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

[AlanSubie4Life]

I charge with a 14-50 at 240V, 32A and 30 mph. The voltage at charge is about 236V.

I think you should measure the time it takes to charge next time, and hand calculate and compare to your meter. These numbers are exactly what I would expect:

30rmi/hr*223Wh/rmi / (236V*32A) = 0.886

The efficiency is ~88.5%.

So if you add 225rmi:

hours = 225mi/30mi/hr = 7.5 hours

7.5hr*236V*32A = 56.6kWh. (Note I am just using your data here!)

So something is not right with the AC meter data. Not all meters are created equal. But anyway it is very very easy to check since the car tells you the voltage and current. You just need the time!

I can take the range miles to be added and divide by 4.2 mi/kWH (~238wh/mi) and come very close to the AC kWH to be added every time.

I am not sure where that 238Wh/mi number comes from. But you have to account for charging losses.

 

[jdcollins5]

I

I think you should measure the time it takes to charge next time, and hand calculate and compare to your meter. These numbers are exactly what I would expect:

30rmi/hr*223Wh/rmi / (236V*32A) = 0.886

The efficiency is ~88.5%.

So if you add 225rmi:

hours = 225mi/30mi/hr = 7.5 hours

7.5hr*236V*32A = 56.6kWh. (Note I am just using your data here!)

So something is not right with the AC meter data. Not all meters are created equal. But anyway it is very very easy to check since the car tells you the voltage and current. You just need the time!
.

I can tell you that I can take the range miles to be added and divide by 30mph to get charge time, and it is always right based on notifications. This is what I use to schedule charge times during TOU. I will verify next time I charge.

I am not sure where that 238Wh/mi number comes from. But you have to account for charging losses.

Remember this is on the AC side so includes losses. The 4.2 is just a trial and error number I used until I could closely estimate kWH needed for range miles needed. Once again on the AC side.

 

[AlanSubie4Life]

I


I can tell you that I can take the range miles to be added and divide by 30mph to get charge time, and it is always right based on notifications. This is what I use to schedule charge times during TOU. I will verify next time I charge.




Remember this is on the AC side so includes losses. The 4.2 is just a trial and error number I used until I could closely estimate kWH needed for range miles needed. Once again on the AC side.

Fair enough. You can see that really I am only taking issue with what your AC meter reads - everything else makes sense. If the time is right, the voltage and current are correct, only the meter can be reading low (by a lot - 8%). There’s not a lot of wiggle room here.

I expect you should see about 252Wh per rated mile you add (223Wh/rmi / 0.885). Your number seems about 6% low...once again I suspect the meter. I suppose the load could be highly reactive (meaning voltage and current are not in phase) and your meter is right. But does not align with the data I have seen in my case (ChargePoint, Tesla’s data...EPA rating)- looks pretty purely resistive.

This number will not align with the EPA number because those are different rated miles due to the 4.5% buffer (in that framework it would be 234Wh/rmi / 0.885 = 264Wh/rmi which is fairly close to the EPA rating).

 

[jdcollins5]

Fair enough. You can see that really I am only taking issue with what your AC meter reads - everything else makes sense. If the time is right, the voltage and current are correct, only the meter can be reading low (by a lot - 8%). There’s not a lot of wiggle room here.

I expect you should see about 252Wh per rated mile you add (223Wh/rmi / 0.885).

This number will not align with the EPA number because those are different rated miles due to the 4.5% buffer (in that framework it would be 234Wh/rmi / 0.885 = 264Wh/rmi which is fairly close to the EPA rating).

Thanks for your time and explanations. I will test the meter next time I charge.

 

[Wreckless316]

Should not use this here because you know how many rated miles you use. As I said, unnecessarily complex.

Also made a mistake - you only used 93%.

51.5*0.955*0.98*0.93 = 44.8kWh.

Easier to just do (232rmi-4rmi)*197Wh/rmi = 44.9kWh

How sure are we that the "since last charge" or any other meter displays energy used as it pertains to AC/Heater?

I have been sitting in my car on a call and my AC is on, for about 45 minutes now, parked, but intentionally leaving it in drive/hold. Percentage has gone down about a percent but wh/mi on the "since last charge" has remained the same at 157 wh/mi.

 

[AlanSubie4Life]

How sure are we that the "since last charge" or any other meter displays energy used as it pertains to AC/Heater?

I have been sitting in my car on a call and my AC is on, for about 45 minutes now, parked, but intentionally leaving it in drive/hold. Percentage has gone down about a percent but wh/mi on the "since last charge" has remained the same at 157 wh/mi.

As soon as you start rolling you'll see the energy get counted. That display also only seems to get updated when the odometer is going up.

Let us know. I and others have experimented with this before and that energy does seem to be counted as long as you are not in park. But you do have to get rolling.

It's also possible that putting the car in Park will force the counter to update. (You can try this - as long as you don't get out of the seat the counter won't reset, if you immediately put it into drive again.) It's also possible that particular course of action could cause some counted energy to get "lost" (I haven't ever tried to verify this)...especially if the distance traveled is 0.0.

Depending on the current values on the screen and how much heat/AC you've been using, the sudden addition may or may not be noticeable. (They'll either be averaged over such a large interval it does not result in a big change - or it could be a really huge change.)

I can't guarantee that in every conceivable situation that the energy is counted - I'd have to check the scenario above (put in Drive, sit for a long time using energy, then put in Park - will the energy use get updated? There is potentially a divide by zero issue...)

 

[Wreckless316]

As soon as you start rolling you'll see the energy get counted. That display also only seems to get updated when the odometer is going up.

Let us know. I and others have experimented with this before and that energy does seem to be counted as long as you are not in park. But you do have to get rolling.

It's also possible that putting the car in Park will force the counter to update. (You can try this - as long as you don't get out of the seat the counter won't reset, if you immediately put it into drive again.) It's also possible that particular course of action could cause some counted energy to get "lost" (I haven't ever tried to verify this)...especially if the distance traveled is 0.0.

Depending on the current values on the screen and how much heat/AC you've been using, the sudden addition may or may not be noticeable. (They'll either be averaged over such a large interval it does not result in a big change - or it could be a really huge change.)

I can't guarantee that in every conceivable situation that the energy is counted - I'd have to check the scenario above (put in Drive, sit for a long time using energy, then put in Park - will the energy use get updated? There is potentially a divide by zero issue...)

Going on an hour now with AC on and 3 rated miles used- 1 percent. Car puts itself in park every 15 minutes or so, put it back in Drive/HOLD. "Since Last Charge" has defeinitely not been updated to reflect that.

Bottom line, I think the car has to be rolling. This probably means that energy isn't counted even at stoplights/long stops

Kind of unfortunate. Gives us even less information than we thought!

 

[AlanSubie4Life]

Going on an hour now with AC on and 3 rated miles used- 1 percent. Car puts itself in park every 15 minutes or so, put it back in Drive/HOLD. "Since Last Charge" has defeinitely not been updated to reflect that.

Bottom line, I think the car has to be rolling. This probably means that energy isn't counted even at stoplights/long stops

Kind of unfortunate. Gives us even less information than we thought!

Interesting observations.
Try doing a little driving now...without having removed yourself from the driver's seat. Make a note of the current stats (picture), and then another picture after you roll for 0.1 miles or whatever.

We have done experiments where the car is stationary, and indeed the energy is counted (as long as it is not in park). It is not hard to verify (go to drive, sit with heat on for 3 minutes at full blast, then roll around for 0.2 miles or whatever with the heat off...then exit car, re-enter, turn off the HVAC, go to drive, sit for 3 minutes, then roll around for 0.2 miles...compare). This is kind of a different situation, and one where I could see the accumulated energy getting "lost".

 

[Wreckless316]

Going on an hour now with AC on and 3 rated miles used- 1 percent. Car puts itself in park every 15 minutes or so, put it back in Drive/HOLD. "Since Last Charge" has defeinitely not been updated to reflect that.

Bottom line, I think the car has to be rolling. This probably means that energy isn't counted even at stoplights/long stops

Kind of unfortunate. Gives us even less information than we thought!

I’m going to limit my tracking now to two parts. 1. Energy used while actually driving (wheels turning) which includes accessories/AC/Heat and 2. Energy used while vehicle is stationary in any capacity ( phantom drain/parked and using accessories / etc...

 

[AlanSubie4Life]

I’m going to limit my tracking now to two parts. 1. Energy used while actually driving (wheels turning) which includes accessories/AC/Heat and 2. Energy used while vehicle is stationary in any capacity ( phantom drain/parked and using accessories / etc...

As I said above, I don't think it's quite as bad at tracking as you're assuming here, but as I said, I can't guarantee that it counts all the energy in all situations. I'll do some playing around at some point with this specific scenario that you're experimenting with.

 

[Wreckless316]

Interesting observations.
Try doing a little driving now...without having removed yourself from the driver's seat. Make a note of the current stats (picture), and then another picture after you roll for 0.1 miles or whatever.

We have done experiments where the car is stationary, and indeed the energy is counted (as long as it is not in park). It is not hard to verify (go to drive, sit with heat on for 3 minutes at full blast, then roll around for 0.2 miles or whatever with the heat off...then exit car, re-enter, turn off the HVAC, go to drive, sit for 3 minutes, then roll around for 0.2 miles...compare). This is kind of a different situation, and one where I could see the accumulated energy getting "lost".

I actually did this and it did not change significantly. Only about 1 wh/mile. I think you might be on to something about divisible by zero. Might have an issue calculating If there is nothing to divide by

 

[AlanSubie4Life]

I actually did this and it did not change significantly. Only about 1 wh/mile. I think you might be on to something about divisible by zero. Might have an issue calculating If there is nothing to divide by

Hard to say without the details of the experiment. We can start a separate thread about this if you want, with the detailed test conditions and results. I'll have to set aside 20-30 minutes to run through some new scenarios I have not looked at before.

 

[Wreckless316]

Hard to say without the details of the experiment. We can start a separate thread about this if you want, with the detailed test conditions and results. I'll have to set aside 20-30 minutes to run through some new scenarios I have not looked at before.

Sure why not

 

[jdcollins5]

Fair enough. You can see that really I am only taking issue with what your AC meter reads - everything else makes sense. If the time is right, the voltage and current are correct, only the meter can be reading low (by a lot - 8%). There’s not a lot of wiggle room here.

I expect you should see about 252Wh per rated mile you add (223Wh/rmi / 0.885). Your number seems about 6% low...once again I suspect the meter. I suppose the load could be highly reactive (meaning voltage and current are not in phase) and your meter is right. But does not align with the data I have seen in my case (ChargePoint, Tesla’s data...EPA rating)- looks pretty purely resistive.

This number will not align with the EPA number because those are different rated miles due to the 4.5% buffer (in that framework it would be 234Wh/rmi / 0.885 = 264Wh/rmi which is fairly close to the EPA rating).

I usually charge at night due to TOU rates. Yesterday I charged in the middle of the day so I could monitor the car and the kWH meter.

Charge Screen : 30 mph, 236V, 32A
kWH meter : 239V, 31.1A, 25kWH (meter reads in whole kWH, no decimal)
Charge time : 3.4 hours
Range miles added : 103 miles

AC Charge Constant : 252 X 103 = 25,956 wh = 26 kwh
Charge Screen : 236V X 32A X 3.4 hours = 25,677wh = 26 kWH

kWH meter amp error = 31.1/32 = 0.969 = 3.1% error
kWH error = 25/26 = 0.962 = 3.8% (Rounding can introduce up to 0.9 error. 0.5 to 1.4 = 1)

I wanted to verify meter error so I connected a Fluke 87 meter and the kWH meter to a water heater circuit.

Fluke 87 = 14.0A
kWH meter = 13.58A
Error = 13.58/14 = 0.97 = 3%.

So from what I have calculated the kWH error is between 3-4%. The lack of a decimal in the kWH reading makes it hard to be exact.

 

[AlanSubie4Life]

I usually charge at night due to TOU rates. Yesterday I charged in the middle of the day so I could monitor the car and the kWH meter.

Charge Screen : 30 mph, 236V, 32A
kWH meter : 239V, 31.1A, 25kWH (meter reads in whole kWH, no decimal)
Charge time : 3.4 hours
Range miles added : 103 miles

AC Charge Constant : 252 X 103 = 25,956 wh = 26 kwh
Charge Screen : 236V X 32A X 3.4 hours = 25,677wh = 26 kWH

kWH meter amp error = 31.1/32 = 0.969 = 3.1% error
kWH error = 25/26 = 0.962 = 3.8% (Rounding can introduce up to 0.9 error. 0.5 to 1.4 = 1)

I wanted to verify meter error so I connected a Fluke 87 meter and the kWH meter to a water heater circuit.

Fluke 87 = 14.0A
kWH meter = 13.58A
Error = 13.58/14 = 0.97 = 3%.

So from what I have calculated the kWH error is between 3-4%. The lack of a decimal in the kWH reading makes it hard to be exact.

Cool. Everything makes sense. Note your meter, aside from reading a little low, is measuring cable losses too.

Looking at my table again for efficiency, it is 88.7% for the LR RWD. If not charging to 100% it might be a little higher due to lack of taper. Let’s call it 89%. (So more like 251Wh (AC) per rated mile added.)

AC kWh: 236V*32A*3.4hr = 25.68kWh

DC kWh to battery (expected): 0.89*25.68kWh = 22.85kWh

Actual DC kWh: 103rmi*223Wh/rmi = 22.97kWh.

So that efficiency checks out within 0.5%.

(For clarity, note the charging screen in the car uses the charging constant, so will display 103rmi*234Wh/rmi = 24kWh for this event - that is spurious, it does not represent actual energy added to the battery.)

Your meter (assuming its voltage is correct) measured 3V*31.1A*3.4hr = 0.317kWh of loss in your cable.

But anyway it does appear your meter reads about 3% lower than it should - it reads 25.3kWh, of which 0.3kWh is cable loss, so 25kWh vs. 25.7kWh which is 2.8% off, by the current ratio.

Of course the car amperes may be off slightly - the Fluke experiment suggests the kWH meter is truly 3% low (if Fluke is perfect), suggesting the car might be drawing very slightly more (0.2%) than 32A when it displays 32A and your kWH meter shows 31.1A.

Anyway all this charging data checks out and is what is expected, within the bounds of error of the measurement methods.

Another clear confirmation that charging efficiency is right around 88.5-90% at 32A (7.7kW) in warm weather. This fact can be used to assess the accuracy of your rated miles display in warm weather - take the AC energy you add in on a big charge and multiply by 0.89 to figure out your kWh actually available due to that charge...

 

[jdcollins5]

Cool. Everything makes sense. Note your meter, aside from reading a little low, is measuring cable losses too.

Looking at my table again for efficiency, it is 88.7% for the LR RWD. If not charging to 100% it might be a little higher due to lack of taper. Let’s call it 89%. (So more like 251Wh (AC) per rated mile added.)

AC kWh: 236V*32A*3.4hr = 25.68kWh

DC kWh to battery (expected): 0.89*25.68kWh = 22.85kWh

Actual DC kWh: 103rmi*223Wh/rmi = 22.97kWh.

So that efficiency checks out within 0.5%.

(For clarity, note the charging screen in the car uses the charging constant, so will display 103rmi*234Wh/rmi = 24kWh for this event - that is spurious, it does not represent actual energy added to the battery.)

Your meter (assuming its voltage is correct) measured 3V*31.1A*3.4hr = 0.317kWh of loss in your cable.

But anyway it does appear your meter reads about 3% lower than it should - it reads 25.3kWh, of which 0.3kWh is cable loss, so 25kWh vs. 25.7kWh which is 2.8% off, by the current ratio.

Of course the car amperes may be off slightly - the Fluke experiment suggests the kWH meter is truly 3% low (if Fluke is perfect), suggesting the car might be drawing very slightly more (0.2%) than 32A when it displays 32A and your kWH meter shows 31.1A.

Anyway all this charging data checks out and is what is expected, within the bounds of error of the measurement methods.

Another clear confirmation that charging efficiency is right around 88.5-90% at 32A (7.7kW) in warm weather. This fact can be used to assess the accuracy of your rated miles display in warm weather - take the AC energy you add in on a big charge and multiply by 0.89 to figure out your kWh actually available due to that charge...

Thanks for the further analysis of the charging data. It is starting to make more sense to me. I think you are slowly teaching me how to understand your EPA spreadsheet and the relationships between the constants and the Tesla displayed information.

 

[hari-bhari]

@AlanSubie4Life Could you please help me make sense of these numbers on my 2018 LR RWD? I can't really figure out how much remaining capacity I have in my battery because at charge time I get around 72-73 KwH extrapolating the added KwH but during driving I get only 63-68 KwH. My average efficiency is around 280 Wh/mi and I basically only get 190-220 miles out of a full charge despite being rated for 309 miles.

Here's my post to avoid re-pasting everything.
Can someone help me understand how my battery is possibly 75 kWh?

Thanks in advance!

 

[ran349]

@AlanSubie4Life Could you please help me make sense of these numbers on my 2018 LR RWD? I can't really figure out how much remaining capacity I have in my battery because at charge time I get around 72-73 KwH extrapolating the added KwH but during driving I get only 63-68 KwH. My average efficiency is around 280 Wh/mi and I basically only get 190-220 miles out of a full charge despite being rated for 309 miles.

Here's my post to avoid re-pasting everything.
Can someone help me understand how my battery is possibly 75 kWh?

Thanks in advance!

Based on the table @AlanSubie4Life provided here in this thread, your numbers make sense.
309 miles x .219 ( "discharge constant") (not really a constant but close enough to approximate) = 67.67 kWh. This seems to match what you observed. The lower values you get are probably mostly from vampire losses not showing up in the trip meter, like when stopping for a period of time before continuing your drive.
Based on calculating kWh added, you would get 309 miles x .234 (charge constant from same table) = 72.3 kWh added, which also seems to match your numbers.

 

[AlanSubie4Life]

@AlanSubie4Life Could you please help me make sense of these numbers on my 2018 LR RWD? I can't really figure out how much remaining capacity I have in my battery because at charge time I get around 72-73 KwH extrapolating the added KwH but during driving I get only 63-68 KwH. My average efficiency is around 280 Wh/mi and I basically only get 190-220 miles out of a full charge despite being rated for 309 miles.

Here's my post to avoid re-pasting everything.
Can someone help me understand how my battery is possibly 75 kWh?

Thanks in advance!

Yep, what @ran349 said.

First, I would generally ignore TeslaFi. It's not that it's wrong exactly, but there are a few fields where it actually applies a user-modifiable scalar to the data, so it makes it very difficult to interpret the results. Anyway, it's fine, but you don't need anything other than the trip meter, and to recognize that you lose energy when parked.

Use 234Wh/rmi for each rated mile you add, if you're reading values off the charging screen in the car. (This is what is displayed on the charging screen. Note that it likely does not represent actual energy added to the battery - it is simply displaying the rated miles added, multiplied by 234Wh/rmi, when you have energy display (rather than distance display) selected.)

Use ~219Wh/rmi for each rated mile you consume.

As has been pointed out, this works out 309rmi*234Wh/rmi = 72.3kWh for full battery capacity (including the 4.5% buffer below 0 rated miles - so there are only 69kWh available between 309rmi and 0 rmi).

On the trip meter, I'd expect the max you'd see for a 100% to 0% discharge (without using any buffer) to be 219Wh/rmi*309rmi = 67.6kWh.

This seems to agree with all of your observations. You'd see 63kWh sometimes for discharges where you did not drive continuously from 309 rated miles to 0 rated miles. Use while in park (and in various other unusual use situations which are very rare, actually) is not counted on the trip meter.

 

[Phlier]

Hey @AlanSubie4Life I have reported you!!

How dare you come in here and write volumes of extremely useful information on this forum! This shall not be tolerated! I have reported the OP of this post (because it needs to be a sticky, so I asked them nicely to make it one). Maybe if others ask for it to be stickified it will be?

IMO, having quick access to this information would help avoid so many repetitive posts, and at the very minimum, make it easier to refer guys to where to get this info quickly.

Seriously, man, thanks for all you do.