Starting to see 2015s with BMS_u029 and 018

[brainhouston]

Didn't battery improvements happen with 70/85 > 75/90 change? i would assume so since the HV pack case changed a bit...

And here once again advocating for this completely unsafe practice AND trying to discredit those of us who know better?

I think u either skimmed the post or missed the point...
Didn't sound like he was promoting, just stating the facts.

On the technical side, can someone explain this:
If the fires are caused by overcharging cells after u029 reset, shouldn't BMS still prevent any cells going over 4.2V??
I don't understand how even with reset BMS can't prevent catastrophic event..
More details would be welcome here

 

[wk057]

Didn't battery improvements happen with 70/85 > 75/90 change? i would assume so since the HV pack case changed a bit...

No, there were improvements pre-75/90. There are 70/85 packs in the new chassis, for example.

I think u either skimmed the post or missed the point...
Didn't sound like he was promoting, just stating the facts.

I mean, all he does is promote his Facebook group here it seems, but I think you're the one who didn't read it.

Clearly was attempting to discredit those of us who have noted that resetting the error is unsafe. Claiming a non-existent economic incentive, as well as claiming that so many people have been successful with their reset vehicles.

If that's no promoting it, then I'm not sure what would qualify to you.

On the technical side, can someone explain this:
If the fires are caused by overcharging cells after u029 reset, shouldn't BMS still prevent any cells going over 4.2V??
I don't understand how even with reset BMS can't prevent catastrophic event..
More details would be welcome here

Not necessarily caused by overcharging cells. This can be the case if you reset the error (see below) but there are other reasons to limit charge on BMS_u029 cells. I think @Recell did a very descriptive post on this not long ago.

One of things BMS_u029 can mean is that the BMS has calculated and determined that some readings are unreliable and may be incorrect. It can attempt to algebraically substitute these readings using other info, but can only do so much and that has a huge margin of error. There's even a failure mode that's almost completely "curable" via software (I wrote about it a couple of years ago). If it's not, it then sets a limit on charge level at a max well below what it believes the error bars would be on the bad readings and throws the error. In the meantime, it will still creep max charge downward (usually pretty quickly) as it gets a handle on things until it throws the error message.

Then, if you reset the BMS's hard work in determining this by wiping out its brains (which you have to do to reset this error), it now no longer knows that a reading is incorrect, among other things. Then you can definitely overcharge a cell group and the BMS would be blissfully unaware, when it normally would have determined this failure over the course of many months as it began and have been 100% safe without the BMS lobotomy.

 

[Recell]

we’ll be more direct in answering @brainhouston ‘s question.

BMS_u029 prevents the highest brick in the pack from charging above 3.66V. that limit is widely considered to be safe, even for a defective cell. safe enough in fact that the UFC (Unified Fire Code) considers a lithium ion cell below 3.66V to be sufficiently inert as to not be an undue fire risk. not zero risk, just an order of magnitude less risk.

so, resetting BMS_u029 eliminates this limit, allowing an unsuspecting owner to charge a battery pack up to 90% or more. this is well into the danger zone for a defective cell. this is how you cause battery fires.

this is truly serious. please please don’t reset BMS_u029 errors. you are putting property, and potentially lives, at risk.

 

[Rocky_H]

I’m gonna start disagreeing with these posts until you provide evidence of anyone receiving a third party battery replacement for anything reasonably close to $5,000.

You've been around here long enough I thought you had seen this already. Recell already does this with published, fixed pricing directly on their website! It's $7,895, which is higher, but I might say "reasonably close" to $5,000. But wk057 has talked about how their price is similar, but variable, because it depends on what the problem with the old pack is and how much salvage value they end up getting from it. I think he has mentioned these are frequently in the range of $5K to $10K.

But sure, it's not very responsible or helpful for someone to always quote the lower end of the range as if it's what most people should expect.

 

[Arctic_Fox]

I've been asked by several and seen posts questioning how we can know for sure that the reset was the core issue.

It's 100% the fault of the reset. Again, there's going to be a lawsuit against the resetter and seller on this, so I'm not going into details. If the media does pick up on it (not sure if they have or will), that's on them, but I'm respecting the privacy of the owner on this.

However, to the fault point, I asked a friend who is in a position to do so to pull the vehicle logs and BMS debug info immediately when this would-be client informed me that they'd charged the car to 100% and had found out this error had been reset. The only reason this person still has a house to live in right now is because I managed to parse this information in time and told him to park the car outside his garage and away from his home. Also gave them instructions on how to attempt to discharge the battery, which unfortunately did not happen in time to avert disaster.

The vehicle logs contained the brick ID of the originally pre-reset offending brick triggering u029, and u029 was clearly and successfully protecting the the vehicle from a catastrophic issue up until the logs showed the error disappearing and other BMS data shifting inappropriately and inexplicably towards virgin values.

By the time the latest info was pulled, that same brick was already an issue and the BMS was trying, futilely unfortunately, to deal with it. From what I can tell based on data obtained this morning (the car tends to send and log a lot of data as errors pop up), the BMS tried quite valiantly to protect the car, and even made a last ditch effort to avoid disaster by commanding the chiller to 100% at 100% coolant flow, managing to feed coolant at 5C/41F into the pack as a result.

Again, I'm not going into many specifics for a multitude of reasons, but suffice it to say that resetting of the u029 error and subsequent charging above the originally imposed u029 limit was 100% the root cause of the fire. Zero doubt. The car would have been completely safe with the imposed limitation had that error not been fake-fixed.

How can you read the BMS logs? Can they be read even if a vehicle has had a factory reset? I want to know what errors occurred before I bought the vehicle, as my extended warranty claimed the issues existed before the sale (denied claim), but they have no proof. Neither do I.

 

[ucmndd]

You've been around here long enough I thought you had seen this already. Recell already does this with published, fixed pricing directly on their website!

But sure, it's not very responsible or helpful for someone to always quote the lower end of the range as if it's what most people should expect.

We’re saying the same thing.

Recell’s cheapest option is 63% more than “$5,000”. That’s a big difference in my book and represents the best case - smallest replacement battery, no return shipping, etc. Jason’s pricing is by all accounts similar but situationally dependent and I’m pretty certain he’s never given anyone the impression that an average replacement cost in his shop is “close to $5,000”.

It is flatly unhelpful to spambot every thread on this topic setting expectations of pricing that is at least 60% below reality in almost every case. It does nobody any good.

 

[Arctic_Fox]

We’re saying the same thing.

Recell’s cheapest option is 63% more than “$5,000”. That’s a big difference in my book and represents the best case - smallest replacement battery, no return shipping, etc. Jason’s pricing is by all accounts similar but situationally dependent and I’m pretty certain he’s never given anyone the impression that an average replacement cost in his shop is “close to $5,000”.

It is flatly unhelpful to spambot every thread on this topic setting expectations of pricing that is at least 60% below reality in almost every case. It does nobody any good.

Yeah, the dealer trying to get out of a battery replacement of 16K wants to use a third-party shop for 5K. I said no. Almost a 4-month battle.

 

[wk057]

How can you read the BMS logs? Can they be read even if a vehicle has had a factory reset? I want to know what errors occurred before I bought the vehicle, as my extended warranty claimed the issues existed before the sale (denied claim), but they have no proof. Neither do I.

Logs are stored on the gateway, and aren't wiped with a factory reset. As far as I know, no one besides myself and Tesla knows how to access and decode them, so, there's that.

 

[Crutonius]

This situation with the lawsuit against this shop that reset the error is interesting. I'm sure the media will cover it since they love covering any time a Tesla catches fire.

Hopefully some good will come of this and Tesla will reckon with the horrible situation of facing an out of warranty battery repair.

 

[Cheese Guy]

Ok, so I am just misreading his chart. Apparently, he took it in for a replacement as soon as the 018 showed, even though it is still drivable but at no more than 50% charge. At least, that is my understanding.

Yes, correct. Charge was limited and sent the car in for repair. Had a warranty with 057 Tech and battery was replaced. Excellent service with Jason and he kept me updated during the entire process. The jump you see is the replacement.

 

[islandbayy]

I’m gonna start disagreeing with these posts until you provide evidence of anyone receiving a third party battery replacement for anything reasonably close to $5,000.

Recell put a refurbished 85 in my 2012 P85 for about $7500. I'd say thats close to 5k. Definately less than half, closer to 1/3 teslas price. Battery's capacity is also quite a bit higher than my 2016's 90 pack.

 

[ucmndd]

Recell put a refurbished 85 in my 2012 P85 for about $7500. I’d say that’s close to 5k.

50% higher is “close”? Do you work in politics or something?

 

[Recell]

we keep it simple. seems to work

"Dollar-for-dollar the RC 210+ is our most popular battery pack - a commute-ready daily-driver for under $8,000, including shipping and installation."

 

[Rocky_H]

50% higher is “close”? Do you work in politics or something?

I think people are looking at the perspective of things. If someone is looking for cheaper used cars, they may see some that are $2K or $3K or $5K. Sure, if you want to look at the percent differences, you could say that one is over 100% more in price and clutch your pearls, but they are all less than 10 thousand, which is fairly low and similar in prices.

In like manner, when people are facing $15K or $20K for battery replacements, $5,000 or $7,500 or $8,000 all seem fairly reasonable in that realm.

 

[ucmndd]

In like manner, when people are facing $15K or $20K for battery replacements, $5,000 or $7,500 or $8,000 all seem fairly reasonable in that realm.

We can agree to disagree on this one. I’m glad companies like 057 and ReCell exist and they’re doing great work at an economical price compared to Tesla.

But if I called either of them and asked how much I should expect to spend and they said “close to $5,000” and when the invoice came it was for $8,000, I’m sure gonna feel like I was not given realistic expectations up front (not suggesting either of them would do this).

ChatGP@Electric700 isn’t doing anyone any favors.

 

[RockinWithKids]

No, there were improvements pre-75/90. There are 70/85 packs in the new chassis, for example.

@wk057 got an idea as to approximately at what VIN these improvements started?

 

[RockinWithKids]

we’ll be more direct in answering @brainhouston ‘s question.

BMS_u029 prevents the highest brick in the pack from charging above 3.66V. that limit is widely considered to be safe, even for a defective cell. safe enough in fact that the UFC (Unified Fire Code) considers a lithium ion cell below 3.66V to be sufficiently inert as to not be an undue fire risk. not zero risk, just an order of magnitude less risk.

so, resetting BMS_u029 eliminates this limit, allowing an unsuspecting owner to charge a battery pack up to 90% or more. this is well into the danger zone for a defective cell. this is how you cause battery fires.

this is truly serious. please please don’t reset BMS_u029 errors. you are putting property, and potentially lives, at risk.

@Recell Does a not f-d with BMS recognize the voltage problem at any SOC or do you have to charge it up to 90% or higher for the BMS to pop the u029 error code?

 

[Recell]

@Recell Does a not f-d with BMS recognize the voltage problem at any SOC or do you have to charge it up to 90% or higher for the BMS to pop the u029 error code?

certainly self-discharge/unexpected SoC deltas can be easier to detect/easier to isolate with higher confidence levels at higher voltage potentials, but charge cycles below 90% are perfectly capable of popping a u029 alert. we’ve seen unexpected SoC deltas (ie all the data points that feed into a BMS_a117 weak short alert and ultimately trigger the BMS_u029 charge limit) being raised with as little as 20-30%

@wk057 may be able to add some additional color

 

[ddb1001]

Logs are stored on the gateway, and aren't wiped with a factory reset. As far as I know, no one besides myself and Tesla knows how to access and decode them, so, there's that.

Can you provide a service similar to Carfax? A person enters the VIN and you tell them the vehicle's history.

 

[AntronX]

@wk057 How does this battery failure affect those of us who use Model S modules for offgrid solar storage with 3rd party BMS? Is cell voltage monitoring alone enough to catch and prevent this thermal runaway failure? Are you concerned with your house solar battery being at risk?

Edit: found this explanation by @Recell: Battery defect for IT brainstorming BMS_u029 - P85 MS 2013
Looks like some cells are developing shorts through separator. Lithium dendrites?

Edit2: Should I limit my ModelS modules to 50% SOC or 3.67V/cell? So far i have no cell voltage misbalance on any of my 6 modules. I check voltages each week.