Model 3 Performance Battery Degradation One Month (Story)

[TimothyHW3]

Ok, I checked the screenshots and there might be a 0.2 Wh/km deviation between the range calculated in the energy graph and the one under the battery icon.

I think the projected energy graph calculated with 155or 156 Wh/km (straight line) and the battery estimated seems to be calculated with 153Wh/km.

Unfortunately I was driving and didn't have SMT running and by the time I stopped the avg changed and didn't see what the nominal kWh were.

But still is not 5Wh/m, more like 1-2Wh/m if anything. Barely a difference.
I will have to test it more thoroughly with ScanMyTesla, but the general idea stands.

 

[VT_EE]

holy hell, could this really be the answer?? Am I reading this right? If the rated range is dynamic relative to driving habits (instead of a fixed EPA figure, e.g., 310 miles), then driving efficiency would affect this figure...I’m much more willing to bet I drive 15% faster/less efficiently than the EPA estimates...this also means that using rated range as a degradation yardstick is completely incorrect, and TeslaFi and Stats have incorrectly labeled their datasets (Stats got people to pay for a separate app for that, no less). Could this really be the case?

It certainly could be. The max value gets set lower because of a lead foot not degradation.

 

[TimothyHW3]

It certainly could be. The max value gets set lower because of a lead foot not degradation.

No, this is incorrect. When you charge the car it always uses the typical rated consumption to calculate range. The max value you see while charging has zero "lead foot" interference.

Only when you unplug and start discharging the car, either driving or idling, does it take the actual consumption into the calculation (from ALL the appliances in the car, HVAC, motors etc.)

This topic has been beaten to death and should be common knowledge by know...

 

[NOLA_Mike]

As I understand it, "Rated Range" might be more accurately labeled "EPA Rated Range" in the United States.

The Model 3, having been run thru the EPA's test cycle, consumed ~243 wh/mile (0.243 kWh/mile) during this test cycle. Therefore, "Rated Range" as displayed in the car, is the battery's current state of charge (as determined by the BMS) / 0.243. Nothing more, nothing less.

If the BMS determines the battery pack's current state (at this instant) is 35 kWh remaining then it will display a "Rated Range" of 35 kWh / 0.243 kWh/mile = 144 miles.

The only variable here is the BMS accurately determining the battery's current state at any given instant.

Mike

Edited to correct some units - kW vs kWh (oops)

 

[apacheguy]

OK so I got a reply about my battery range.

It looks like the Technician completed an over-the-air review of your vehicle and it will not require a visit to the service center as the findings appear to be normal. Please find his notes below for review.

...

• The Mileage calculation shown on the vehicle’s display is not an indication of HV battery capacity. It is a calculation meant to give an approximation of energy economy pertaining only to the estimated available distance of travel the vehicle is capable of at its current state of charge based on prior energy usage data.
.

Can’t believe Tesla is still using this false explanation. Rated range is based on the capacity of the battery and it is not influenced by “prior energy usage data.”

 

[AlanSubie4Life]

Yes and I have verified it - energy graph and battery estimate match at the exact line...I don't see how that could be different in miles, but it might be a display error or a rounding error. Change to km.

Take a picture and post here showing that - - you can see my post from before showing that it works the way I say - see this picture, where the range estimate is less than the remaining rated kilometers, in spite of consumption being below the rated line - so clearly you have to be even lower consumption to get the two numbers to match numerically.

358km projected, 365km remaining, when efficiency better than rated line

Specifically, to make it match, you’d have to do about 155Wh/km * 358/365 = ~152Wh/km

Which is at least 4Wh/km below the rated line (around 156Wh/km - see the picture)...it is possible that due to rounding errors that 153Wh/km would make the two numbers match.

 

[AlanSubie4Life]

with 155or 156 Wh/km (straight line) and the battery estimated seems to be calculated with 153Wh/km.

But still is not 5Wh/m, more like 1-2Wh/m if anything. Barely a difference.

Missed this post before my post above. Sorry.

Note this means I am exactly correct; 156Wh/km - 153Wh/km = 3Wh/km

This is 5Wh/mi - which is exactly what I said.

You can verify it yourself - I’ve done it a few times so there is no doubt in my mind. It is inconsequential of course but just wanted to make sure statements here were correct so people don’t get confused when they see different behavior!

 

[VT_EE]

No, this is incorrect. When you charge the car it always uses the typical rated consumption to calculate range. The max value you see while charging has zero "lead foot" interference.

Only when you unplug and start discharging the car, either driving or idling, does it take the actual consumption into the calculation (from ALL the appliances in the car, HVAC, motors etc.)

This topic has been beaten to death and should be common knowledge by know...

I'm sorry, but you don't know what they use to determine the max rated value. Yes, many people have provided evidence that the consumption factor seems fixed, but that does not mean that Tesla is not using some algorithm now to come up with the initial figure, which includes some variables related to driving history.

Theoretical example: Sue is historically light on the accelerator and lives in a warm climate. Her display shows 310miles range at 100%, which decreases while driving at the fixed 230wh/mi EPA rate. John likes to abuse the accelerator a little and lives where the weather is brisk. His display shows 280miles range at 100%, and decreases while driving at the fixed 230wh/mi EPA rate. Neither has abnormal battery degradation, but their 100% range difference is a result of some variables Tesla has started using to make the battery meter range more realistic.

Something obviously changed after V10 when all this complaining about rated range started. It is highly unlikely that everybody's battery is now experiencing excessive degradation that started with a software update. I would be very careful stating something as fact just because Tesla has historically calculated metrics one way. They are free to, and do, change just about everything with their OTA updates.

 

[AlanSubie4Life]

The Model 3, having been run thru the EPA's test cycle, consumed ~243 wh/mile (0.243 kWh/mile) during this test cycle. Therefore, "Rated Range" as displayed in the car, is the battery's current state of charge (as determined by the BMS) / 0.243. Nothing more, nothing less.

If the BMS determines the battery pack's current state (at this instant) is 35 kWh remaining then it will display a "Rated Range" of 35 kWh / 0.243 kWh/mile = 144 miles

This is roughly correct - the numbers you quote are for the MR though - that is where the rated line is at about 243Wh/mi. But the way the calculation is done is different than you describe.

You have to use the corrected 0.7*(0.4*hwy + 0.6*city) formula for the miles. For the AWD your formula works out to be 255Wh/mi (79.2kWh/309.5mi) and the line ends up at 250Wh/mi, and the calculations are done using 245Wh/mi as shown above in the picture.

 

[AlanSubie4Life]

I'm sorry, but you don't know what they use to determine the max rated value. Yes, many people have provided evidence that the consumption factor seems fixed, but that does not mean that Tesla is not using some algorithm now to come up with the initial figure, which includes some variables related to driving history.

Theoretical example: Sue is historically light on the accelerator and lives in a warm climate. Her display shows 310miles range at 100%, which decreases while driving at the fixed 230wh/mi EPA rate. John likes to abuse the accelerator a little and lives where the weather is brisk. His display shows 280miles range at 100%, and decreases while driving at the fixed 230wh/mi EPA rate. Neither has abnormal battery degradation, but their 100% range difference is a result of some variables Tesla has started using to make the battery meter range more realistic.

Something obviously changed after V10 when all this complaining about rated range started. It is highly unlikely that everybody's battery is now experiencing excessive degradation that started with a software update. I would be very careful stating something as fact just because Tesla has historically calculated metrics one way. They are free to, and do, change just about everything with their OTA.

What contradicts this argument above is that if you measure the discharge energy for a given number of rated miles at a reasonable consumption wattage (say 0.2C), you’ll get an energy that is very close to the rated miles * discharge “constant”.

It is true that that discharge constant seems to change slightly (by up to about 2%) due to various unknown factors, but that means that within about 2%, the number of rated miles displayed is not dependent on past driving behavior - it shows it is a representation of energy available (approximately).

Anyone who cares to do so can verify this for themselves - even on the latest V10 software. The guy posting his data in this thread has already shown this, in fact. And if he posts the data from the trip meter (rather than TeslaFi) it will be even more clear.

People who have CAN bus data will also likely be able to prove that in general the 100% rated range number correlates very well with the “full kWh” value. In other words someone showing 270 rated miles at a full charge is going to get a BMS readback of something like 68kWh full rather than a new car’s 78kWh (these are approximate numbers)

 

[VT_EE]

What contradicts this argument above is that if you measure the discharge energy for a given number of rated miles at a reasonable consumption wattage (say 0.2C), you’ll get an energy that is very close to the rated miles * discharge “constant”.

It is true that that discharge constant seems to change slightly (by up to about 2%) due to various unknown factors, but that means that within about 2% the number of rated miles displayed is not dependent on driving behavior - it shows it is a representation of energy available.

Anyone who cares to do so can verify this for themself.

People who have CAN bus data will also likely be able to prove that in general the 100% rated range number correlates very well with the “full kWh” value. In other words someone showing 270 rated miles at a full charge is going to get a BMS readback of something like 68kWh full rather than a new car’s 78kWh (these are approximate numbers)

And that may very well be true. The problem I have with all this is the fact that everything started with the version 10 software update. Whatever is causing this "degradation" is software related and likely not actual battery degradation. I'm sure there are a few posters here that do actually have a battery problem, but the majority do not.

 

[AlanSubie4Life]

. The problem I have with all this is the fact that everything started with the version 10 software update.

If you review the history you will see it started happening before V10. My (small) loss of range happened with a software update earlier in the summer at around 6000 miles. There were many others reporting similar results at the same time.

I am not saying it was not correlated with V10 for some people - as with many Tesla changes, not all cars are affected by a particular update in the same way, due to Tesla’s internal methods I suppose.

I agree it seems correlated with software - they may have capped the lower discharge limit or the upper charging voltage - no idea. But no guarantees it will be range loss that ever comes back - but yes, likely not true degradation. It is loss of available capacity though.

I think it is a combination of true capacity loss due to degradation (primarily the normal initial rapid but small battery degradation) and software changes.

 

[TimothyHW3]

I'm sorry, but you don't know what they use to determine the max rated value.

No, I actually do. It is very easily reproducable and measurable. This is how Tesla has been doing this for years.

The max rated value is and always was calculated by the following formula:

Wh total/ Typical constant = rated range at 100%

Something obviously changed after V10 when all this complaining about rated range started.

No, something obviously did not change. Even Bjorn Nyland, the father of this conspiracy theory, backed up once he got access to the CAN Bus...

The only thing that V10 did was to introduce more phantom drain when stationary due to the default setting on smart summon being set to always listen and the car never going to sleep. But as soon as you disable that option all is good.

Nothing else has changed, not in the voltage and not in the consumption constant - also easily prooven.

They are free to, and do, change just about everything with their OTA updates.

They are free, but haven't done this in the last 4-5 years. At least not to the way they display the rated range at charging - the formula is the same across all Teslas.

The only thing Tesla can and has done in the past is: to take Voltage out of the top from S85 (instead of 4.2 they charge to 4.15 or something) which gives you less Wh total hence less range (see math above, Wh is a smaller number)

And the other thing they did was put a lower Typical Constant on the RWD LR - this gives you a higher range due to lower denominal under the division.

And these two cases are just a further proof that the math is correct. And it is a very simple math - 2nd grade math?

 

[AlanSubie4Life]

The only thing that V10 did was to introduce more phantom drain when stationary due to the default setting on smart summon being set to always listen and the car never going to sleep. But as soon as you disable that option all is good.

Nothing else has changed, not in the voltage and not in the consumption constant - also easily prooven.

There absolutely are people who saw immediate rated range changes on software updates. What the reason is for those changes I don’t know. But saying that nothing changed on any car is not a dispassionate reading of the data. There is plenty of Stats and Teslafi data showing a clear connection to these changes in rated range to a software update in SOME cars. Nyland may not have had one of those cars.

 

[VT_EE]

No, I actually do. It is very easily reproducable and measurable. This is how Tesla has been doing this for years.

The max rated value is and always was calculated by the following formula:

Wh total/ Typical constant = rated range at 100%


No, something obviously did not change. Even Bjorn Nyland, the father of this conspiracy theory, backed up once he got access to the CAN Bus...


They are free, but haven't done this in the last 4-5 years. At least not to the way they display the rated range at charging - the formula is the same across all Teslas.

The only thing Tesla can and has done in the past is: to take Voltage out of the top from S85 (instead of 4.2 they charge to 4.15 or something) which gives you less Wh total hence less range (see math above, Wh is a smaller number)

And the other thing they did was put a lower Typical Constant on the RWD LR - this gives you a higher range due to lower denominal under the division.

And these two cases are just a further proof that the math is correct. And it is a very simple math - 2nd grade math?

OK, so what is your theory as to why there is a sudden rash of degradation complaints that seem to correlate well with software updates. I find it hard to believe nothing has changed and that the problem is a large number of batteries all of a sudden experiencing excessive degradation.

 

[TimothyHW3]

This has been also beaten to death and I refuse to repeat myself. Click on my name and read yourself through.

 

[TimothyHW3]

There absolutely are people who saw immediate rated range changes on software updates. What the reason is for those changes I don’t know. But saying that nothing changed on any car is not a dispassionate reading of the data. There is plenty of Stats and Teslafi data showing a clear connection to these changes in rated range to a software update in SOME cars. Nyland may not have had one of those cars.

Well, you are just proving my point. If it was a software update across the board, we would have spotted it.
The voltage would have dropped and/or thr constant changed. But since none of this has been measured by all the people with access to the can then -
If it is only "just some cars" then this is 100% proof that nothing changed as updates are the same for the identical models.

The temps dropped from 20C to -2C where I live. Of course my consumption rose and the range tanked. Same for the kWh and the snowflake due to cold weather. And my phantom drain went from 2km to 10km a day. Same setup... And there was no update in between just weather. I don't complain here about it...

It is just "user error"( people who don't understand basic math or how the range is calculated. Not have an understanding what a buffer is and how Tesla uses it)

 

[AlanSubie4Life]

Well, you are just proving my point. If it was a software update across the board, we would have spotted it.
The voltage would have dropped and/or thr constant changed. But since none of this has been measured by all the people with access to the can then -
If it is only "just some cars" then this is 100% proof that nothing changed as updates are the same for the identical models.

It is not uncommon for software updates to apply differently to different vehicles (because not all vehicles are the same). There are continuous changes to the underlying hardware in the vehicles - see the steppings on the motors for example, etc.

So you would have to have CAN bus data before and after from one of the vehicles that showed a rated range change after a software update, to say anything definitive about what is happening. People with CAN access does not prove anything - because the issue may not have applied to their vehicle.

 

[gaspi101]

I'm sorry, but you don't know what they use to determine the max rated value. Yes, many people have provided evidence that the consumption factor seems fixed, but that does not mean that Tesla is not using some algorithm now to come up with the initial figure, which includes some variables related to driving history.

Theoretical example: Sue is historically light on the accelerator and lives in a warm climate. Her display shows 310miles range at 100%, which decreases while driving at the fixed 230wh/mi EPA rate. John likes to abuse the accelerator a little and lives where the weather is brisk. His display shows 280miles range at 100%, and decreases while driving at the fixed 230wh/mi EPA rate. Neither has abnormal battery degradation, but their 100% range difference is a result of some variables Tesla has started using to make the battery meter range more realistic.

Something obviously changed after V10 when all this complaining about rated range started. It is highly unlikely that everybody's battery is now experiencing excessive degradation that started with a software update. I would be very careful stating something as fact just because Tesla has historically calculated metrics one way. They are free to, and do, change just about everything with their OTA updates.

Thanks for this---but one point--My range has been plummeting since February...waaaay before V10....I'm now down to 266 miles at 100%!

 

[AlanSubie4Life]

Thanks for this---but one point--My range has been plummeting since February...waaaay before V10....I'm now down to 266 miles at 100%!

Definitely grab some pictures of the trip meter AND the battery gauge (distance mode) with distance set to km and before and after photos for a decent length trip segment, when you get a chance. Sounds pretty rough for you. Just want to see how it is behaving.

TeslaFi data time cross check of course too (but that data I do not trust).