Wiki Sudden Loss Of Range With 2019.16.x Software

[aerodyne]

I would try to charge at very low current for long time. I am not sure what the minimum setting is. The BMS might be able to bring that cell group in module 2 up to normal voltage.

I think at level 2, it is 3-5A. I usually charge at home at 28 A.

My lowest ever SoC was last week at 26%, followed by a SuC to 70%. That got me back down to 10 or so mv, but a few days later the 40mv imbalance returned.

 

[aerodyne]

Because they don't have new packs that would fit the old 85's???? Maybe....

The modules may not fit, but they must have cells that do. Don't the current MS and MX still use 18650 cells?

 

[MP3Mike]

The modules may not fit, but they must have cells that do. Don't the current MS and MX still use 18650 cells?

The newer packs still fit they just need a little adapter. Though there could be a weight issue with going with the 100kWh packs which are the only ones shipping in new cars.

 

[aerodyne]

The newer packs still fit they just need a little adapter. Though there could be a weight issue with going with the 100kWh packs which are the only ones shipping in new cars.

So, if we are careful, maybe a 100 KwH pack and a MCU2 can be had under warranty...

I can dream, right?

 

[Chaserr]

Something to think about....

Tesla is continuing to replace 85 battery packs that fail under warranty, with refurbished packs.

If there was truly some sort of defect in the pack and cell design, why would they do this, and risk more bad press, aka, "Rebuilt Tesla battery explodes on way home from service center" instead of replacing them with new packs?

It is conceivable charge/battery/regengate were done out of an abundance of caution, but also has the effect of reducing warranty claims, so no reason to roll back the changes.

Many recalled Takata airbags were replaced with newer Takata airbags that also has to be recalled again later. Some cars were recalled twice for the same Takata recall.

The NHTSA allowed this because it was safer than no airbag and safer than very old airbags that could explodes and kill someone. The danger was worse with age, but the temporary recall replacements had to be replaced with new airbags of a completely different design eventually.

Tesla's fire risk is correlated strongly with age but without some transparency we continue to conjecture.

 

[Chaserr]

So, if we are careful, maybe a 100 KwH pack and a MCU2 can be had under warranty...

I can dream, right?

The new 350 85 pack seems like it was purpose designed and built to fill the needs of this recall.

 

[aerodyne]

Many recalled Takata airbags were replaced with newer Takata airbags that also has to be recalled again later. Some cars were recalled twice for the same Takata recall.

The NHTSA allowed this because it was safer than no airbag and safer than very old airbags that could explodes and kill someone. The danger was worse with age, but the temporary recall replacements had to be replaced with new airbags of a completely different design eventually.

Tesla's fire risk is correlated strongly with age but without some transparency we continue to conjecture.

Well, I would not want Tesla to follow the Takata example. It did not end well for them.

I will assume it is difficult to economically repair an explosive squib in an airbag. However it should be possible to remove a few under performing modules in a large expensive HV pack.

I don't know much about the 350 packs you refer to. I'd be happy with a replacement pack with decent range and no "gates".

 

[JRP3]

However it should be possible to remove a few under performing modules in a large expensive HV pack.

Possible, but likely not practical, due to imbalances between the replacement cells and existing ones. I'd expect all modules to be replaced at once.

 

[aerodyne]

Possible, but likely not practical, due to imbalances between the replacement cells and existing ones. I'd expect all modules to be replaced at once.

I would be happy with that! I can't imagine they would do that and not fix condition x,y, or z...

It is not clear to me that they are using new cells in all cases. Eventually, they will have to, as the supply of serviceable used packs dwindles.

 

[DILIE]

Agreed.

But by the timing and wording of Tesla's press release, it does appear to link their actions regarding battery management to their investigation of the fire. Over a year later, have their investigations delivered any more detailed info.?

It is certainly possible there were separate actions in the firmware related to fire issues and longevity issues. Let's say X was potential thermal runaway due to gremlins in the cells. They ran test software and found it either wasn't likely (existing remediation was handling it) or didn't occur, but found module imbalances were causing module overcharges (condition Z), potentially resulting in a cascade of stress on modules.

This would equate to ramping decreases in overall battery performance/range with the potential to have failures within warranty.
This presents two problems: 1) Unhappy customers about escalating range reduction and 2) potential warranty replacements.

So they can honestly say they are changing the battery algorithm to extend battery life by modifying the max voltage (without really being open about it) and protect themselves from warranty costs. Most of the reduced range cars did stabilize at the new normal.

That said, it's up to a civil court to determine if this induced range reduction to protect themselves (and ostensibly the customer) is fair treatment of the consumer, especially with no explanation or option. It would be hard to convincingly deny the battery life extension, at cost to the customer in value, was done to avoid potential warranty costs. Admittedly, if they were open about the condition they would be inundated with people seeing varying, often reasonable, levels of range loss claiming it was Z and demanding replacements with law suits abounding. It's possible they completely misjudged the fallout from the change, thinking it would go largely unnoticed.

There may as well been other tweaks to prevent fires which are unrelated to range protection, or they could have some overlap. There is no question there have been fewer reported fires since the software was released. It could well be the practice of discharging the battery by running the fans to reach some determined level of safety, which would seemingly be unrelated to batterygate.

I'm curious if Teslafi might have a few vehicles which were exhibiting premature range loss or other issues consistent with condition Z.
Corresponding telemetry data would be quite useful.

At this point I think they've backed themselves into a corner or the push back from owners isn't really as big as we think in comparison to other things like delivery quality, service experiences, and just keeping the business going these days.

Every delay in the civil suit serves nothing more than to rack up legal fees. I expect at best it is settled under NDA if not eventually dropped.

Thanks everyone for the recent turn in tone on the thread.

Edit: Hope I got my X, Y, and Z usage correct. Will fix if someone points anything out.

 

[Dave EV]

Sometimes a visual helps. From SMT just now...note my "Problem child" module 2. Still get good range/nom capacity north of 260RM/75Kwh after a drive.View attachment 565366 

That's a lot of imbalance for a pack that's near not close to full or empty.

Since Tesla balances the pack at the top (bleeding off excess voltage/charge from modules at high SOC, when one module reads low, this means that there's quite a bit of charge to bleed off and the balancing circuits don't bleed off current that fast.

I would be very curious to see what the module voltages look like at 90-100% SOC and also at a low SOC (0-10%). Does module 2 come into balance better at high SOC? Or does it stay lower than the rest?

What are your typical charging habits?

 

[Battpower]

Since Tesla balances the pack at the top (bleeding off excess voltage/charge from modules at high SOC,

Could you offer some technical evidence to support that statement? (just a question, not picking a fight )

In order for the battery to perform safely and efficiently for the longest time, energy must always be spread evenly throughout all the cells. Things go wrong and can get completely off the rails if things get too out of whack. This means the battery / BMS MUST be tryingvto maintain balance at all SOC.

Speculating - In a brick with a bunch of cells starting to go high resistance (due to aging I suppose) the component cells are not absorbing or releasing energy evenly. Effectively the brick itself is becoming internally mismatched between good and bad cells. There is no process that can reverse this situation other than finding a similarly aged brick from an old battery that has maintained better matched internal cell degradation and swapping out the bad one(s)

So (still speculating), in the case of @aerodyne 's #2 brick, when he charges at low current for a long period, it gives everything maximum time to even out energy distribution, so voltages between bricks even out helped by BMS obviously. But inside #2 brick, the energy is not evenly spread. So when charging stops, there is an internal balancing of energy between strong and weak cells which is why the voltage of the brick drops.

In one sense, there is no 'new story' here. It seems to me all the conditions exist in a brand new battery, and through very precise manufacturing and selecting, cells are almost perfectly matched. They need to be in order to push the batteries as hard as Tesla does to extract fastest possible charge / discharge. It seems reasonable that even the best simulation of aging would not pick up on every outlier possibility, and unless you had a BMS working on an individual cell level (hugely difficult to implement I would guess and in most cases having little benefit) it would be inevitable that the cell matching drifts off as cells get older.

What you then see is what would otherwise happen if Tesla / Panasonic did not go to such lengths to produce ultra-matched cells.

 

[Battpower]

Post # 12443:

Because they don't have new packs that would fit the old 85's???? Maybe....

Actually, if it's hard to match everything up when first manufactured, it would be way harder once a battery is several years old.

There could also be multiple degradation modes effecting cells - based on original manufacturing tolerances, energy throughput in use, typical storage voltage, temperatures..... so finding matching used cells would not be easy.

But.... we are not talking about most cells.... Just some or even very few. So how and where to draw the line?

A well aired view is that lithium cells typically degrade in a manner such that they still have useful life even when their energy storage capability degrades below the level needed in EV's. The issue here would seem to come down to what to do with all the part-used cells that are left behind when we are done with them. Also, how to keep cells in use for as long as possible in the original EV application.

Since Tesla is well ahead of the competition in having to start dealing with this in volume, no surprise that universally acceptable answers don't come easily.

Energy storage is an obvious contender, but how to re-match cells into new packs given the widely different journeys they could have been on will be a challenge imo.

For us, I think a focus should be what goes on between parallel cells in a brick and if there is any way of spotting / mitigating. SMT (in conjunction with Tesla Logger) would seem to be a good way of recording dynamic conditions in batteries under heavy stress which I would guess is the first place you would see signs of this uneven deterioration.

 

[Battpower]

So they can honestly say they are changing the battery algorithm to extend battery life by modifying the max voltage (without really being open about it) and protect themselves from warranty costs. Most of the reduced range cars did stabilize at the new normal.

That said, it's up to a civil court to determine if this induced range reduction to protect themselves (and ostensibly the customer) is fair treatment of the consumer, especially with no explanation or option. It would be hard to convincingly deny the battery life extension, at cost to the customer in value, was done to avoid potential warranty costs.

Bold added by @Battpower. Expand to see all. These seem to sit well with the discussion and seem likely / reasonable. Doesn't include or exclude fire issue. That does feel separate in this context, but less so given the press release timing and content.

Admittedly, if they were open about the condition they would be inundated

High ideals can reap great rewards but potentially come at a very high cost!

 

[Battpower]

So when charging stops, there is an internal balancing of energy between strong and weak cells which is why the voltage of the brick drops.

I want to explore / understand this better. Can anyone help?

I am pretty certain that changes (eg: due to aging) between cells in a brick would have some cells in the brick absorb / release more energy than others when under load or charging. Just because the cells are in parallel does not stop changes to individual cells' internal resistances.

What is actually happening when, as per @aerodyne 's #2 brick, you remove the charging 'bias' and the cells in the brick find their own equilibrium?

Potentially, lazy cells gradually contribute less and less while the better condition cells are put under greater stress.

And during the internal equalisation in that #2 brick, I can vizualize the stronger cells effectively discharging into the internal resistance of the weaker cells, warming them up, increasing the current through the weak cells and also increasing the output of the strong cells..... Possible positive feedback & thermal runaway?

Simple version.... No control over what goes on between parallel cells in brick other than the individual cell fuses - that might be fine for isolating an odd rogue cell, but not able to protect against evenly spread gradual deterioration.

 

[MP3Mike]

Could you offer some technical evidence to support that statement? (just a question, not picking a fight )

In the past @wk057 said that his research found that balancing kicked in at a ~93% SOC. But things may have changed since back then as can be seen from this information from the Model 3 service manual: Discussion on detailed info from scan my Tesla and OBDII adapter : teslamotors

Brick Balancing

Note that the capacity of a pack is limited by the brick with the lowest capacity. When that brick is charging, it will gain voltage faster than other bricks. The HVBMS will stop charge when any brick reaches its ceiling voltage (~4.2V). If one brick has a significantly lower capacity that others, the pack will be limited by that brick which will get to 4.2V faster than the other ones. We refer to the brick with the lowest capacity as: min CAC. In periscope, its value can be seen by viewing the signal: 'BMS_cacMin'

Another limitation could come from bricks being imbalanced, or some bricks with a voltage higher/lower than others. This would limit ability to charge the pack as the brick with a higher voltage than others would reach the ceiling voltage early. Same idea when discharging, the brick with the lowest voltage would hit the floor voltage early which would cause the HVBMS to open contactors from low power

To mitigate this imbalance, Batman has some bleed resistor that can be placed and removed in parallel of each brick via a FET relay. Batman can put that resistor across the brick with the highest voltage which would slightly discharge that brick and bring it back to the level of the other bricks. Batman closes a FET which puts that resistor across the brick. The HVBMS will order Batman to put that bleed resistor across the brick with the highest voltage when Delta V is > 5mv MinBrickV > 4.0v (~85% SOC) && HVBMS State == STAND BY.

Note that Batman can also do balancing when the HVBMS is asleep .

The best way to balance the Model 3 pack is to set charge limit to 90% or higher and let the vehicle sit idle for hours (plugged in or not). 24 hours of balancing can reduce imbalance by 1mV.

 

[qwk]

Something to think about....

Tesla is continuing to replace 85 battery packs that fail under warranty, with refurbished packs.

If there was truly some sort of defect in the pack and cell design, why would they do this, and risk more bad press, aka, "Rebuilt Tesla battery explodes on way home from service center" instead of replacing them with new packs?

It is conceivable charge/battery/regengate were done out of an abundance of caution, but also has the effect of reducing warranty claims, so no reason to roll back the changes.

That’s because a neutered reman pack is again much safer due to the new software. If you have a neutered pack and somehow load the old non-neuter software, it gets Tesla’s attention right now. I’m not sure what the risk factor is regarding chance of fire, but they don’t want cars on the old software in a bad way.


There is so much misinformation in the last few days posts in this thread, it’s not even funny.

In a nutshell, Jason usually posts the truth except for a few things. It’s obvious that he doesn’t want to ruin his positive relationship with the company, so he keeps quiet. I can totally understand this position, but am not sure why he just doesn’t just come out and say that.

While there is a risk of him getting sued, everyone that knows anything about the issue, knows that the risk is not only low, the chance of Tesla winning is even lower. They have painted themselves into a corner with this, and just want to pay out as little as possible and be done.

At this point(since the majority of cars have the neutering software) it’s a civil matter, and the current lawsuit is the only hope for owners that are, or will be affected by this. The only other option for those affected, is to show tesla their own logs, in which case they will order you a new battery ASAP.

 

[Battpower]

That is the case with my little Renault EV I believe, but I know that SMT shows brick balancing at much lower SOC. Also, battery must be able to balance at lower SOC since Tesla allow you to set end of charge limit. If you set a reasonable 80% charge limit and the bms doesn't start balancing til 90%....... When does your battery balance? Also, you can see a battery balancing with SMT as it charges at lower SOC.

This thread is full of speculation. Not a little by me. Which makes the lack of hard verifiable facts all the more frustrating if they are out there.

 

[Battpower]

a neutered reman pack is again much safer due to the new software.

And there we are back to the safety issue again.

Don't push the battery as hard by limiting energy storage and charge rate and everything's good.

 

[MP3Mike]

Also, battery must be able to balance at lower SOC since Tesla allow you to set end of charge limit. If you set a reasonable 80% charge limit and the bms doesn't start balancing til 90%.......

Again, this could have changed over time, but with the Model S 40, Telsa would allow you to bring it in once a year and they would charge it to an actual 100%, instead of the software restricted 67%, so that it could trigger a balancing. (With so many software restricted packs now I would hope that the BMS can now properly handle balancing for software restricted packs.)