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

[AlanSubie4Life]

I should have done this a while ago, but I've finally compiled all the data that Tesla provides to the EPA, and put it into a spreadsheet. The detailed 2020 test data is now available, so we can see why things are rated the way they are.
Through formulas, I've also tried to capture what it looks like they are doing as far as charging constants are concerned. The derived charging constants appear to line up with what we have observed so far.

Quick summary:
1) 2020 SR+ is about 6.8% more efficient than it was in 2019.
2) 2020 AWD is about 3.1% more efficient than it was in 2018/2019.
3) 2020 Performance 18" is 6.6% more efficient than it was in 2018/2019.
4) 2020 Model 3 Performance 18" is therefore more efficient than the 2020 AWD, by about 3.6%.

A portion of these improvements appears to be due to small reductions in the dyno loading model, but far from all of it. Whether that rolling resistance change is due to motor controller improvements or not, I don't know. However, the Model 3 Performance and AWD have identical dyno models, so that specific improvement is inherent to the vehicle somehow.

As far as charging constants go, it appears that Tesla is no longer going to "hide" initial degradation, so I think people should start observing loss of capacity (in rated miles) on 2020 Models immediately now, rather than only seeing it when the degradation becomes sufficient to get below the rated mile cap. This may also result in greater variation in initial reported rated miles at 100% charge when comparing new vehicles. We'll see. Interested in reports from 2020 Performance owners (does not sound like 2020 AWD owners have the software to enable it to match EPA, just yet) with brand new vehicles.

This spreadsheet also captures my current model (4.5% buffer) for the discharge/BMS constants, and they seem to align fairly well with what is observed, though there are a couple questions I still have, particularly on the 2020 SR+ which seemed to have a very low capacity battery for the test vehicle.

All of the MPGe & other answers (not in orange - orange are inputs from Tesla) are calculated and match the EPA data, so pretty sure I have the formulas correct on that front for EPA related stuff.

Sources:
EPA Datafile:

Download Fuel Economy Data

EPA Lookup Site:

https://iaspub.epa.gov/otaqpub/

 

[jdcollins5]

Thank you.

What is Porkiness Factor RLHP at 50 mph?

 

[AlanSubie4Life]

Thank you.

What is Porkiness Factor RLHP at 50 mph?

Ha. The Porkiness Factor is my own phrasing. But RLHP I believe that stands for "Road Load Horsepower," and is derived from the model of the coefficients A, B, and C (which I hid from view but are available in the documents referenced). You use those constants A + Bx + Cx^2 where x is the speed in mph, and then divide by 7.5 to get the RLHP. I haven't been able to work out the 7.5 factor but I assume it's somehow unit conversion from lb-ft torque to the HP (these are not the same units...)...it's always 7.5 (at 50 mph anyway - who knows at another speed - it is probably different at different speeds) in any case.

These coefficients I think are derived from some sort of coastdown measurement of the vehicle. But a single datapoint at 50mph is a good reference point for how much power at the wheels it takes to keep the vehicle moving at that speed. On the order of 7.5kW to 10kW. (Which would be 150Wh/mi to 200Wh/mi, which seems in the ballpark for that speed - you'd have to add the drivetrain losses to those numbers and I think ~175Wh/mi (SR+) to ~225Wh/mi (Performance) is very reasonable at 50mph.)

I don't know about these coefficients in general though. I'm always leery of coefficient fits like this. I prefer a physical model - pretty sure if you tried to plug in to the formula and try to get a reasonable number at 80mph the answer would be way off.

Anyway, I just provided it for reference. It is what it is, and is critical for the EPA test, but I'm assuming it is correct.

 

[jdcollins5]

Another question, what are the Initial and Endpoint BMS and Trip Meter Constants?

 

[AlanSubie4Life]

Another question, what are the Initial and Endpoint BMS and Trip Meter Constants?

So, as far as I can tell, in prior Performance vehicles (and other prior model year vehicles), they max out at 310 rated miles regardless of capacity (they can have capacity as high as 78kWh). The constant 245Wh/rmi only adds up to 76kWh though. That means that the rated miles must be “bigger” than 245Wh*(1-0.045) = 234Wh to start with. Indeed, there is some empirical evidence in the form of YouTube videos showing a trip meter constant of 234Wh/rmi (corresponding to BMS constant of about 239Wh/rmi, charging constant of 250Wh/rmi, and total capacity of 77.5kWh) which support that theory that rated miles on brand new vehicles contain more energy than that on older vehicles with capacity below 76kWh.

But once you degrade below the capacity used to calculate the constant (76kWh), then your rated miles start reducing.

So the “initial” constant is what each mile contains prior to any degradation. It would vary from vehicle to vehicle depending on initial capacity. And then the “endpoint” is the “floor” of the energy content per rated mile (234Wh/rmi). Once you reach that floor, any further degradation is accounted for by a reduction in rated miles at 100%.

That is my theory. Note that from what we see so far on the 2020 vehicles, they no longer appear to be approaching this issue of variable initial capacity (that is the motivation I think) the same way - or at least the floor is set a lot higher...in theory a few vehicles may have more energy available initially than that used to define the constant - but probably not nearly as many as before. What happens with those (do they show more miles than 322 (or whatever) or do they inflate the constant?) I have no idea.

Anyway, happy to hear observations to the contrary; it is just a theory which I am trying to fit to the observations - I am not wedded to it!

 

[jdcollins5]

Thanks for the explanations. The best data, and most realistic data, that I have seen. Fits my September 2018 LR RWD experience very well.

 

[AlanSubie4Life]

Sorry it is so small. I suppose I could upload the Excel file at some point.

 

[jdcollins5]

That would be nice.

 

[Fresnel]

Wait so 2020 LR AWD are listed as 322 but don't have the software yet to get that?

 

[AlanSubie4Life]

Wait so 2020 LR AWD are listed as 322 but don't have the software yet to get that?

Yes, the EPA rating for 2020 LR AWD is definitely 322 (it's also on the Tesla website!). What the software shows in the car - I don't have one so I have no idea. I've only seen one report of one above 310 - but there's a thread for that, where that can be discussed - Anyone LR AWDs Showing 322 Miles Fully Charged?

 

[Fresnel]

Yes, the EPA rating for 2020 LR AWD is definitely 322 (it's also on the Tesla website!). What the software shows in the car - I don't have one so I have no idea. I've only seen one report of one above 310 - but there's a thread for that, where that can be discussed - Anyone LR AWDs Showing 322 Miles Fully Charged?

Thanks much! It confuses me seeing that number but only seeing 310 so I was wondering if there was an update coming or what the deal was.

 

[AlanSubie4Life]

Thanks much! It confuses me seeing that number but only seeing 310 so I was wondering if there was an update coming or what the deal was.

Most likely there is an update coming but probably your energy available and efficiency are fine (there's no way to really know!), so that update will probably make no difference (other than change the number to something closer to 322 - it probably won't get all the way there though because of capacity loss on your car). It's just a number. It can be whatever Tesla decides they want it to be as long as they deliver the energy and efficiency. (Though their historical convention has apparently been to match the EPA number.)

 

[Fresnel]

Most likely there is an update coming but probably your energy available and efficiency are fine (there's no way to really know!), so that update will probably make no difference (other than change the number to something closer to 322 - it probably won't get all the way there though because of capacity loss on your car). It's just a number. It can be whatever Tesla decides they want it to be as long as they deliver the energy and efficiency. (Though their historical convention has been to match the EPA number.)

Yeah, it's just annoying seeing my Stats data showing starting at 310 and steadily dropping down to 295 now over time and then reading about it being 322 so I feel at some point it should show a bump up or something to actually show it. It feels more like I don't hav an update.

 

[AlanSubie4Life]

steadily dropping down to 295 now over time

Ignore any numbers you see if the battery is cold. Wait for a day when it is 70 degrees and you take the car for a nice long freeway run and then Supercharge it. If you really have 295 now with a warm battery, you'll probably bump UP to about 299 when you get the update. Hard to know right now because we haven't seen how they will handle this - as I've said they can make the number say whatever they want, really. But that 299 number is what the constant calculations above would suggest 295rmi2019AWD * 248Wh/rmi2020AWD / 245Wh/rmi2019AWD = 299rmi2020AWD

EDITED to correct the calculations above...

As I said, this is wild guessing since there is no clarity currently on the 2020 AWD treatment.

 

[Fresnel]

Ignore any numbers you see if the battery is cold. Wait for a day when it is 70 degrees and you take the car for a nice long freeway run and then Supercharge it. If you really have 295 now with a warm battery, you'll probably bump DOWN to about 291 when you get the update. Hard to know right now because we haven't seen how they will handle this - as I've said they can make the number say whatever they want, really. But that 291 number is what the constant calculations above would suggest 295rmi2019AWD * 245Wh/rmi2019AWD / 248Wh/rmi2020AWD = 291rmi2020AWD

As I said, this is wild guessing since there is no clarity currently on the 2020 AWD treatment.

Have you used the Stats app before? It tracks the battery charge over time and shows a graph of the total battery charge over time.. and it shows it steadily dropping. If I got an update to 322 there should be a bump in the graph.. Though true I suppose I'm not sure exactly when the stats app measures the battery. I assume its after a charge session. But, I think there should still be a bump with some numbers above 310 so.. It still feels like im missing an update

 

[Fresnel]

Here is a screenshot of the graph

 

[AlanSubie4Life]

Have you used the Stats app before? It tracks the battery charge over time and shows a graph of the total battery charge over time.. and it shows it steadily dropping. If I got an update to 322 there should be a bump in the graph.. Though true I suppose I'm not sure exactly when the stats app measures the battery. I assume its after a charge session. But, I think there should still be a bump with some numbers above 310 so.. It still feels like im missing an update

Yeah, I use Stats. My plot looks similar but with a much shallower slope (300 miles @ 13k miles).

Best to ignore the outliers on Stats. They can be from a charge to a lower % (like 60%) which would have large inaccuracy. And cold will affect the results as I said. I have no idea the temperature where you are but probably you're around 300 miles if some of those are warm measurements.

Also I totally screwed up the math above (corrected).

 

[ElectricIAC]

Yes, the EPA rating for 2020 LR AWD is definitely 322 (it's also on the Tesla website!). What the software shows in the car - I don't have one so I have no idea. I've only seen one report of one above 310 - but there's a thread for that, where that can be discussed - Anyone LR AWDs Showing 322 Miles Fully Charged?

Love how it’s advertised...322mi range and 3.2sec 0-60 as if it’s the same model (and I suppose on a stealth that’s “possible”)

 

[Fresnel]

Yeah, I use Stats. My plot looks similar but with a much shallower slope (300 miles @ 13k miles).

Best to ignore the outliers on Stats. They can be from a charge to a lower % (like 60%) which would have large inaccuracy. And cold will affect the results as I said. I have no idea the temperature where you are but probably you're around 300 miles if some of those are warm measurements.

It's around 19f - 30f recently. I see what you mean maybe the weather is depressing the reading more then it would be

 

[AlanSubie4Life]

Love how it’s advertised...322mi range and 3.2sec 0-60 as if it’s the same model (and I suppose on a stealth that’s “possible”)

Not it's not. 322mi range and 4.4sec (including rollout) vs. 310mi range and 3.2sec (not including rollout).