Also to resurrect this thread, I can confirm now that based on about a couple hundred thousand canbus data points, the prerefresh 90D is reporting miles remaining @ 100% based on nominal capacity which includes brick protection. Yet, SOC reports 0% (and thus no more driving) at nominal remaining 4kwh (0kwh usable remaining). The complete transition to calculating remaining range changes below 20% to properly use usable remaining for it's number reported to the driver. The only conclusion I can draw is that Tesla is hiding the true degredation on the 90kwh batteries. I say that because it's not a rounding error or a fudging of math, the numbers when plugged into the equations are exact.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Ok, to answer my own question using some industry definitions: SOC indicates the battery remaining capacity, and SOE indicates the battery remaining energy. More specifically, SOE is a better estimation of the amount of energy we can expect to be able to actually get out of the battery at this moment in time, taking into account factors like immediate cycling history, temperature, etc...
I wouldn't jump to conclusions so quickly. I don't necessarily think Tesla is doing anything different with the 90 packs than with any of their other batteries in terms of BMS and reported capacity available to the user... The amount of energy we can extract from the battery is far more dramatically impacted by the factors I listed above (like temperature, immediate cycling history, etc) at lower states of charge... Driving hard at low states of charge, for example, will reduce the amount of energy you can get out of the pack for that discharge cycle. It is very possible Tesla picked the 20% SoC mark to start modeling SoE (what gets displayed to the user as remaining capacity) based on these factors, as this is where they would matter the most. And SoE is certainly the right thing to show to the user, not SoC. SoC doesn't help me know how much range I really have left, SoE does!
Why would they "estimate" so specifically? In no situation could an owner ever achieve the rated range by using the EPA consumption as a model.
Great thread - been following with interest. But just wondering if someone could explain why I'm seeing 299 Rated Miles (268m at 90%) from a 2015 S85? Whilst I'm not complaining, that seems higher than most.
By wk057's formula on the 1st page, 295 Wh/Rated Mile x 299m = 88.3 kWh usable!
Thanks, Steve.
By wk057's formula on the 1st page, 295 Wh/Rated Mile x 299m = 88.3 kWh usable!
Thanks, Steve.
Bizarre. Your ideal shows as less than your rated miles in the second image. Ideal should be higher. Something seems off.
Oh yeah, their terminology is way different and confusing, and I'll bet that program is getting the names wrong.
In the U.S., it's "ideal" for the high fantasy one and "rated" for the more realistic one.
In Europe, the NEDC is such a pie-in-the-sky idealistic rating system, their high one is called "rated", and the lower one you can actually get is called "typical".
This is where is gets good. Below 80%, it starts to use some number between nominal and usable. So it's trying it's hardest not to make it known (hence my suspicion it's VERY intentional). However, I've gone on long trips where I'll catch it just shoot 4-6% down in a couple of seconds if it's below 20%. But yes, the miles do tick away quicker to keep up with the chance in calculation
This would be such a great project for some engineering student.
Give them a Tesla and an oval track.
They drive just the right speed on cruise control to maintain the rated Wh/km. If a bit too high slow down, a bit too low speed up.
Then record remaining range and SoE and dist traveled at frequent intervals (I just take pics of days).
Finally plot. It would answer the above questions about a transition point.
Give them a Tesla and an oval track.
They drive just the right speed on cruise control to maintain the rated Wh/km. If a bit too high slow down, a bit too low speed up.
Then record remaining range and SoE and dist traveled at frequent intervals (I just take pics of days).
Finally plot. It would answer the above questions about a transition point.
Working on a pretty graph today.
And you don't necessarily need a track because you can simply adjust for consumption. But good idea. Heck, if I had a track, I'd do it.
Grabbed a few data points from my MS 75D... Based on a limited data set (~30 data points, some while supercharging, others while driving over a pretty wide temperature range), the SoE SoC relationship appears to be pretty linear, with a crossover happening somewhere around 96-97%.
I definitely don't see a cliff or major change @20%...
I definitely don't see a cliff or major change @20%...
That's pretty disturbing then if it's just the 90. What is your rated range at 100% based on? Usable or nominal?
Appears to be based on usable.
My pack is reporting 70.9kWh at 100% currently, showing 248mi on the dash.
70.9/285*1000= 248.77mi
What’s disturbing about the data? Seems fine to me?
So is that 70.9 kWh defined as the nominal full pack or the usable full pack? And when you charge your car from 0%, does the car say you added 248 x .285= approx. 70.7 kWh to the battery?
Hang on, let's be *very* clear about this. You have 2 numbers from the BMS, usable full and nominal full. Are you confirming that if you divide your usable full number by EPA consumption, you get the Rated Remaining number within a few decimal places, AND that is the number that is showed to you on the car's screen?
Because let's be crystal clear on what this means. The way your car is reporting at 100% is the correct way. Because usable is, in fact, what you can use. This would mean that the 90's are intentionally misreporting higher available rated miles than they could ever deliver under conditions set and verified by the EPA tests.
I got here via the Model S REST API post where I am trying to figure out the values returned from the Tesla API. As the API doesn't return any data about consumption per kilometer or mile I started to do some research.
I have a Tesla Model S100D from 2018 and I'm from the Netherlands. A quick assumption here is that (api) data is the same for all European Tesla's but different from the US models (things like rated miles).
Based on this data the *only* thing that shows consumption is the energy display estimate. Based on that I've taken that consumption rate (261Wh/km) and converted these to both the battery_range which is *rated* in the European UI and ideal_battery_range.
As you can see when extrapolating this to 100% based on the estimate consumption my battery pack has 97.25kWh usable, a rated consumption of 156.19Wh/km and ideal consumption of 194.49Wh/km
Comparing the numbers from the UI and the API and converting from/to miles it means that:
- Typical in the UI is mapped to "ideal_battery_range" in the API
- Rated in the UI is mapped to "battery_range" in the API
Can anyone confirm if these consumption numbers from "rated" and "ideal" are correct?
I have a Tesla Model S100D from 2018 and I'm from the Netherlands. A quick assumption here is that (api) data is the same for all European Tesla's but different from the US models (things like rated miles).
Code:
User interface:
soc: 55%
typical: 276km => 171.49m
rated: 343m => 213.13m
Energy display (average):
10km: 198Wh/km, 271km (168.39km)
25km: 251Wh/km, 214km (132.97km)
50km: 261Wh/km, 205km (127.38km)
API: (via Teslafi.com account info)
battery_level: 55%
battery_range: 213.21m => 343.12km (rated in GUI)
est_battery_range: 127.59m => 205.33km
ideal_battery_range: 171.23m => 168.39kmBased on this data the *only* thing that shows consumption is the energy display estimate. Based on that I've taken that consumption rate (261Wh/km) and converted these to both the battery_range which is *rated* in the European UI and ideal_battery_range.
Code:
Current ranges:
rated: 342.45km, 212.79m (battery_range)
ideal: 275.02km, 170.89m
estimated: 204.93km, 127.34m
Max ranges:
rated: 622.64km, 386.89m (battery_range)
ideal: 500.04km, 310.71m
estimated: 372.61km, 231.53m
Consumption:
rated: 156.19Wh/km, 251.36Wh/m (battery_range)
ideal: 194.49Wh/km, 312.99Wh/m
estimated: 261.00Wh/km, 420.04Wh/m
Pack capacity (SoC: 55%):
current: 53.49kWh
max: 97.25kWh
Factors:
to ideal: 1.24518696237346 (from battery_range)
to estimated: 1.67103816554107 (from battery_range)As you can see when extrapolating this to 100% based on the estimate consumption my battery pack has 97.25kWh usable, a rated consumption of 156.19Wh/km and ideal consumption of 194.49Wh/km
Comparing the numbers from the UI and the API and converting from/to miles it means that:
- Typical in the UI is mapped to "ideal_battery_range" in the API
- Rated in the UI is mapped to "battery_range" in the API
Can anyone confirm if these consumption numbers from "rated" and "ideal" are correct?
Similar threads
