi).

Remember that currently your max rated miles right now is only ~254 rated miles (2020 maxed out at 250 rated miles of the same energy). So I generally expect a constant change in the future to about 203-204Wh/rmi.

That would give: 53.5kWh/203Wh/rmi = 263 rated miles. (Current displayed mileage on Tesla website.)

Note it isn't actually going to give any more range. It's just a constant change. Anyway, if you do ever see a boost to about 381km at 90% charge, take another look at this Energy screen (and post here if you are so inclined!). I suspect you'll see the line position and the constant used for the calculations change, if that happens.

Thanks for the response! That was fast

That sounds about right in terms of the constant. I usually charge up to 90-95% SOC, so I'll keep an eye on how much this varies and I'll post the updated pics if I see a significant change.

One other thing I was wondering is, how does ambient temperature affect these data and calculations? From what I could find, it looks like most of the EPA's testing is done at 75 F (~23.9 C). Since my data were taken in 3 to 4 C (37 to 39 F) ambient temperature, would this affect any of our derived values?

I'm assuming Tesla's algorithms take temperature into account when calculating projected range (at least indirectly through the higher average energy consumption).

If the ~20 km rated range increase is mainly due to the addition of the heat pump system, I really want to grasp how temperature is accounted for.

 

[AlanSubie4Life]

Tesla's algorithms take temperature into account when calculating projected range (at least indirectly through the higher average energy consumption).

Indirectly through higher consumption. The constant does not change (except possibly one single time in an upcoming software update to get you to your apparently promised 263 rated miles).


Note that for trips you should just use the trip planner on the other tab of the Energy screen (Trip). It is much better as it takes into account elevation, recent consumption, etc. The Energy Consumption page is useless, except for figuring out your battery capacity, TBH. I never look at it as it is basically irrelevant.

If the ~20 km rated range increase is mainly due to the addition of the heat pump system, I really want to grasp how temperature is accounted for.

I’ve explained elsewhere, you can look. There are five test cycles. Two of these use climate control - one hot, one cold. The heat pump does substantially better in these. This allows Tesla to claim a higher multiplier than 0.7 on their weighted two-cycle test results (which are done at room temperature). They claim (it is a defined formula) about 0.746 or so. This therefore results in higher EPA range. That is how the EPA range becomes a function of the heat pump; it’s folded into the scalar.

EPA range = (0.55*UDDS+0.45*HWY)*scalar

That results in more rated range than prior years. Basically they can specify a lower constant (even if you don’t realize it for a particular drive), so for a given energy, you can display more rated miles.

 

[AlanSubie4Life]

Even the heat pump in ideal conditions and poorly setup 2019 car wouldn't bring that much difference. Interesting to see what the US cars will show @100% and wether it will be above 332 miles

They'll show about 353 rated miles. No question about that.

Oh yeah? I highly doubt it. Don't we already have delivered LRs?

before crowing about it, if the initial value does happen to be 330+ miles, please wait for a software update in the next month or so. Eventually new cars will start being delivered with 353 miles. That is the EPA rated range.

Ahhhh, so I have to wait now, oo oo kkkk....

Why don’t we wait until some vehicles are delivered in the United States? I would bet you eternal glory, or $1k, that they will display 353 rated miles for a battery with capacity that exceeds the degradation threshold.

This is a US car at below 353...In Portland. You sure you have 1$k?

But I’ve already mentioned the software update - remember this is what happened in 2020 (in 2019 for 2020 vehicles) - the vehicles did not show the EPA range at first and then were updated a couple weeks after delivery (the constant was changed). If the software change goes to 353 rated miles (within 2 rated miles of that value, say) I would win the bet.

I’ll be right and I don’t want to take your money, which is why I am not really serious about it. I would feel bad!

The new AWD 18” (not Performance) vehicles will eventually be assigned a constant that results in 353 rated miles, in the US

In that thread, if we get the pictures we want, we'll soon know exactly (within 1Wh/rmi) what the constant is for the AWD 2021. And that should settle it. I expect it will be about 220Wh/rmi, or 137Wh/rkm

Alan wins the bet (of course)! (We did not actually finalize the bet though.)

US 2021 Model 3 that will show ~353 rated miles at 100%

The constant is now 137Wh/rkm, as I predicted. With the software update, within a month, as predicted.

The unlocked battery capacity (on the AWD non-P) is unchanged from 2020, in SMT, as I predicted. (Covered elsewhere with SMT captures in that thread.)

But, to be clear, I didn't "predict" anything. I just read the EPA documents, and took Tesla's number from their website, which says the EPA estimated range of 2021 Model 3 AWD is... 353 rated miles.

For some reason @TimothyHW3 doesn't want to believe the EPA documents, which have always been quite reliable.

Now...there is still the unresolved question of whether it's possible there are some AWD non-P vehicles out there with the larger (but locked) battery capacities, which the Performance all have (the Performance may have the same battery as AWD, unlocked, or a denser battery mostly unlocked (I suspect the latter) ). But that's a separate question.

 

[Phlier]

Note that for trips you should just use the trip planner on the other tab of the Energy screen (Trip). It is much better as it takes into account elevation, recent consumption, etc. The Energy Consumption page is useless, except for figuring out your battery capacity, TBH. I never look at it as it is basically irrelevant.

The important parts are in bold.

Do yourself a favor on your next road trip, and take Mr. Subie's advice.

 

[Masa]

2020 M3 LR AWD. Took Delivery in September 2020. I am getting far less range than I expected. I charged to 95% (300 miles) and drove around the city and freeway. Drove it down to under 18% in 2 days. Only drove about 120 miles. Do I have a Bad battery. Can anyone tell me if this is normal?

 

[LargeHamCollider]

View attachment 624018 2020 M3 LR AWD. Took Delivery in September 2020. I am getting far less range than I expected. I charged to 95% (300 miles) and drove around the city and freeway. Drove it down to under 18% in 2 days. Only drove about 120 miles. Do I have
a Bad battery. Can anyone tell me if this is normal?

Did you leave it in the cold overnight not plugged in? Cause that will drain it quick even though it's not being driven.

 

[AlanSubie4Life]

View attachment 624018 2020 M3 LR AWD. Took Delivery in September 2020. I am getting far less range than I expected. I charged to 95% (300 miles) and drove around the city and freeway. Drove it down to under 18% in 2 days. Only drove about 120 miles. Do I have a Bad battery. Can anyone tell me if this is normal?

Seems fine. With your vehicle constant of around 242Wh/rmi (discharge/displayed about 231Wh/rmi) that driving would have used ~187-189 rated miles. (120.1mi*358Wh/mi/(0.955*0.99*242Wh/rmi) = 188 rmi.) You have 61rmi left and that leaves only 300rmi -61rmi - 187rmi = 52 rated miles unaccounted for, max. (52 rated miles is 242Wh/rmi*(0.955rmi/1rmi(displayed)) * 52rmi(displayed) ~ 12kWh)

If you use Sentry Mode or whatever other features (your screen suggests you are trying!), that could easily account for a lot of the remaining energy (Sentry uses about 1 rated mile per hour, very roughly). Some of course can also be used to heat the car remotely or when sitting in the car in Park (all such use when in Park is not counted on the meter).

Cold isn’t an issue for your car since you are in LA.

Anyway, seems fine. If you want to know your battery capacity just take this picture and do the calculation.

I expect you’ll have 76kWh as a result from that calculation, which is fine (you started around 77.8kWh).

No problems! Perfectly normal. If you don’t want it to use any significant energy on this timescale, make sure it is sleeping and you are not using any third-party apps or features like Sentry Mode.

 

[Masa]

Seems fine. With your vehicle constant of around 242Wh/rmi (discharge/displayed about 231Wh/rmi) that driving would have used ~187-189 rated miles. (120.1mi*358Wh/mi/(0.955*0.99*242Wh/rmi) = 188 rmi.) You have 61rmi left and that leaves only 300rmi -61rmi - 187rmi = 52 rated miles unaccounted for, max. (52 rated miles is 242Wh/rmi*(0.955rmi/1rmi(displayed)) * 52rmi(displayed) ~ 12kWh)

If you use Sentry Mode or whatever other features (your screen suggests you are trying!), that could easily account for a lot of the remaining energy (Sentry uses about 1 rated mile per hour, very roughly). Some of course can also be used to heat the car remotely or when sitting in the car in Park (all such use when in Park is not counted on the meter).

Cold isn’t an issue for your car since you are in LA.

Anyway, seems fine. If you want to know your battery capacity just take this picture and do the calculation.

I expect you’ll have 76kWh as a result from that calculation, which is fine (you started around 77.8kWh).

No problems! Perfectly normal. If you don’t want it to use any significant energy on this timescale, make sure it is sleeping and you are not using any third-party apps or features like Sentry Mode.

Thank you for the info! I will try disabling sentry mode if not necessary.

 

[Harvey Danger]

0.955 accounts for the buffer not being included in the 51.5kWh. (Converts from 210Wh/rmi charge/EPA, to 200.5Wh/rmi, displayed.)

0.98 is approximate (might be a bit closer to 1 for you) because the trip meter seems to miss some energy. At these consumption levels it might miss less. Or more. I don’t know. Empirically determined. (Converts from 200.5Wh/rmi displayed to ~197Wh/rmi (trip))

Thanks for putting all this together.

I was wrestling with all these concepts on my own and stumbled across this thread and it really helped.

Is there a place where you go into more detail about each of the column headings in your spreadsheet?

It looks like you've invented your own terminology for some of this, and if I have this correctly you are tracking three "constants":

- by "charge constant" you mean the conversion rate the firmware uses to convert from kWh remaining (which is carefully tracked but confusing to report since people don't intuitively understand it) to rated miles (which gives people a friendly number to look at).

[In the old days you could just grab this number off the CAN bus by looking for the aptly named VAPI_ratedWattHourPerMile . Nowadays you can make a rough determination by taking a picture of your display during supercharging. Divide the number shown next to "kW" by the number shown next "mi/hr". Multiply by 1000 if the decimal place is bothering you, but the first three significant figures are what you are looking for. Or you can bring up the energy graph and manipulate your driving style until your recent average overlaps the line that says Rated. Then subtract two percent of the value you come up with because some reason the "rated" line is always drawn 2% too high. ]

- by "BMS constant" you mean the rate at which the BMS reduces the displayed rated miles remaining as you go from 100% to 0% SoC. This is is basically the charge constant multiplied by the ratio of ((usable pack energy) divided by (full pack energy which includes unusable buffer energy). If the buffer is a fixed percentage of the full pack, then the BMS constant is the charge constant reduced by this buffer percentage.

[I'm not sure how the average person can determine the buffer size for their model and firmware, but I do believe it is reported on the CAN bus and can be exposed easily enough with ScanMyTesla or some such. I understand this number is purely notional, it does not represent any literal allocation of battery capacity, it is just a mathematical constant included in range estimation algorithms to avoid alarming drivers with sudden changes in remaining-range near the bottom of the pack. ]

-by "Trip Meter constant" you just mean the BMS constant reduced by an arbitrary amount empirically observed to match consumption rates displayed in the trip meter during real world drives of any length. You have chosen a nice round 2 % to represent this. I have no idea if it is possible to gain any more precision than that- trip meter behavior seems to depend on a whole lot of unpredictable real world conditions.

Do I have this right?

Do you know anywhere I can go to learn more?

 

[AlanSubie4Life]

by "charge constant" you mean the conversion rate the firmware uses to convert from kWh remaining (which is carefully tracked but confusing to report since people don't intuitively understand it) to rated miles (which gives people a friendly number to look at).

Do I have this right?

Do you know anywhere I can go to learn more?

Pretty much have it right. For the charge constant, worth noting that kWh remaining divided by the charge constant equals miles remaining ONLY at 100%. That’s because as we know discharge doesn’t click off at the same charge constant rate. So once you start clicking off miles you can’t use the same formula anymore. (However rated miles remaining * charge constant / SoC will always roughly equal your max pack capacity.)

The charge constant shows up a lot of places - it’s used for conversion between kWh added at a charging session and miles added, for example (note this means the kWh added to the battery is not correct - actual kWh added is about 4.5% less for Model 3, again due to the buffer 4.5% factor). And of course on the energy screen it is used.

Not sure anywhere else you can go. I learned this empirically from info in my car, and supplemented it with data from SMT that people published here (I don’t have SMT). I didn’t read it anywhere.

 

[Phlier]

However rated miles remaining * charge constant / SoC will always roughly equal your max pack capacity.

That right there should be the title of a locked, stickied thread at the top of this forum, as it answers 99% of all charge related and "is my battery dying?" questions posted here. Fortunately, there is a stickied thread that does have that little nugget in it, so at least it's easy to refer people to it.

Great way to immediately check the health of your pack without having to jump through a bunch of hoops.

 

[Harvey Danger]

Pretty much have it right.

I learned this empirically from info in my car, and supplemented it with data from SMT that people published here (I don’t have SMT). I didn’t read it anywhere.

Excellent, thank you, sounds like I'm on the right track then.

I think it's very helpful to share ideas that allow everyone (not just people who have SMT or other specialized tools) to understand where the numbers in the displays come from, what they mean etc.

I'd love to see someone expand your spreadsheet for all Tesla models not just all the Model 3 variants. I might do it myself if I could locate an authoritative source for the charge constants.
This thread was a good start but there are so many more models now, and of course it's sometimes a moving target if it gets changed by subsequent firmware revisions..

 

[AlanSubie4Life]

I might do it myself if I could locate an authoritative source for the charge constants.

You just need the energy screen snapshot at a high SoC for each type of vehicle (showing the three numbers Remaining Rated Range, Projected Range, and Recent Avg Efficiency). That will give you the charge constants assuming these screens behave exactly the same way on Model S & X. Beyond that, it's trickier, because not all vehicles work the same way. I know with reasonable certainty that Model S and Model X do NOT have the same 4.5% buffer - it's a different % (so that will affect the discharge constant). I know that from another user's data from a scanning tool on a Model S. I also heard recently from Green on Twitter that the Model S behavior has changed over time (might not have been an explicit buffer originally?).

Once you have the charge constant, you can just multiply it by the original rated range of the vehicle according to Tesla's EPA numbers (fuel economy.gov), assuming there haven't been any adjustments (which occasionally but rarely happen), and that will give you the nominal starting battery capacity (which will match the readings from SMT, etc.).

 

[nj1266]

Hi @AlanSubie4Life Can I get an estimate on the current battery capacity in kWh of my Model 3 Performance on 18 inch wheels? I have these screenshots: one from the car display and one from Teslafi the last time I charged to 90%. At 100%, Teslafi tells me that my rated range should be 294 miles. TIA

 

[AlanSubie4Life]

Hi @AlanSubie4Life Can I get an estimate on the current battery capacity in kWh of my Model 3 Performance on 18 inch wheels? I have these screenshots: one from the car display and one from Teslafi the last time I charged to 90%. At 100%, Teslafi tells me that my rated range should be 294 miles. TIA

View attachment 696632

View attachment 696631

294mi*245Wh/mi = 72kWh

Or (perhaps less accurate due to low SOC): 212Wh/mi*174mi/0.52 = 71kWh

So about 9% capacity loss which is very typical.

About 69kWh usable (above 0%).

 

[ElectricIAC]

294mi*245Wh/mi = 72kWh

Or (perhaps less accurate due to low SOC): 212Wh/mi*174mi/0.52 = 71kWh

So about 9% capacity loss which is very typical.

About 69kWh usable (above 0%).

Ours is down to ~270 at 100%.

 

[nj1266]

294mi*245Wh/mi = 72kWh

Or (perhaps less accurate due to low SOC): 212Wh/mi*174mi/0.52 = 71kWh

So about 9% capacity loss which is very typical.

About 69kWh usable (above 0%).

Can you please explain the math to me? Specifically where you got the 245Wh/mi from? Is that a constant number for all 2020 model 3 performance on 18 inch wheels? The second math line I am familiar with as I have done it before at slightly lower miles on the car. TIA

 

[AlanSubie4Life]

Can you please explain the math to me? Specifically where you got the 245Wh/mi from? Is that a constant number for all 2020 model 3 performance on 18 inch wheels?

Oops! I assumed you had a 2018/2019 (or missed info you provided/missed the subtlety of calling it a Performance 18”).

For 2020, 245Wh/mi is the wrong constant number.

The number is arrived at by taking the degradation threshold of 77.8kWh for the 2020 Model 3 18” (kind of derived empirically, but also the explanation is above about how that value is arrived at), and dividing by the 322 rated mile EPA rating (18”).
So for the 2020 it is 241.6Wh/rmi. (Compare to 76kWh/310rmi = 245Wh/rmi for 2018, 2019. 76kWh is the degradation threshold for those (Stealth) vehicles even though they started around 78kWh just like the 2020 vehicles.)

So for you the capacity is:

241.6Wh/rmi*294rmi = 71kWh.

Which matches the energy screen method as expected.

When I started this thread, I did not completely understand how the degradation thresholds work. This thread is still correct (and captures some of how it behaves, correctly) but when the vehicle is new, the term “charging constant” ends up being a bit oxymoronic. It is indeed a constant that is used by the car to display certain information in the car, but as the tables above allude to, the energy in each rated mile is a little higher than expected (and variable but generally reducing towards the constant value) as long as the vehicle battery energy exceeds the degradation threshold.

 

[nj1266]

Oops! I assumed you had a 2018/2019 (or missed info you provided/missed the subtlety of calling it a Performance 18”).

For 2020, 245Wh/mi is the wrong constant number.

The number is arrived at by taking the degradation threshold of 77.8kWh for the 2020 Model 3 18” (kind of derived empirically, but also the explanation is above about how that value is arrived at), and dividing by the 322 rated mile EPA rating (18”).
So for the 2020 it is 241.6Wh/rmi. (Compare to 76kWh/310rmi = 245Wh/rmi for 2018, 2019. 76kWh is the degradation threshold for those (Stealth) vehicles even though they started around 78kWh just like the 2020 vehicles.)

So for you the capacity is:

241.6Wh/rmi*294rmi = 71kWh.

Which matches the energy screen method as expected.

When I started this thread, I did not completely understand how the degradation thresholds work. This thread is still correct (and captures some of how it behaves, correctly) but when the vehicle is new, the term “charging constant” ends up being a bit oxymoronic. It is indeed a constant that is used by the car to display certain information in the car, but as the tables above allude to, the energy in each rated mile is a little higher than expected (and variable but generally reducing towards the constant value) as long as the vehicle battery energy exceeds the degradation threshold.

Shouldn't the number for the 2020 on 18s be 241 wh/mi? Because you typed 241.6 but the tables you posted say 241 in that column

 

[AlanSubie4Life]

Shouldn't the number for the 2020 on 18s be 241 wh/mi? Because you typed 241.6 but the tables you posted say 241 in that column

Yeah it’s approximate but quite close. I think I used to think the degradation threshold was slightly lower (77.6kWh?) It could be 241Wh/mi, but I think it is basically right between those two numbers. You can actually tell yourself what the rounded value is from the energy screen in miles mode with three sig figs for each of the three numbers: Product from the energy screen (Proj range * efficiency), divided by rated miles remaining. While trying your best to eliminate rounding errors, etc.

In the end a 0.4% difference doesn’t matter much.