Tesla Official Statement on Range

[AlanSubie4Life]

Can I please get some feedback on whether this comparison is valid? I did some searching and browsing here and on Tesla's site and didn't see a clear answer (possible "ebcak").

I charge my car nearly every night in the garage. On a single trip after unplugging, I see distance traveled and Wh/mi. Multiplying these gives me units of energy consumed (E_c]. Also, my range estimate next to the battery icon decreases by X. X*78.27 kWh/310 mi (EPA rating for 2018 LR AWD) is also units of energy, amount decreased in actual battery (E_d).

As long as I'm not using energy while parked, E_d should equal E_c, correct? (Assuming E_c does not display parked consumption, please let me know if this is dis/proven). My car has E_c about 76% of E_d rather consistently, which seems like something is really off (and also indicates even in the summer with HVAC off my lifetime best consumption was never below 266 Wh/mi actual battery drop, which is crazy in comparison to people I see averaging in low 200s). The service center wants to focus the conversation on what affects consumption rate, even though I try to communicate the battery capacity is what seems off.

Are my assumptions/expectations invalid? If so, can someone please point out to me the why?

You can look at the posts elsewhere (the Constant sticky is a good place to look).

Your first equation should use 76kWh and 325rmi (to give you 234Wh/rmi). Remember you got an unlock.

The E_c does not include energy used in park, that's correct.


In short, you only have 95.5% of your total battery energy available between 100% (your full rated miles) and 0%. This is a key thing to remember when doing these calculations if you want to do them.

And your full battery energy, including buffer below 0 = Vehicle Constant * Rated Miles @ 100%. (Since you have a 2018 RWD that constant is 234Wh/rmi)

Finally, the trip meter seems to typically read low (shows less energy consumed than actually consumed) by about 1%, though it can also read high in the right circumstances (temp changes, etc., etc.). These are small discrepancies though compared to the first point. So E_d and E_c as you define it (after making corrections above) will differ by that very small %.

As an example, let's say you start at 290 rated miles.

You drive 50 miles at 266Wh/mi. Your trip meter says you used: 13.3kWh.

That's a little (1%) low as I said, so 13.3kWh/0.99 = 13.43kWh was the actual consumption, probably.

Since each displayed rated mile is not 234Wh/rmi, but instead 95.5% of that, each displayed rated mile is 223Wh.

So you'll see 13.43kWh/0.223kWh/rmi = 60rmi tick off, so you'll end at 230 rated miles for this trip.

And remember after you hit 0 rated miles, if you drive really slowly and carefully, you'll be able to drive (perhaps, if you're lucky) and use that final 4.5% of your battery. (Don't do this.) But unless you do that you'll never see the full battery energy on your trip meter (it'll be short by 4.5%). That buffer is 4.5% of your full pack energy.

 

[3ngineer]

You can look at the posts elsewhere (the Constant sticky is a good place to look).

Your first equation should use 76kWh and 325rmi (to give you 234Wh/rmi). Remember you got an unlock.

In short, you only have 95.5% of your total battery energy available between 100% (your full rated miles) and 0%. This is a key thing to remember when doing these calculations if you want to do them.

And your full battery energy, including buffer below 0 = Vehicle Constant * Rated Miles @ 100%. (Since you have a 2018 RWD that constant is 234Wh/rmi)

Finally, the trip meter seems to typically read low (shows less energy consumed than actually consumed) by about 1%,
... So E_d and E_c as you define it (after making corrections above) will differ by that very small %.

As an example, let's say you start at 290 rated miles.

You drive 50 miles at 266Wh/mi. Your trip meter says you used: 13.3kWh.

That's a little (1%) low as I said, so 13.3kWh/0.99 = 13.43kWh was the actual consumption, probably.

Since each displayed rated mile is not 234Wh/rmi, but instead 95.5% of that, each displayed rated mile is 223Wh.

So you'll see 13.43kWh/0.223kWh/rmi = 60rmi tick off, so you'll end at 230 rated miles for this trip.

And remember after you hit 0 rated miles, if you drive really slowly and carefully, you'll be able to drive (perhaps, if you're lucky) and use that final 4.5% of your battery. (Don't do this.) But unless you do that you'll never see the full battery energy on your trip meter (it'll be short by 4.5%). That buffer is 4.5% of your full pack energy.

To clarify, I have an AWD not RWD, so that would still be 310 right? Only the RWD got the unlock I thought? Where does 76kWh come from in the sticky calculation? I found AWD on page two of the constants sticky (didn't notice it as a slash after the P) 2020, 2019, 2018 Model 3 Battery Capacities & Charging Constants. There it says 79.218; I saw this communization from Tesla (page 6) that indicates it was the 78.27 for a RWD (assuming all LR had same pack in 2018). What is this deduction on top of the 95.5%?

Trusting the sticky screenshot, I should be using 76 kWh and 310 mi, so my E(nergy)_d(ecrease) should be X (rated mile decrease) * 76 kWh / 310 * 95.5%.

And my E(nergy)_c(onsumed) should be displayed miles run * Wh/mi / 99% for under reporting.

Using my prior metric of E_c/E_d (previously .76), this is now .793 if my midnight math is right (.76/.99*75/76/.955).

This still indicates my car is suffering an energy decrease of 25% more (1/.793 -1) than it says it is consuming while driving, right?

 

[3ngineer]

Forget the 76 vs 79.x vs 78.27 question; I'll ask for more specifics in that thread. E_d vs E_c discrepancy question remains.

 

[AlanSubie4Life]

I have an AWD not RWD, so that would still be 310 right? Only the RWD got the unlock I thought?

Sorry, misread your post. So use 245Wh/mi and 234Wh/mi (discharge) above instead of 234Wh/mi and 223Wh/mi.

What is this deduction on top of the 95.5%?

The car does not show degradation until your energy drops below 76kWh (for 2018/2019 only). Before that happens each rated mile contains more energy. But we don’t need to worry about that since you undoubtedly have a lot less than 76kWh at this point.

Provide a specific example, and we can reconcile your numbers.

 

[3ngineer]

Provide a specific example, and we can reconcile your numbers.

So before going in for service, consistently this behavior:

305Wh/mi, 23.0 miles traveled, range available decreased by 36.

After going in for service, consistently accurate reporting/range loss, e.g.:

271 Wh/mi, 53.4 miles traveled, range available decrease by 63.

Before about a month ago, I was using the % available instead of miles available and wasn't logging detailed data, but the product of consumption times distance was quite far from the percentage battery loss multiplied by 76kWh

 

[AlanSubie4Life]

So before going in for service, consistently this behavior:

305Wh/mi, 23.0 miles traveled, range available decreased by 36.

After going in for service, consistently accurate reporting/range loss, e.g.:

271 Wh/mi, 53.4 miles traveled, range available decrease by 63.

Before about a month ago, I was using the % available instead of miles available and wasn't logging detailed data, but the product of consumption times distance was quite far from the percentage battery loss multiplied by 76kWh

I meant some hard data describing scenario, then we can analyze. Somewhat longer drives are better of course. But you’ll never see it match 76kWh because of the 4.5% buffer.

Are the numbers you quote above from single uninterrupted drives with no time spent in park? It’s easiest to take a picture of the screen before and after and capture the “since x:xx” trip meter numbers.

 

[3ngineer]

I meant some hard data describing scenario, then we can analyze. Somewhat longer drives are better of course. But you’ll never see it match 76kWh because of the 4.5% buffer.

Understood; my assessment of "accurate"/"inaccurate" was based on multiplying 76kWh * 0.955 (100%-4.5%).

Are the numbers you quote above from single uninterrupted drives with no time spent in park? It’s easiest to take a picture of the screen before and after and capture the “since x:xx” trip meter numbers.

Yes, uninterrupted drives with no time in park. The Wh/mi and miles traveled are from the exact screen you referenced, and I logged the range displayed next to the battery icon right before shifting to drive and right after shifting to park.

I don't have the pictures for most of these, but have been keeping logs with each of these data points. Let me know if there is something more specific for "hard data" that I'm missing. Thanks for your help!

 

[Cris disrupts]

Colleagues from Europe, are there any reports about max range in cold weather of the Made in China LFP batteries?

 

[JCtx]

Hey guys, have a question regarding power usage: Are LR M3s always using both motors? And equally? Or is it just one at slower speeds? And if so, which one: F or R? Just curious about if using 1 vs 2 motors in certain situations would result in less power used (more range). Thx.

 

[AlanSubie4Life]

Yes, uninterrupted drives with no time in park. The Wh/mi and miles traveled are from the exact screen you referenced, and I logged the range displayed next to the battery icon right before shifting to drive and right after shifting to park.

I don't have the pictures for most of these, but have been keeping logs with each of these data points.

I can't make sense of your data as the resulting constant is highly inconsistent (not constant). I'd recommend taking pictures before and after each segment capturing the trip meter data and the battery gauge (in km or miles), at your convenience, so as to help center the discussion here. It's so easy to make a transcription error, and then we're just chasing bad data.

I'm not saying your data is wrong, I just can't make sense of it. It's rare to have such widely varying results.

Are LR M3s always using both motors? And equally? Or is it just one at slower speeds? And if so, which one: F or R? Just curious about if using 1 vs 2 motors in certain situations would result in less power used (more range). Thx.

It depends. Tesla tries to make the most optimal choice available at all times. In Model 3, the rear motor is permanent magnet so is more efficient than the front induction motor. The front motor and associated drag of the axles, churning the gear oil, etc., is once of the reasons the LR RWD is still (arguably) the range king. Too bad they don't make it any more, though the efficiency gap (between RWD and AWD) has narrowed since the early days. But they haven't done an LR RWD EPA test in a while, so it would probably have a very, very high range if they did retest it with heat pump, high capacity cells, etc. Probably would exceed 400 rated miles if they added 3-5% capacity (if you extrapolate directly (which isn't quite right) from the 2021 SR+ with no added battery density, you'd see 390 miles (263rmi*46p/31p = 390rmi). They should bring it back, just for the glory of it! No one would buy it probably because no one seems to care about range (lol).

That would be a great lineup, with 5% capacity increase everywhere, we'd be looking at:
SR+ 275 rated miles (444km)
LR AWD 370 rated miles (595km)
LR AWD P 320 rated miles (507km) (Only 2-3% increase since it's already increased a bit.)
LR RWD 400-410 rated miles (644km)

 

[3ngineer]

I can't make sense of your data as the resulting constant is highly inconsistent (not constant). I'd recommend taking pictures before and after each segment capturing the trip meter data and the battery gauge (in km or miles), at your convenience, so as to help center the discussion here. It's so easy to make a transcription error, and then we're just chasing bad data.

I'm not saying your data is wrong, I just can't make sense of it. It's rare to have such widely varying results.

I agree it's quite odd, hence why I asked Tesla to look into it!

I have a number of other examples from hand jammed logs which I think would be quite unlikely to have so many transcription errors in only one direction and magnitude, but I understand the desire to see first hand and eliminate potential error sources.

As I said, drives since the last visit to service have been consistently true to the charging constant estimate in the other thread, but as soon as it starts acting up again I'll report back with some pics.

 

[AlanSubie4Life]

I agree it's quite odd, hence why I asked Tesla to look into it!

I have a number of other examples from hand jammed logs which I think would be quite unlikely to have so many transcription errors in only one direction and magnitude, but I understand the desire to see first hand and eliminate potential error sources.

As I said, drives since the last visit to service have been consistently true to the charging constant estimate in the other thread, but as soon as it starts acting up again I'll report back with some pics.

Thanks. Yeah, if the BMS misbehaves and the estimated available energy jumps around wildly (which can happen sometimes I guess), then the "constants" will tend to end up not being applicable (though generally speaking they are still correct), since energy is being "lost" or "gained" the the BMS. I guess we'll see. Not sure I can explain a consistent bias in one direction for that sort of misbehavior though.

 

[intelligator]

2018 3 AWD, 35K miles, charge to 90%, 252 miles. Getting tired of throwing energy into the battery to get it to 90% but only showing 252. The battery drops 1% every few minutes when driving in the city/town, average 35 mph. Degradation started after applying 2020.40.3 / 2020.40.4 ... Now on 202.48.10 , the trend is directionally incorrect. Meanwhile, Tesla sent me the range URL as if that will be okay. I keep throwing extra Kwh at the battery and not getting the return. Average temperatures just fell to low 50's; had nothing to do with winter coming or winter here.

Anyone else with similar correlation of battery degradation after applying update?

 

[jjrandorin]

(moderator note: locking this thread as it has turned into another version of the "range loss over time" master thread, which already exists. Also removing it from sticky thread status, as I believe it has outlived its usefulness as a sticky thread.)