Battery estimated capacity question

[Sti-g]

Hi everyone. New owner of a 2021 Model 3 SR+.

I’m using the method described in the sticky post:

Battery Capacity @ 100% (in Wh) = Avg Wh/mi * Projected Range / SoC%

My question is simple: Does this formula yield the full pack size or useable size? AFAIK, full pack size for my car is 55kWh and usable is 52.4kWh.

Actually I have another question. I’m guessing this method will result in pessimistic value due to cold temps, right? So perhaps best to wait until spring for more accurate calculation.

Thank you!

 

[Sti-g]

Should clarify more. My end goal is to calculate degradation.

What I calculated is: 0.244kWh/miles * 168miles / 84% = 48.8kWh

Based on whether this is full size of the pack or usable size, I get 11.3% or 6.9% degradation.

 

[Sti-g]

Hi again. Anyone with an answer? Does the formula calculate the full pack size, or useable size?
Thank you!

 

[ATPMSD]

Hi again. Anyone with an answer? Does the formula calculate the full pack size, or useable size?
Thank you!

Pretty sure it is usable, not full. The BMS displays battery percentage / distance based on usable. This is to ensure there is a margin remaining if the driver takes to SOC 0%

 

[AlanSubie4Life]

Actually I have another question. I’m guessing this method will result in pessimistic value due to cold temps, right?

Not much change should be expected. I would recommend doing it when your pack is warm (no blue showing) to get reasonable results.

My question is simple: Does this formula yield the full pack size or useable size?

Full pack. Usable is 95.5% of the value.

What I calculated is: 0.244kWh/miles * 168miles / 84% = 48.8kWh

Based on whether this is full size of the pack or usable size, I get 11.3% or 6.9% degradation

I am not sure if you have the LFP 2021 SR+ (55.1kWh FPWN) or the NCA (53.5kWh 2170L FPWN though starting values were often higher I think). That would affect % loss.

You can get info on the starting range of the vehicle to determine that.
So assuming you started with 263 rated miles, you have the NCA.

Unless noted, I think these are “Full Pack When New” values, remember degradation threshold can be lower. (54.7kWh or so for the SR+ LFP for example). Really should keep a spreadsheet and dig up all the exact numbers.

SR+ - 52.5kWh
SR - 52.5kWh (but locked out 220/240 of the capacity)
SR+ 2170L - ~53.5kWh (but 55.4kWh FPWN)
SR+ LFP - 55.1kWh
RWD - 60.5kWh

Why is home charging only 32 amps on SR when supercharging is so much faster?

I would expect 8-10% as a default estimate.

Pretty sure it is usable, not full.

It is the entire capacity including buffer for this method.

It should really be clarified in that sticky. Though the example I give for my car shows it is the full capacity not including buffer.

In addition it should be clarified that for new cars, the method has a ceiling limit at the degradation threshold - it will not provide values that exceed the threshold, even though for new cars that will frequently happen.

 

[jjrandorin]

It should really be clarified in that sticky. Though the example I give for my car shows it is the full capacity not including buffer.

In addition it should be clarified that for new cars, the method has a ceiling limit at the degradation threshold - it will not provide values that exceed the threshold, even though for new cars that will frequently happen.

If you want me to add Text to the sticky that you feel is relevant, you can send it to me in a PM (what, where in the body text it should go if not at the end of the post, etc).

 

[ATPMSD]

It is the entire capacity including buffer for this method.

Not saying your wrong, but…

“Logically” if you drive the car down to 0% SOC then the range will also show zero miles, but of course the car is still drivable owning to the built in buffer. The question is, what will the projected range show on the graph at 0% SOC? If it also shows zero then the result of the calculation in question should based on usable capacity.

Anybody tried this?

 

[Sti-g]

Not much change should be expected. I would recommend doing it when your pack is warm (no blue showing) to get reasonable results.

Full pack. Usable is 95.5% of the value.


I am not sure if you have the LFP 2021 SR+ (55.1kWh FPWN) or the NCA (53.5kWh 2170L FPWN though starting values were often higher I think). That would affect % loss.

You can get info on the starting range of the vehicle to determine that.
So assuming you started with 263 rated miles, you have the NCA.

Unless noted, I think these are “Full Pack When New” values, remember degradation threshold can be lower. (54.7kWh or so for the SR+ LFP for example). Really should keep a spreadsheet and dig up all the exact numbers.

SR+ - 52.5kWh
SR - 52.5kWh (but locked out 220/240 of the capacity)
SR+ 2170L - ~53.5kWh (but 55.4kWh FPWN)
SR+ LFP - 55.1kWh
RWD - 60.5kWh

Why is home charging only 32 amps on SR when supercharging is so much faster?

I would expect 8-10% as a default estimate.


It is the entire capacity including buffer for this method.

It should really be clarified in that sticky. Though the example I give for my car shows it is the full capacity not including buffer.

In addition it should be clarified that for new cars, the method has a ceiling limit at the degradation threshold - it will not provide values that exceed the threshold, even though for new cars that will frequently happen.

I have the NCA chemistry (262mile EPA range). I was careful to get NCA battery since it is lighter yet I wanted 2021 model for the heatpump upgrade. I live in Seattle, so heatpump is perfect for mild winters. My goal was to get most efficient model (though I couldn't help myself and got 19" wheels..).

Based on lots of googling, the information I found stated that full pack size is 55kWh (code name: BT41) and usable size is 52.4kWh (very close to 95.5% of 55kWh). But honestly, I am not sure if that information is correct. It seems like what I found (55kWh) might be rounded-down version of FPWN, which you say is 55.4kWh. As I understand, FPWN is just a "label", not actual pack size.

48.8kWh / 52.4kWh --> 6.9% degradation [perhaps a bit optimistic, given the built date of the car is 9/27/2021)
48.8kWh / 53.5kWh --> 8.8% degradation [is this within statistical variation for ~28 month old car?]
48.8kWh / 55.0kWh --> 11.3% degradation
48.8kWh / 55.4kWh --> 11.9% degradation

Good point on doing this when the pack is warm. I don't recall seeing the snowflake icon, but outdoor temperature was <50F and I don't have a garage.

By the way, I think it would be great to update the sticky with this information (ie, formula gives full capacity including buffer). Because I have seen contradicting information in various posts/threads using this formula. And it would be even better if there is a list of battery pack sizes per year... though Tesla definitely does not make that easy!

 

[AlanSubie4Life]

The question is, what will the projected range show on the graph at 0% SOC? If it also shows zero then the result of the calculation in question should based on usable capacity

Yeah, logically it’s a bit confusing, I agree. I would have had it work differently, but here we are.

Never tried, but should show 0 at 0%. You are assuming the projection is correct though - that is where your logic (which is correct) goes wrong.

However, the key: The projected range overestimates the range based on your current consumption, since it is using the charging constant to do the calculations, not the actual energy per displayed rated mile (which is lower). The error in miles fortunately tends to 0 as you approach 0%. (At 100% the projected range will be off (high) by 4.7% (inverse of 4.5% low) if your end target is 0% and you don’t plan to use the buffer. ) For example if your capacity is 70kWh and your recent efficiency is 300Wh/mi at 100%, it will project 233 miles as your range. But your actual range will be 223 miles at 300Wh/mi to 0%. At 50% it only includes half the buffer in the energy value, so it would project 116 mi, with actual range of 111mi (33.4kWh, 300Wh/mi). Error is smaller in absolute terms, same in percentage terms.

(And notably the buffer if you were extremely lucky might give you 11 more miles. Though not recommended and would likely never happen.)

 

[Sti-g]

Not saying your wrong, but…

“Logically” if you drive the car down to 0% SOC then the range will also show zero miles, but of course the car is still drivable owning to the built in buffer. The question is, what will the projected range show on the graph at 0% SOC? If it also shows zero then the result of the calculation in question should based on usable capacity.

Anybody tried this?

I was thinking the same. This would be a good test.
If the energy app estimate includes buffer, then when SOC is at 0%, it will still show some few miles (as much as 7-8miles in SR+) of estimated range. I don't see that happening actually. Estimated range in energy app is always seem to be less than range shown in SOC. Though that could be due to actual consumption vs EPA consumption rate.

 

[AlanSubie4Life]

If the energy app estimate includes buffer, then when SOC is at 0%, it will still show some few miles (as much as 7-8miles in SR+) of estimated range. I don't see that happening actually.

See above.

Note the formula has been checked many times vs. the Nominal Full Pack value from SMT. There is no doubt here that the calculation includes the buffer.

You should find at 0% that it projects 0 miles. But that does not mean the formula does not include the buffer in the result (see above).

 

[Sti-g]

Yeah, logically it’s a bit confusing, I agree. I would have had it work differently, but here we are.

Never tried, but should show 0 at 0%. You are assuming the projection is correct though - that is where your logic (which is correct) goes wrong.

However, the key: The projected range overestimates the range based on your current consumption, since it is using the charging constant to do the calculations, not the actual energy per displayed rated mile (which is lower). The error in miles fortunately tends to 0 as you approach 0%. (At 100% the projected range will be off (high) by 4.7% (inverse of 4.5% low) if your end target is 0% and you don’t plan to use the buffer. ) For example if your capacity is 70kWh and your recent efficiency is 300Wh/mi at 100%, it will project 233 miles as your range. But your actual range will be 223 miles at 300Wh/mi to 0%. At 50% it only includes half the buffer in the energy value, so it would project 116 mi, with actual range of 111mi (33.4kWh, 300Wh/mi). Error is smaller in absolute terms, same in percentage terms.

(And notably the buffer if you were extremely lucky might give you 11 more miles. Though not recommended and would likely never happen.)

This confused me. Is the projected range what's shown in the energy app? If so, you just said projection is not correct... If it is not correct, how are we calculating the battery capacity? Or what am I missing?

 

[AlanSubie4Life]

Estimated range in energy app is always seem to be less than range shown in SOC.

It’ll be higher if your recent efficiency is lower than the vehicle charging constant (which is the position of your rated energy line minus 5Wh/mi). And lower if your recent efficiency is higher.

I think your constant should be 203Wh/mi (53.5kWh/263mi) but not 100% sure about that for your car since it is less common so I may be misremembering. So your rated line would be 208Wh/mi.

 

[AlanSubie4Life]

Is the projected range what's shown in the energy app? If so, you just said projection is not correct... If it is not correct, how are we calculating the battery capacity? Or what am I missing?

It’s what is in the energy app.

It’s “correct,” if you scale it by 95.5%. So all works out.

So if you want to, just take the projected range and multiply it by 95.5%. Then do the calculation. That will give you your usable capacity. But anyway, see above.

It is confusing. It confused me for quite some time. But once you accept the projections are incorrect (by a %), it is easier.

 

[AlanSubie4Life]

Based on lots of googling, the information I found stated that full pack size is 55kWh (code name: BT41) and usable size is 52.4kWh (very close to 95.5% of 55kWh). But honestly, I am not sure if that information is correct. It seems like what I found (55kWh) might be rounded-down version of FPWN, which you say is 55.4kWh. As I understand, FPWN is just a "label", not actual pack size.

EPA test got 54.7kW out of it (includes the buffer!; they drive it on the dyno until it stops moving), so that is a good value to use as a starting number. Page 21

https://dis.epa.gov/otaqpub/display_file.jsp?docid=51461&flag=1

I think degradation threshold is 53.5kWh for your vehicle, but you could help with that by taking the energy screen picture but with the battery icon set to mi/km NOT %, and capture all three “key” numbers. That allows calculation of the constant, which in turn allows calculation of the threshold, since we know the max rated range was 263mi.


You probably show 240mi at 100% but the actual correctness of my guess for that depends on that degradation threshold.

Seems like you are down at least 9%, maybe as much as 11%. Very normal.

 

[Sti-g]

I think degradation threshold is 53.5kWh for your vehicle, but you could help with that by taking the energy screen picture but with the battery icon set to mi/km NOT %, and capture all three numbers. That allows calculation of the constant, which in turn allows calculation of the threshold, since we know the max rated range was 263mi.

I will do this and report back. Should I charge to >80% for more accurate results?
By the way, rated range was 262miles (on Tesla website at least).

 

[Sti-g]

It’s what is in the energy app.

It’s “correct,” if you scale it by 95.5%. So all works out.

So if you want to, just take the projected range and multiply it by 95.5%. Then do the calculation. That will give you your usable capacity. But anyway, see above.

It is confusing. It confused me for quite some time. But once you accept the projections are incorrect (by a %), it is easier.

I think I got it. Since the energy app already scales it by 95.5%, the formula gives the full pack size including the buffer.

 

[AlanSubie4Life]

By the way, rated range was 262miles (on Tesla website at least).

No, it is 263mi. 262 is for LFP unicorn.

Should I charge to >80% for more accurate results?

Doesn’t really matter that much as long as all three numbers have three digits (switch to km if that helps).

 

[Sti-g]

No, it is 263mi. 262 is for LFP unicorn.

View attachment 1007557

The plot thickens. I double checked, the Tesla used inventory page shows 262miles. I’m pretty sure I have NCA pack because GVWR=4446lbs. Therefore car weight is 4446-826=3620lbs vs ~3860lb for LFP models. Also, 0-60 time is 5.3s vs 5.8 for LFP models.
Also, I have 2 tier charging in the app. It recommends 80% for daily trips.
Also, I believe under software—>more info it would say LFP and mine doesn’t (don’t recall what it says but it doesn’t say LFP).

 

[AlanSubie4Life]

Tesla used inventory page shows 262miles.

Yeah it is just wrong. So many vehicle types that year they cannot keep it straight I guess.

You have an NCA and the EPA knows what is up.