Model 3 Performance Battery Degradation One Month (Story)

[bedoig]

@AlanSubie4Life

I'm still trying to figure out exactly how TeslaFi is calculating this. If we believe rated miles are based on a fixed constant (which I do), why wouldn't they just take total rated miles used multiplied by an established (or user definable) constant for each vehicle? I'm going to keep digging around a little, just curious. I did find a setting that allows you to scale the kWh calculation. I used a factor to make my last trip equal actual consumption. Now I'm interested in whether that factor holds fairly true over time or not.

The CAN data I quoted was Scan My Tesla's Trip Energy display, which I believe is based on cumulative (lifetime) counters of discharge and regen reported by the BMS. The values of the counters from the start of the trip are compared to the values at the end of the trip, so you don't have to worry about data dropouts, etc.

 

[AlanSubie4Life]

If we believe rated miles are based on a fixed constant (which I do), why wouldn't they just take total rated miles used multiplied by an established (or user definable) constant for each vehicle?

I think part of the issue is that there is some small variability on the discharge "constant" depending on discharge rate and other factors. It's not large (if they did do what you said it would be more accurate than what they have), but it is there, apparently (I've personally never seen it). If done properly, it is a bit more accurate, probably, to just scale the usage (based on the rated mile use) according to what they think the full battery energy capacity was - but that value for full battery capacity does have to be accurate. In the end it's likely to be imperfect compared to the trip meter no matter what you do, since that direct access to "coulomb counting" is not available to the API. I'd expect there will always be some deviation of even your new scaled (with the new setting you found) TeslaFi value from the trip meter, depending on the exact scenario.

The CAN data I quoted was Scan My Tesla's Trip Energy display, which I believe is based on cumulative (lifetime) counters of discharge and regen reported by the BMS. The values of the counters from the start of the trip are compared to the values at the end of the trip, so you don't have to worry about data dropouts, etc.

That's not as direct as I'd like - because Scan My Tesla COULD scale that data to make it match the trip meter. Not saying they do, but this makes me doubt the data. I'd prefer to see a direct reading of the initial kWh and final kWh (along with the "fullkWh" value just for reference) from the CAN for a given segment and just do the math myself. Then finally I might get a definitive answer to how the BMS kWh are scaled relative to the charging and discharging kWh. (We know charging and discharging kWh are scaled differently.) It's also possible the discharge and regen integrated values on the CAN bus are just not scaled the same way as the "kWh" values on the CAN bus, by Tesla - the sum of these discharge/regen values may actually be the values directly used by Tesla to display on the trip meter, so by definition they would match...but that would not answer my question about whether the kWh in the battery match these other "types" of kWh. In other words, if you use 10kWh on the trip meter, by how many kWh does the CAN bus kWh available (the various BMS kWh that you can read back) reduce? No one knows! (I've never seen this question answered.)

 

[TimothyHW3]

So I got a call from Tesla just now (in anticipation of the service visit next week), and they said that it's all my fault. LOL They say that over the past month, I've been charging from 50% to 80% nightly and that this is messing up the calculated range, and instead, I should charge from 30-40% to 90% for at least a month and that should correct the problem.

I find this explanation, while compelling, not entirely credible. I am sure I am not the only one that charges their car from 50-80% nightly (or thereabouts),

Well, you are not alone, which is evident by all the posts we get in the forum...People will look at updates and other things for faults, but not at the most obvious - their charging. The Tesla person was even conservative - I would've advised to go at least two times from 10%-90%. Where 10% you go from 30% or something down to 10% and charge right away and 90% you charge and drive right away.

Me and at least 2 other people told you that the issue is with the BMS and the way you charge. And we don't even work for Tesla...

 

[AlanSubie4Life]

Well, you are not alone, which is evident by all the posts we get in the forum...People will look at updates and other things for faults, but not at the most obvious - their charging. The Tesla person was even conservative - I would've advised to go at least two times from 10%-90%. Where 10% you go from 30% or something down to 10% and charge right away and 90% you charge and drive right away.

Me and at least 2 other people told you that the issue is with the BMS and the way you charge. And we don't even work for Tesla...

I guess we're going to find out here! My guess is that's going to do approximately nothing. Might change rated miles available (extrapolated to 100%) by 5-10 miles (which is not really a fix in this case). I guess we will see! @gaspi101 do report back, and if possible take those pre/post pictures of each discharge segment.

My guess is the whole experiment will get disturbed by the next software update though, and that will invalidate any results.

 

[TimothyHW3]

Like I said - ideally he will start doing this after a CAC reset. Not sure how long it will take for the BMS to recalibrate without it. It certainly will not happen overnight - it might take months of constant 10-100% charging, no idea.

Plus, he has been extrapolating range from TrslaFi this whole time even though we told him multiple times this is not correct.

I think this is yet another indication that the problem most people face here is with the way they charge and not their batteries. I think Tesla should definitely rethink this always plug in rule, because people end up just topping up 60%-90% or 50%-80%, which once again is bad for the battery and the BMS.

It seems the BMS on Model 3 works differently. Not sure if Tesla is just figuring things out as they go with these new batteries or if they just have uninformed staff that is not educating people enough how to charge.

 

[gaspi101]

Well, you are not alone, which is evident by all the posts we get in the forum...People will look at updates and other things for faults, but not at the most obvious - their charging. The Tesla person was even conservative - I would've advised to go at least two times from 10%-90%. Where 10% you go from 30% or something down to 10% and charge right away and 90% you charge and drive right away.

Me and at least 2 other people told you that the issue is with the BMS and the way you charge. And we don't even work for Tesla...

I have tried (1) charging from 30% to 90% for two months; (2) taken several road trips where I charged to 100%, then immediately discharged to nearly zero, road trips of varying length; (3) I have charged to 80%, 85% and 90%, for extended periods in each, and both nightly and every 2-3 days (again, 30-90(or 80 or 85). I have done calibrations, I have done balancing, I have done a CAC reset. I have been at this for 8 months. Please don’t patronize me.

 

[gaspi101]

I guess we're going to find out here! My guess is that's going to do approximately nothing. Might change rated miles available (extrapolated to 100%) by 5-10 miles (which is not really a fix in this case). I guess we will see! @gaspi101 do report back, and if possible take those pre/post pictures of each discharge segment.

My guess is the whole experiment will get disturbed by the next software update though, and that will invalidate any results.

You got it!

 

[diamond.g]

You got it!

Do you have a 2019 3 and use Android device? If so would you be willing to get an OBD2 reader and Scan My Tesla to pull pack data from canbus?

 

[gaspi101]

Do you have a 2019 3 and use Android device? If so would you be willing to get an OBD2 reader and Scan My Tesla to pull pack data from canbus?

2018 on an iPhone...sorry!

 

[TimothyHW3]

I , I have done a CAC reset. I have been at this for 8 months. Please don’t patronize me.

Yes, and six weeks after the CAC reset the car was fine until you started to charge, consistently as evident by the Tesla person, from 50-90% or 70-90%.

And all the "data" you have is just extrapolated TeslaFi data from 50%-80%, which is BS at best, because of all the things already mentioned here.

Do whatever you like...

 

[TimothyHW3]

Do you have a 2019 3 and use Android device? If so would you be willing to get an OBD2 reader and Scan My Tesla to pull pack data from canbus?

I don't think that will help. The CAN just reads what the BMS thinks the capacity is. I think if he had to read the data my estimate is he will get about high 60kWh, around 67-69kWh. This is just what the BMS thinks his capacity is, but in reality is he is probably at the mid 70s evident by the reset.

We have a car in the data log with the same charging pattern as his that has tanked abruptly at 5000km and never recovered. However, as here, I am not quite sure of his exact driving profile and how often did he consistently tried to rectify it. I don't know if you can fix it with just 0-100% charging - maybe a reset is needed in all such cases and then consistent charging from 20-90%

Judging from all the info we have, that iLike Tesla video and all the reports we get here, it seems that the BMS on Model 3 works differently and doesn't really like these charging behaviors from 50-90 or 70-90% which most, mostly American drivers, do.

It is high time Tesla make a FAQ somewhere and address this as their calls and service appointments will only rise.

 

[diamond.g]

I don't think that will help. The CAN just reads what the BMS thinks the capacity is. I think if he had to read the data my estimate is he will get about high 60kWh, around 67-69kWh. This is just what the BMS thinks his capacity is, but in reality is he is probably at the mid 70s evident by the reset.

We have a car in the data log with the same charging pattern as his that has tanked abruptly at 5000km and never recovered. However, as here, I am not quite sure of his exact driving profile and how often did he consistently tried to rectify it. I don't know if you can fix it with just 0-100% charging - maybe a reset is needed in all such cases and then consistent charging from 20-90%

Judging from all the info we have, that iLike Tesla video and all the reports we get here, it seems that the BMS on Model 3 works differently and doesn't really like these charging behaviors from 50-90 or 70-90% which most, mostly American drivers, do.

It is high time Tesla make a FAQ somewhere and address this as their calls and service appointments will only rise.

Maybe Tesla should just make it into a GOM, which seems like it would do a better job of hiding degradation. Then they can make the energy graph actually break out what is using the power.

 

[gaspi101]

Yes, and six weeks after the CAC reset the car was fine until you started to charge, consistently as evident by the Tesla person, from 50-90% or 70-90%.

And all the "data" you have is just extrapolated TeslaFi data from 50%-80%, which is BS at best, because of all the things already mentioned here.

Do whatever you like...

during the first two weeks after the CAC reset I was instructed to not let it get below 40% because that could potentially damage the CAC system, and potentially brick the battery, and so, for those first two weeks, I charged 70–90 maybe nightly, and about a month after, deeper discharge cycles (40-85, usually), and only then did the range start to plummet again. It was only in the last month or two that I gave up worrying about it and just did smaller discharges to save the battery, as charging habit changes seemed to not be the cause. And then a reddit post brought me back to this issue and here I am. Again, you don’t know my charging habits throughout the last 8 months. I’ve been varying them for long periods at a time, testing. your snarky and uninformed judgmental responses are neither useful nor appreciated.

 

[gaspi101]

Maybe Tesla should just make it into a GOM, which seems like it would do a better job of hiding degradation. Then they can make the energy graph actually break out what is using the power.

That’s interesting, what’s a GOM?

 

[KenC]

That’s interesting, what’s a GOM?

GuessOMeter.

It's what most other EV mfrs use. It's just range based upon recent driving, so in a Model 3, it would show the 30mile AVG figure from your energy consumption chart.

 

[diamond.g]

That’s interesting, what’s a GOM?

GuessOMeter.

It's what most other EV mfrs use. It's just range based upon recent driving, so in a Model 3, it would show the 30mile AVG figure from your energy consumption chart.

Yeah what @KenC says. It hides the degradation better since it is based on how you drive instead of a fixed constant.

 

[gaspi101]

Yeah what @KenC says. It hides the degradation better since it is based on how you drive instead of a fixed constant.

Clever. But given that there’s a warranty claim tied to a specific degradation sum, I’d prefer some method to keep an eye on degradation...but thanks for the info. Learned something new.

 

[diamond.g]

Clever. But given that there’s a warranty claim tied to a specific degradation sum, I’d prefer some method to keep an eye on degradation...but thanks for the info. Learned something new.

Yeah. I guess you would have to ask a Bolt owner how they are able to track battery life.

 

[TimothyHW3]

during the first two weeks after the CAC reset I was instructed to not let it get below 40% because that could potentially damage the CAC system, and potentially brick the battery, and so, for those first two weeks, I charged 70–90 maybe nightly, and about a month after, deeper discharge cycles (40-85, usually), and only then did the range start to plummet again. It was only in the last month or two that I gave up worrying about it and just did smaller discharges to save the battery, as charging habit changes seemed to not be the cause. And then a reddit post brought me back to this issue and here I am. Again, you don’t know my charging habits throughout the last 8 months. I’ve been varying them for long periods at a time, testing. your snarky and uninformed judgmental responses are neither useful nor appreciated.

Nothing is snarky and my comments are pretty informed.

The range never "tanked" because you never actually used it nor tested it. Mostly extrapolated data from TeslaFi.
As ttowhy the BMS started being awkward after a while and not immediately - this is simple, it takes time for the BMS to screw itself...

 

[TimothyHW3]

Maybe Tesla should just make it into a GOM, which seems like it would do a better job of hiding degradation. Then they can make the energy graph actually break out what is using the power.

I like how Tesla does it. It gives you some indication of degradation or BMS misscalibration.

The only thing that kind of bugs me is their marketing trick to use the buffer in the rated range when charging and hide it below 0% when discharging.
But once you know it, then you are fine.

Also, the way Tesla predicts the energy graph, based on elevation and speed, this is top notch stuff. No other EV comes even close to what Tesla does.

If only they had better maps matterial and their BMS could factor those 70-90% charges. Or explain to drivers why this charging is wrong,O but to thenthe battery and BMS.

Energy graph breakage is nice - but once again, once you know how your HVAC works and what power it draws at what settings(easily tested by plugging in and stopping the charge) you don't need it.

Other than that they have the best range estimates.