Wiki Sudden Loss Of Range With 2019.16.x Software

[sorka]

For reference, I finally updated the software on my early-2013 Model S 85 from v8 (they managed to fix enough of USB music for it to be an improvement and not a regression). I saw no change in my range level at all. One more data point for "some cars aren't affected at all".

Whatever Tesla has done in the software this time, it's intelligent -- it isn't blind ignorant caution like some of the previous changes. (Like when they keep kicking my charging rate down from 40 amps to 30 amps because their program to prevent overheating wiring doesn't like the voltage of my electrical supply. They're just wrong there. My electrical wiring is top notch overkill, heavier duty than needed, and their algorithm there is junk; the dumb algorithm just can't handle that the mains supply here tends to run a high voltage and have unsteady frequency. Or the "counter based" limits on some people's 90 pack Supercharging, which are just guesswork and not based on actual battery state. Or the "counter based" limits on top acceleration, which are again guesswork-based.)

In this case, unlike previous cases, they're clearly detecting something specific which actually happened in particular batteries and not in others, and reacting to that only when they find it in the battery. It would be interesting to know the technical details.

Because it seems like an intelligent response rather than a blind excess-of-caution response, people probably should do the update; if your range drops, your car probably *was* at risk of catching on fire.

...at which point they should replace the battery if it's still under warranty.

 

[sorka]

Hi sorka. I am genuinely confused. Here you are arguing for freedom and that Tesla is taking away your freedom by protecting your battery from further degradation. And yet I thought that in some other posts you were advocating that Tesla should do warranty replacements (or recall?) and give lots of people free batteries. Or maybe that was other posters who advocated for free replacement batteries? I am really not sure and just trying to follow this 43 page thread and learn from it.

I'm arguing for keeping what I PAID FOR

 

[PaulusdB]

For reference, I finally updated the software on my early-2013 Model S 85 from v8 (they managed to fix enough of USB music for it to be an improvement and not a regression). I saw no change in my range level at all. One more data point for "some cars aren't affected at all".

Whatever Tesla has done in the software this time, it's intelligent -- it isn't blind ignorant caution like some of the previous changes. (Like when they keep kicking my charging rate down from 40 amps to 30 amps because their program to prevent overheating wiring doesn't like the voltage of my electrical supply. They're just wrong there. My electrical wiring is top notch overkill, heavier duty than needed, and their algorithm there is junk; the dumb algorithm just can't handle that the mains supply here tends to run a high voltage and have unsteady frequency. Or the "counter based" limits on some people's 90 pack Supercharging, which are just guesswork and not based on actual battery state. Or the "counter based" limits on top acceleration, which are again guesswork-based.)

In this case, unlike previous cases, they're clearly detecting something specific which actually happened in particular batteries and not in others, and reacting to that only when they find it in the battery. It would be interesting to know the technical details.

Because it seems like an intelligent response rather than a blind excess-of-caution response, people probably should do the update; if your range drops, your car probably *was* at risk of catching on fire.

Did you 'range charge' your car? More specifically charge to more than 93% SoC?
It is not impossible that the (protective) range drop only materializes after the rebalancing process triggered by SoC > 93%.

 

[faughtz]

... The 85 has nothing to do with anything except for the model designation / marketing.

...

I have to disagree: my Monroney sticker lists an 85kWh battery.

 

[sorka]

If you were Tesla, what would you do about the fire danger associated with batteries that are showing indications of some Li plating?

I'd replace the SMALL number of batteries showing the condition rather than hiding it with a bandaid and pissing off customers making it almost certain they will NEVER BUY ANOTHER TESLA

Also, keep in mind, the only official answer given to anyone by Tesla about the update that reduced range is the one they gave Electrek in which they said it was to prolong battery life.

There's zero evidence that the BMS update they talked about in the press release last week to reduce chances of combustion that they are rolling out now had anything to do with each other.

 

[sorka]

Replace the battery!

Guess I should have kept reading before I responded with the same thing

 

[St Charles]

I'd replace the SMALL number of batteries showing the condition rather than hiding it with a bandaid and pissing off customers making it almost certain they will NEVER BUY ANOTHER TESLA

Also, keep in mind, the only official answer given to anyone by Tesla about the update that reduced range is the one they gave Electrek in which they said it was to prolong battery life.

There's zero evidence that the BMS update they talked about in the press release last week to reduce chances of combustion that they are rolling out now had anything to do with each other.

Guess I should have kept reading before I responded with the same thing

No worries, we are on the same wavelength. Tesla NEEDS to stand by their product. This new bullshit of avoiding warranty work needs to go!

 

[sorka]

Pfff, updates used to be fun and something I looked forward to installing, now I'm terrified it will nerf my car. Make it right, Tesla.

Already there. I've lost app access because my Tesla connects to a null hotspot when parked to isolate it from the internet. So far it appears downloads do not occur while driving. I have no new staged software in weeks. First time ever since owning it.

 

[jaitch]

..................

• 90D/P90D – ~85.8 kWh total capacity, 81.8 kWh usable
• ............

Tesla’s hacked Battery Management System exposes the real usable capacity of its battery packs - Electrek

Are these P90D figures still considered to be correct? I have searched and can't find the answer. I recently bought a CPO P90DL, 2016, 34k miles, V1 battery. The figures Scan My Tesla gives me are:-

Nominal full pack: 77.2
Useable full pack: 73.2
Energy buffer: 4

Should I be concerned that my battery capacity been similarly reduced by Tesla at some point in its life? My 90 % voltages are 4.09 and 3.89 v at 69 %. Amps are 1525 ish max. Which seem "normal" from what I can tell.

So I guess I have "normal" V1 90 pack degradation at just over 10%?

Thanks

James

 

[neroden]

They responded to my service request for the range loss issue:

- Asked me if my issue has already been resolved (!!!).

Tesla communications stink.

- I responded negative and also provided the electrek article link.
- They replied and said the article has all the information they can give me at this time.
- They also said they have reviewed my car's data to ensure there are no other issues that are occurring with my battery or the charging system and found that everything is working as it should.
- They said there were a number of cars impacted by the update and that they are working on updates that will address the issue for those that were impacted but they do not have a timeframe on when those updates will be deployed.
- They said they know this is a concern for those that were impacted and hopefully changes will be coming in the very near future.

So, I'm going to make an educated guess regarding what the "improvements" will be for those impacted. It seems to me that, with the new software, when lithium plating is detected somewhere in the charging process, restrictions are applied to the entire pack to prevent overcharging of plated cells.

In fact, with better use of data, it should be possible to figure out which modules, or even which groups, have plating, and only restrict *those modules*. The Shanghai fire apparently resulted from only one module. This should restore the capacity of any unaffected modules.

That's my guess. Just a guess.

 

[CuriousG]

I don't disagree with the statement in a vacuum. I do disagree with the logic leading up to the above statement. Tesla made a change to the vehicle, post production, to reduce the chances of catastrophic failure. This change caused a loss in functionality. Furthermore, Tesla never released any guidance to allow owners to avoid this condition. This alteration was pushed to owners 'overnight' rather than slowly over time. This is also an indication that Tesla themselves were not aware of this condition. All of which points to a failure which must be corrected.

Limit batteries which are failing, until they can be replaced. They MUST be replaced!

There is also no unilateral statement on what proper charging procedures are. Elon will say one thing. The technicians will say another. Then they have service managers, support people, PR people. Look at all the different ways there are on how to close the frunk properly without denting the hood?

 

[sorka]

It might prevent your car from catching on fire and burning to a crisp.

..[snip]...

Those of you whose range has dropped -- I understand why you're frustrated. But you don't want your cars to catch fire, I assume.

Totally agree with all of that. If they detect your battery is in a dangerous state and could catch fire they totally should take whatever measures are necessary to prevent that via software until they can get the SMALL number of batteries effected in for replacement. Samsung did the same thing with the Note 7 until the customers who had them could be made 100% whole.

 

[sorka]

Naive question: On a trip where the range between two superchargers works out to 75%:
is it better to charge at the first supercharger to 95% and arrive at the next one with 20%,
or to charge at the first one to 90% and arrive at the next one with 15%?
(Please feel free to include considerations associated with ambient temperature in your response.)

Secret option C...charge to 85% and arrive with 10% would be even better.

...further more that option results in less total time spent at the supercharger. As you slide the range window up, it takes longer to charge that range.

I personally alway shoot for about a 7% buffer to prevent charging higher as much as possible in good weather. If it's crappy rainy weather, then I increase that buffer. It's 50 MPH head wind, all bets are off and I'll shoot for as much buffer as I can. I always check the current wind and wind forecast. It can have a huge impact.

 

[neroden]

P.S. Thanks to @IngTH for the technical pointers which led me to understand what is probably going on.

P.P.S. Brickbats to Tesla for their appalling, confusing, and inaccurate communications. Here's what I would have said:

"We have determined that a certain type of normal degradation, lithium plating, was not being detected properly by the Battery Management System. In rare cases where the degradation was substantial, this could create a dangerous condition where the battery could be overcharged and catch fire. We believe this was the cause of the fire in Shanghai. We have fixed this with a software update to the Battery Management System.

Some owners will discover that their reported range is significantly lower than it was before; this degradation already happened over the preceding years but was not being reported properly. Although we did not warrant any particular range for Model S or X, if your range is less than 70% of the original rated range and your car is under warranty, we will replace the battery, as we do with Model 3.

Unfortunately we will have to make this update mandatory, so we are trying to fix all the other reported bugs in the software as fast as we can. Please report any regressions (things which worked in older versions and broke in newer versions) to [email protected] and we will work on them as fast as possible."

This isn't anywhere close to what Tesla did, not by a long chalk.

 

[jerry33]

Naive question: On a trip where the range between two superchargers works out to 75%:
is it better to charge at the first supercharger to 95% and arrive at the next one with 20%,
or to charge at the first one to 90% and arrive at the next one with 15%?
(Please feel free to include considerations associated with ambient temperature in your response.)

Personally, I don't think it matters. Best idea is to charge just enough to make it to the next stop, plus a bit extra just in case.

 

[sorka]

FYI, nobody consents to OTA updates at purchase (it's not even clear that would be legal).
Late (post-2016 IIRC) purchasers consent to OTA updates as a condition for getting warranty repairs. Earlier purchasers did not.

Later versions of the warranty agreement state that you must agree to updates to maintain your warranty. It's unlikely that this can actually be enforced. My warranty agreement was early enough that it doesn't have that language.

 

[sorka]

Great post neroden.
When did the change(s) away from all-graphite anodes occur? Is the problem with Li plating pretty much gone with the new anodes, or is it still a good idea to limit supercharging to high SoC in cold weather?
Can you get Li plating charging charging to just 85% or less at a supercharger?

First use of silicon in the Anode was V1 of the 90. They went too far with it though and it caused premature anode fracturing. Silicon increases capacity by allowing the anodes to absorb more ions. A single silicon atom can accept 4 lithium ions while 6 carbon atoms can only take 1 ion. A small amount of silicon makes a huge difference. But the more you ad, the larger the anode will get physically. If you add too much, the anode will fracture over larger changes in SOC.

As a long time RC hobbiest, you learn to pick the cell that meets your needs. Higher energy density cells with more silicon hold more energy but their lifetime is shorter because their anodes fracture sooner.

 

[sorka]

Yes, of course, but I’m speaking of “best” performance which would be at cell voltages near 4.2v.

Yup, your best performance is gone

 

[sorka]

I have to disagree: my Monroney sticker lists an 85kWh battery.

Good point!

 

[St Charles]

First use of silicon in the Anode was V1 of the 90. They went too far with it though and it caused premature anode fracturing. Silicon increases capacity by allowing the anodes to absorb more ions. A single silicon atom can accept 4 lithium ions while 6 carbon atoms can only take 1 ion. A small amount of silicon makes a huge difference. But the more you ad, the larger the anode will get physically. If you add too much, the anode will fracture over larger changes in SOC.

As a long time RC hobbiest, you learn to pick the cell that meets your needs. Higher energy density cells with more silicon hold more energy but their lifetime is shorter because their anodes fracture sooner.

Can this be countered by limiting the SOC range the battery experiences?