Wiki Sudden Loss Of Range With 2019.16.x Software

[Guy V]

There is always a spread between the behaviour of the different batteries (and some of it depends on environmental factors and charging history).

No manufacturer will guarantee that everyone gets above average batteries (they do sell outside of the Lake Wobegon area).

"Defective" is "outside of the normal range", but customers cannot at will define what range they would like to see means a battery is seen as "defective".

For model 3 owners, Tesla has been more careful to define what is definitely abnormal, and the bar is awfully low: 70% at the end of the warranty period, which is 8 years or 192000km, whichever comes first. They might treat customers skating close to that limit as "abnormal" once they gathered fleet data that allows them to more accurately see what is normal (and replace batteries when a user is seen as being in the 5% of unluckiest users), but they haven't put that stick in the ground.

Also, charging limits imposed when someone is worried about a hitherto unforeseen condtion are usually precautionary. Just hypotethically, suppose they see a rise in the risk of fire for some batteries, which could have lethal effects, I would rather want them use an abundance of caution in setting limits until they gathered more data, and only then attempt to restore more range to more customers. If someone dies, then that's not something you can fix by installing some new firmware.

That's all very rational, but unfortunately Tesla hasn't told at all what it going on with the batteries and what they are doing about.

 

[sorka]

I wanted to add that I was pushed a software update last night. Version is 2019.28.2 320Fba0. I didn't think anything of it until reading this thread. The car is plugged in now (took the Boxster today) and the recent 215-219 90% charge is now showing 228 rated mile range. They must have "unlocked" the cells with the new version.

What were you getting before this update?

 

[sorka]

At this point I don't think anyone is going to convince anyone else as to the cause of this drop in range. I'm curious what those affected think would be an effective resolution.

Software update that prevents the pack from charging to 4.2 volts. I thought this was the axiom of the thread and didn't realize it was being debated. What's being debated is why Tesla did this and nobody really knows because Tesla hasn't formally said.

 

[DJRas]

That they will make the reduction in range "less severe, serious, or painful." (Saying it would be reversed is saying that it would be 100% undone.)

So maybe they will reduce the range cap by 50%, 75, or 90%. But WK057 thought they would have to do something, but that it wouldn't be significant.

so typical Tesla... vague promises.
If I get 25% back then the suit goes on for sure.
If I (and all the others so affected) get 90% back then they need to address the reasons and safety concerns.

 

[sorka]

So if it turns out that Tesla is actually pushing a new firmware version that fixes the range issue, how quickly will you be dropping your class action lawsuit?

At this point I doubt the suit would just be dropped if Tesla just simply undid everything. There would still be damages in the form

so typical Tesla... vague promises.
If I get 25% back then the suit goes on for sure.
If I (and all the others so affected) get 90% back then they need to address the reasons and safety concerns.

Tesla broke 691 - 463 promises to me when I bought mine

 

[Evoforce]

You are assuming facts that are not in evidence.

That is part of the problem. It is left to speculation and no facts were forthcoming.

 

[Evoforce]

I wanted to add that I was pushed a software update last night. Version is 2019.28.2 320Fba0. I didn't think anything of it until reading this thread. The car is plugged in now (took the Boxster today) and the recent 215-219 90% charge is now showing 228 rated mile range. They must have "unlocked" the cells with the new version.

Didn't happen that way for mine... I installed that same update and it is still at 216 @90%

 

[Ferrycraigs]

Of course your service center didn't as there is nothing that they can do. Well technically I guess they can downgrade your firmware to an old version, but if they do that they will be fired. (Tesla has a strict no roll-back of firmware version policy unless approved, and probably performed, by engineering.)

For firmware all they are allowed to do is push the latest firmware that has been assigned to your car by engineering.

Yeah, but they are the first port of call. I go to them if I have a problem. If they can fix it, they do. If they don’t have the authority, I presume they ask. If there is a no roll back policy, what’s wrong with asking to just revert the Vmax setting. Who does the uncorking when you pay for it, and how is it done?

 

[Evoforce]

Even when I received my Tesla new in 2017, I was embarking in unknown territory because the history of the cars and batteries have not been proven over many decades. But rather than waiting decades to find out, I CHOSE to buy one then.

If/when Tesla finds issues and needs to make adjustments, then by all means, do it.

The part where Tesla fails often is their communication and does need improvement. However, I think this lawsuit is ridiculous as it really just rewards the lawyers. Tesla will lose millions on defending themselves with lawyers, lost sales AND the bad PR for which places like CNBC are so eagerly waiting to dish out.

Sometimes lessons come at a price... Good communication to your product owners is a must at all levels. Tesla is struggling with this or it is not a priority. Maybe this lesson will help improve that?

 

[VT_EE]

@KarenRei, I love this post. Well done!

You could have saved me lots of time and grief to post this one much earlier. I'm becoming a fan now, seriously

KarenREI is a force to be reckoned with and knows her stuff. I used to enjoy her informative (and slightly combative) posts while details of the Model 3 packs were dribbling out. Debate battery tech with her at your own risk.

 

[VT_EE]

Sometimes lessons come at a price... Good communication to your product owners is a must at all levels. Tesla is struggling with this or it is not a priority. Maybe this lesson will help improve that?

One can only hope.

 

[Guy V]

As a side note - and this is more out of curiosity - how is this actually affecting people's daily lives? I'm seeing range drops reported like 10-15% (for example, David Ramussen: 247->217mi). Were you guys actually charging to 100% daily before the drop?

• If so, then your batteries should be degraded even if there were nothing wrong with the pack, and it'd be your own fault. Everyone knows (or at least should know) that you're not supposed to charge to 100% daily unless you have absolutely no other choice (wherein even "slow" degradation will eventually ruin things for you).
• If not - if you were being responsible and charging only to 80-90% daily - then just change your daily charge to 95-100% and you've got your daily range back, and your level of stressing your batteries is exactly the same as it was before the update, because you're charging to the exact same voltage. Only long trips are affected.

Am I missing something? Is this just about trips?

It takes twice as long for one. It precludes charging to previous range for longer trips. Between the range limitation and the charging speed reduction it very significantly affects road trips. It may not be a devastating loss for most but neither is it insignificant and instills dread at what may come next.

 

[Chaserr]

Now on 2019.18.2

Can you confirm you were sent 2019.18 yesterday? I checked teslafi when you posted and there were none, now there is 1 car running [2018.18.2 301aeee]. If that's you, Tesla sent you a firmware nobody else got from months ago, implying they fixed this issue in May but held the update back.

As a side note - and this is more out of curiosity - how is this actually affecting people's daily lives?

Reduced voltage has slowed my 0-60 I use at every stop. The Tesla grin isn't there any more and I want it back.

 

[Droschke]

The key point is that all li-ion batteries, even when brand new, are:

• At risk of self-combustion if charged beyond certain voltage / current limits (which vary greatly based on conditions and the how the various aging processes have affected the cells)
• At risk of unacceptable levels of degradation if charged beyond certain voltage / current limits (same as above)

Which is why you always have limits coded into the BMS. Again, that's the BMS's job.

A key limiting factor in charging is anode intercalation rates. Intercalation of lithium ions into the anode is energetically preferable to plating lithium; they "want" to intercalate rather than plate out. But the rate at which intercalation can proceed varies depending on the conditions - and each of these can be seen in how the BMS manages charging.

• Ion mobility, and thus intercalation rates, are strongly dependent on temperature. In practice, you see this as very slow charging until the battery warms up.
• Intercalation slows down the more lithium there is already intercalated in the anode. The BMS responds with what you see as taper.
• Above a certain point, intercalation has slowed to such an extent that you can A) no longer reliably ensure no lithium plating, and or B) the high reactivity of the anode makes for an unacceptable level of reaction with the electrolyte, and thus degradation. You see this as the cutoff.

It's long been clear that the packs in question were not living up to the desired longevity standards.

• First off, we got the nerfed Supercharging after a fixed number of sessions in 2017. This clearly spells out Tesla's concern: after a given number of sessions, they can no longer trust that intercalation is proceeding fast enough for the full ~120kW charging. Ion mobility is suffering as the packs age and in particular as they are supercharged - hence the limits which quite clearly spell out the problem.
• Now we get this - a voltage limit, and possibly a further charging limit. A voltage limit at the upper end would more commonly be interpreted as protecting the battery from excessive anode-electrolyte reactions - and that might well be. But combined with the charge rate (and thus intercalation) limits, it looks more like they cannot trust the reliability of intercalation at the upper end of the packs. The problem could also be both.

Why didn't Tesla's accelerated aging tests detect this problem? It's hard to say; accelerated aging is a very difficult thing to do reliably (it's even possible that there were QC issues on the anode powder that were out of Tesla's control). I'd love to know what went wrong, but I doubt we ever will. But acceptable intercalation rates have clearly fallen far faster than Tesla expected them to vs. how these packs were testing when new. As they age, the anodes just aren't soaking up the lithium like they're supposed to - at least not reliably.

Can you force them to charge faster? Of course you can. Just like you can try to force new cells with new chemistries to charge much faster and to maximum theoretical voltages than they already do. But it's the BMS's job to stop you from doing what's possible, and instead limit you to what's reasonable. Sure, it "works" - but you're taking away 9s in that reliability game. Whether the issue is safety or longevity, it's the BMS's job to stop you from hurting yourself. The more a manufacturer learns, the more they change the BMS's algorithm. If the news you learn is good, charge rates, tapers, max voltages etc improve. If the news is bad.... well, this happens. And it sucks for those affected.

But this is the BMS's job. And it is absolutely right to update the BMS based on the information that you collect over time as to how packs with that chemistry are aging - for good or for bad.

To sum up:

• Packs are always limited. From the very day you pick up the car. That's the BMS's job.
• The limits in the BMS should be updated based on the latest information - good or bad.
• The ability of lithium to be reliably intercalated into the anodes clearly has degraded faster than the company anticipated - hence the chain of downgrades over time.
• Nothing in the warranty guarantees a minimum range
• People absolutely have a right to be upset about how much the range is dropping
• Tesla should try to do something for them, without setting a trend of "non-warranty-covered fixes will occur at Tesla's cost any time people are disappointed with something". What would be appropriate is reasonably up for debate.
  • Rolling back the changes and pretending that anode degradation isn't happening, however, is not a reasonable choice.
• Tesla failing to communicate when there's bad news they have to deal with is understandable (who wants to broadcast negative PR when you can just handle the issue silently?), but incredibly frustrating for affected individuals.

That's just my two krónur.

Impressive writing with perfect English (you say you are in Iceland).

Two questions for you:

- Can you share your battery credentials (looks like good scientific knowledge) and/or Tesla affiliation (looks like you have inside info)?

- If you have read this thread early on, there were indications of Tesla (by testing) was looking into condition X and found condition Z surprisingly in the suspected packs. Would you elaborate what these two conditions can be?

Thank you.

 

[Guy V]

At this point I don't think anyone is going to convince anyone else as to the cause of this drop in range. I'm curious what those affected think would be an effective resolution.

For e.g. if 2019.28.2.5 increases your range back to where it was before 2019.16, would that be enough? Or would you also want service center battery pack inspections for peace of mind?

I believe Tesla could convince most of us with a strong dose of transparency. They have taken away confidence in what we can expect from our cars. They really need to restore that or they will suffer seriously in the marketplace.

 

[Evoforce]

In my case, I have a loaner battery that has not been replaced. Original battery had the contactor issue and was issued a loaner. They never have swapped the loaner out. I don't want their loaner z or x problem battery. Give me a better battery.

 

[DJRas]

Can you confirm you were sent 2019.18 yesterday? I checked teslafi when you posted and there were none, now there is 1 car running [2018.18.2 301aeee]. If that's you, Tesla sent you a firmware nobody else got from months ago, implying they fixed this issue in May but held the update back.

Reduced voltage has slowed my 0-60 I use at every stop. The Tesla grin isn't there any more and I want it back.

Yes. My earlier post was in error. 2019.28.2

 

[Guy V]

Even when I received my Tesla new in 2017, I was embarking in unknown territory because the history of the cars and batteries have not been proven over many decades. But rather than waiting decades to find out, I CHOSE to buy one then.

If/when Tesla finds issues and needs to make adjustments, then by all means, do it.

The part where Tesla fails often is their communication and does need improvement. However, I think this lawsuit is ridiculous as it really just rewards the lawyers. Tesla will lose millions on defending themselves with lawyers, lost sales AND the bad PR for which places like CNBC are so eagerly waiting to dish out.

That would be the result of their lack of transparency in responding to the issue. If you hear nothing from Tesla in months about what is going on other than claims there is nothing wrong with the batteries and this is normal degradation what else would you suggest to find out?

 

[MP3Mike]

If there is a no roll back policy, what’s wrong with asking to just revert the Vmax setting. Who does the uncorking when you pay for it, and how is it done?

Nothing wrong with asking, but it isn't going to get you anywhere as that isn't a parameter that is exposed for a SC technician to change.

Uncorking was releasing performance, i.e. faster 0-60 times, by modifying parameters that engineering had exposed to technicians. If you are talking about the unlock/upgrade from a Model S 60 to a Model S 75, for the ones that came with a larger pack, that was done via an option controlled on the Tesla servers. (You could unlock the extra capacity from a button on the MCU, from the Tesla web site, or by visiting a service center.)

Service center personnel aren't going to have any capability to modify the BMS parameters.

 

[Guy V]

That they will make the reduction in range "less severe, serious, or painful." (Saying it would be reversed is saying that it would be 100% undone.)

So maybe they will reduce the range cap by 50%, 75, or 90%. But WK057 thought they would have to do something, but that it wouldn't be significant.

If you want to quibble about meanings, he didn't say fully reversed. Reversing it to any extent is reversing it. I certainly will be interested in finding out how much and how soon.