Service says $22k for new battery on 2012 Model S

[Jason Bloomberg]

" I've been debating on what exactly to do with my public info sharing and contributions for a while now since my posts here are restricted. ...

Anyways, thanks for the support!"

Could you "blog" on your own website so you would not be "moderated" but instead would be a moderator?

057 Technology

057 Technology, LLC

057tech.com

I'm curious what does the "057" refer to?

I've read some of your posts here and have a better understanding of some of considerations for options in repairing versus replacing a High Voltage Battery. Thank you for that.

We recently opted to have Tesla replace/upgrade our failing 85kWh battery in our 2012 Signature Edition P85 Model S to a new 90kWh pack with the latest batteries being put in the Model S/X. It cost us ~$22,400 installed and warranted by Tesla. Our charging speed has increased from 88kW to 120 kW, range has increased from 220-230 real world miles at 100% SOC to 270 miles, and it has a 4 year/50,000 mile warranty which can be utilized at any Tesla Service Center.

Our original pack was 400 Volts. The replacement is 350 Volts. Since the kWh is increased, or even if it was the same, my understanding is that compared to the 400 Volt versions, ratio of numbers of cells in parallel to series has increased with the intent of "spreading out the load" across more cells than the 400 Volt versions. What do you think of that approach to improving longevity?

I haven't noticed any loss of acceleration, which surprised me. Overall we're satisfied with the upgrade.

I find much of what I've read from you informative and hope you will continue to contribute your knowledge and experience to these forums.

If you are Jason Hughes, your homemade battery storage system made from recycled Tesla batteries is very interesting. I appreciate that someone with practical knowledge and experience working on projects like that is part of the discussions here.

Thank you.

 

[aerodyne]

Many cars have 16 modules.

Assume two are bad.

Each module is about 800 bucks on Fleabay.

A car with a bricked battery does not have zero value, even if nothing else is usable, highly unlikely.

 

[electricar]

And my point is that anyone who has a battery that's concerned about failure should probably sell the piece of crap and move on rather than risk the value going to zero and spending $22k+ to make it usable again. So far we have a statistical sample of one wrt a resale of an out of warranty model S with a recently replaced battery. Appears to add, maybe, $10k in resale value. Still negative $12k.

Some here are very good at doing Elon Musk pretend-fictitious-in-the-sky math where a hole in one's pocket is worn like a badge of honor. Maybe you've been gaslighted for so many years you can't tell the difference anymore.

Recommending to “ sell the piece of crap “ is one size fits all nonsense as is saying that the “ hole in one’s pocket is worn like a badge of honor “. Many Tesla owners are concerned about the environment and replacing a battery rather than buying a new car has less impact on it. $12K is more money than some people have but a rounding error on the net worth of others. I have seen people put 6 figures into old cars worth half that and been very pleased and, who knows, keeping a 2012 Signature Edition Model S on the road may even turn out to be a good move financially if they become collectible as I suspect they will.

 

[wk057]

To make sure I understand this correctly... If you have a pack that you suspect is unbalanced but doesn't throw an error the best way to know for sure is to do 3 days of full range driving and then the pack will detect this and shut down?

A "normal" imbalance isn't going to cause the BMS to shut the car down regardless of how much driving you do. The BMS can sufficiently compensate for imbalances that are within spec indefinitely regardless of driving habits. In a worst case the BMS will reach an equilibrium with the imbalance and hold it at that level. You may lose a few miles of range temporarily while it hones this in, but that's about it.

This only becomes an issue if there is an actual issue with something in the pack. In the case of an issue large enough to cause a problem (such as some ignorant mechanic cutting a fuse on a cell in a brick), driving as continuously as possible (drive 100% to 0%, supercharge, repeat) will exaggerate the issue and cause the BMS to do a protective shutdown more quickly.

As long as the BMS can extrapolate out that it will take a reasonable amount of time to keep things safe, it will keep functioning with no issues. Even the most out of balance packs in the fleet don't trigger this, because an imbalance in itself isn't a problem provided the BMS can predict that the imbalance is correctable in a reasonable time. If at any point a normal drive or charge could cause an imbalance to be pushed outside of a safety zone before the BMS can mitigate it, it has to prevent that action. Once it does that, it's not going to be like, "Sorry, please let your car sit for a week while I take care of this issue in your battery." ... no, it's going to say "Vehicle may not restart. Contact Tesla Service" .... and the "may" part might as well be "will".

 

[thefrog1394]

A "normal" imbalance isn't going to cause the BMS to shut the car down regardless of how much driving you do. The BMS can sufficiently compensate for imbalances that are within spec indefinitely regardless of driving habits. In a worst case the BMS will reach an equilibrium with the imbalance and hold it at that level. You may lose a few miles of range temporarily while it hones this in, but that's about it.

This only becomes an issue if there is an actual issue with something in the pack. In the case of an issue large enough to cause a problem (such as some ignorant mechanic cutting a fuse on a cell in a brick), driving as continuously as possible (drive 100% to 0%, supercharge, repeat) will exaggerate the issue and cause the BMS to do a protective shutdown more quickly.

As long as the BMS can extrapolate out that it will take a reasonable amount of time to keep things safe, it will keep functioning with no issues. Even the most out of balance packs in the fleet don't trigger this, because an imbalance in itself isn't a problem provided the BMS can predict that the imbalance is correctable in a reasonable time. If at any point a normal drive or charge could cause an imbalance to be pushed outside of a safety zone before the BMS can mitigate it, it has to prevent that action. Once it does that, it's not going to be like, "Sorry, please let your car sit for a week while I take care of this issue in your battery." ... no, it's going to say "Vehicle may not restart. Contact Tesla Service" .... and the "may" part might as well be "will".

Here's what I don't understand: if this is a guaranteed result of the Gruber method, how come there aren't people out on forums saying "Gruber method doesn't work" or whatever? Who are all these people paying $5k for a car that bricks itself after a week? Might you be underestimating the BMS's ability to handle severely out of balance packs?

 

[wk057]

Here's what I don't understand: if this is a guaranteed result of the Gruber method, how come there aren't people out on forums saying "Gruber method doesn't work" or whatever? Who are all these people paying $5k for a car that bricks itself after a week? Might you be underestimating the BMS's ability to handle severely out of balance packs?

Firstly, I think the number of people actually paying them to do this "repair" very likely is zero or close to zero. It wouldn't surprise me if this is just a bait and switch tactic, either, because there are virtually no packs that have the issue they describe where you'd even consider snipping a cell fuse anyway. Any random Joe TeslaOwner coming to them with a battery issue is not likely to have an issue supposedly resolvable by snipping a cell fuse. I've had at least three customers come to me for battery replacements after speaking with them. Not my place to go into the details, but suffice it to say two of those customers more happily shipped their vehicles all the way from the west coast for us to work on them vs going with what was offered elsewhere. (Other one was local-ish.)

Secondly, at the risk of sounding arrogant, I know just about everything there is to know about the Tesla S/X BMS. Guarantee I know more than anyone outside of Tesla, and probably more than some people who have even worked on the BMS at Tesla. I've worked on this for over half a decade now. I know exactly what its capable of, how the code works, and what its not capable of. Let's put it this way, I've completely designed from scratch a drop-in replacement for the Tesla BMS that lets me run a Tesla vehicle off of any battery I want. See my earlier posts about running a Tesla Model S off of a Chevy Volt pack in the trunk.

Lastly, and perhaps more damning than any of the above: I've tested this. We've tried all sorts of ways of repairing packs with issues before, and the "cut a cell fuse" method does not work. Granted, that's not something I intentionally did because, well... I have a brain. Instead I had a pack with accidental physical damage to a single cell fuse on a module from a botched BMB replacement repair (attempted by a third party). In an attempt to get this car up and running, I manually balanced the entire pack to within 3mV and got the thing running again. I tested every module and confirmed the only issue was the one missing cell fuse. Everything worked fine...... for about 5 charge cycles. Then the BMS threw the initial (hidden) SoC imbalance code. After about 2 weeks of normal use, the car was not drivable at all and had to be towed back.

We also tested other "simple" fixes, such as replacing an individual module entirely with one that had as close as possible of a capacity to the rest of the pack. We tried this several times. Even when matched as best as we could (which I honestly thought was super close, since we had a pack built in the same month from a similar mileage donor car and pretty close measure capacity), we still couldn't satisfy the BMS for an extended period. This was our best effort, and the car worked normally for about a month before the BMS gave up due to the persistent imbalance. Logs shows the BMS was concerned after only a few days of use, and eventually threw in the towel after about a month of use (only about 1500 miles).

Suffice it to say, outside of Tesla no one has as much knowledge and experience with these batteries, BMS, etc than I and my company do. We've processed hundreds of Model S/X vehicles, and thousands of battery modules. We don't have to make flashy professionally edited YouTube videos about everything we do in order to grab customers. So far, 100% of our business is word of mouth or otherwise organic search stuff... which is impressive to me considering how objectively terrible our website is at the moment. But, function over form, for sure... I mean, I have some YouTube videos up, but lol.

Anyway... to summarize, the Tesla BMS was not designed to handle the complete loss of one or more cells. It was designed to detect this so the pack can be replaced. It's just basic math. 100mA bleed vs a ~3200 mAh delta (lost cell) in a ~250,000 mAh system is just not viable. Even the worst "natural" imbalances I've seen are on the order of few hundred mAh... not thousands.

 

[fasun]

@wk057 Need to connect with you regarding repair on my model S. Recently got quoted for 85kwh battery pack replacement. Not sure if I'm getting too old that I am unable to figure out how to message you directly. If you can send me a PM, that would be great.

 

[wk057]

@wk057 Need to connect with you regarding repair on my model S. Recently got quoted for 85kwh battery pack replacement. Not sure if I'm getting too old that I am unable to figure out how to message you directly. If you can send me a PM, that would be great.

It's probably not you. My account is screwy here on the forum, which screws with my PMs. I can't send PMs at the moment, unfortunately.

contact at 057tech dot com (not hiding email from forum, hiding from bots...)

On a side note, I have a 90 pack with ~80 kWh usable available (~270 rated miles in a RWD, ~280 in AWD non-P) for an upgrade/replacement for someone who wants it within the next week-ish.

 

[wk057]

On a side note, I have a 90 pack with ~80 kWh usable available (~270 rated miles in a RWD, ~280 in AWD non-P) for an upgrade/replacement for someone who wants it within the next week-ish.

And its gone before my post even gets approved.

 

[cduzz]

And its gone before my post even gets approved.

What's such a battery cost, installed, assuming my existing battery is full of jellyfish and not usable (except to the local carnival's game of "guess how many jellyfish are in this battery case and win a prize! game)

 

[ce2078]

Maybe I'm a little late to the party here, but let me introduce myself. I am Matt / CE2078. I am actually the one who inadvertently taught Gruber how to repair Roadster, S & X battery modules. I have actually been doing S battery repairs since 2016 - which is likely the first cell level repair done on any Tesla. Does Gruber fully know how I do it? Not really. They sort of figured out part of my process, enough to "get it done" but from what they have shown me, they don't know everything that I do. Thus its likely harder for them to diagnose borderline failures. I guess that's what I get for recovering one of their Roadster sheets that had a group of cells at 0.8V and another at 0.0V and I brought it back to life with stable voltage (not for use but as proof of concept). Likely the first Roadster cell level repair. They told me it actually gave the Tesla engineers something to think about...

@wk057 , I agree the current software flags a pack with imbalanced SOC between modules. It doesn't matter if a cell fuse has been cut or not. IMHO the current software is very aggressive at failing perfectly usable packs that are just too worn out for Tesla's liking (among other reasons I won't get into here). I've diagnosed several packs that have been failed by recent software and found no issue with them except some cell groups that have a SOC that is different from the average by a few percent. The solution? We downgrade these cars to older software and they no longer have an issue. Hopefully Tesla gets their act together on the software side and realizes that packs age and can have different module SOC levels and still function fine. Enough with this trying to maintain a pack with perfectly balanced voltage throughout the SOC range - it's a fantasy on all but brand new packs. Balance at the top end (or wherever they prefer) and if the pack has a 50mV difference at the bottom end, no big deal. Lowest/weakest module determines pack capacity - not the average of the modules.

I don't cut the cell fuse for low capacity or worn out modules. Those do need to be replaced, but just the offending module not the entire pack. I do cut the cell fuse for the "Weak Short" failure - where one cell is pulling the group of cells down to a point where the BMS cannot keep up. The result? The car becomes usable again with a roughly 5 mile reduction in range.

The BMS itself is able to handle 1 or 2 cut cell fuses no problem and it can definitely handle more. But after that I would recommend replacing that module, just so you don't lose too much range. Does the new software like this fix? Of course not, thus we downgrade the car. Does my fix last? Except for the software issue Tesla introduced last year, I have not had anyone come back with an issue. I know the first 2 cars I fixed in 2016 are still on the road (one of them a forum member who pointed me to this post), they Supercharge and drive just fine. Both had 1 cell fuse cut and lost about 5 miles of range. This is by far the least expensive way to repair the pack, just maybe not the quickest.

But WK057, you have to admit it's the recent software that's being overly aggressive - not the BMS that's at fault. If it was the BMS, then why could I put a 40k mile 85 -E battery with 1 cut cell into my 70 S and drive it to 170k pack miles with no issues? I've driven that car AZ <> FL five times on that pack and all over the southwest, supercharging nearly half of those miles. You and I both know that a few 10's mV isn't the end of the world for the pack. As long as the pack can balance at the top end and stay balanced there, the pack is fine - the car just needs to watch the lowest module, keep the pack from over and under voltage conditions and base pack capacity from the weakest module. It's all common sense for anyone who has dealt with high voltages packs.

 

[Zuikkis]

This thread has taken a very interesting turn. I keep checking daily to see if there's anything new from @wk057 , and now @ce2078 joined the fun!

I guess the recent change in the software is related to batterygate issues like slower charge times? I guess Tesla is really scared of batteries catching fire or something? Many people seem to think that batterygate was done to keep batteries working until their 8 year warranty runs out, but this @ce2078 's information seems to indicate that it is more a matter of safety?

You probably can't downgrade MCU2 software? And what happens with MCU1 if Tesla pushes OTA upgrade, is the change reverted and battery fails again?

 

[cwerdna]

^ Based on videos I've seen of both Tesla and Leaf pack fixes, it's a similar process to disassemble the pack, pull out modules, test, isolate, fix and put back in the car. The difference of 7000 cells vs 192 makes very little difference to the ultimate level of complexity.

I haven't read nearly all of this this thread but I disagree with your claim that they're similar in complexity. Nissan actually has dealers swap out bad modules like this one: 2012 SL with range collapse - My Nissan Leaf Forum (after repair at 2012 SL with range collapse - Page 2 - My Nissan Leaf Forum). Nissan doesn't have the whole pack sent somewhere to be repaired if such things happen.

Cor at Anybody with 2011 or 2012 interested in upgrade to new battery? - My Nissan Leaf Forum (read thru all his posts in that thread) actually does swaps of all modules from one casing into another. The reasons are explained in his posts (related to physical differences between a newer pack in better shape from a crashed Leaf vs. a heavily degraded older pack).

Anybody with 2011 or 2012 interested in upgrade to new battery? - Page 18 - My Nissan Leaf Forum was about one of his many swaps and Anybody with 2011 or 2012 interested in upgrade to new battery? - Page 20 - My Nissan Leaf Forum has the time lapse video.

There's no liquid cooling and dealers don't muck w/the individual cells, only modules which contain multiple cells in what looks big sardine can in older revs. I've held one of the modules in my hands at a Nissan/LEAF EVent years back. Was given by Nissan to the SF BayLEAFs group.

 

[Zuikkis]

I've actually built an EV from old Leaf packs. They are indeed like huge sardine cans, with M6 terminals so it's very easy to add cables etc.

Here's my pack:



See how simple the construction is.. There's 20 Leaf modules, that's about 12kWh total capacity.

 

[ce2078]

On MCU1 if you go back a little over a year - no more updates. I don't have a solution for MCU2 as of yet.

I honestly believe Tesla got overaggressive with some of the new settings they introduced recently and hopefully they will relax them soon. The software is looking for a "Weak Short" but flagging packs with no short at all, just differing AH capacity causing a voltage deviation at different pack SOC levels. I took a pack flagged for a Weak Short and removed it from the car, let it sit for almost 2 months and noticed no difference in any voltage levels except a few mV lost evenly across the pack, likely due to cooler weather (BMS off, no balancing being performed). There would have been a noticeable voltage loss if there was any kind of short.

There is really no reason why a pack with a voltage delta of 0.05V should be considered bad. The only problem with a voltage delta is reduction of range. Essentially the pack is full when any one of the bricks registers "full" and the pack is empty when a brick registers "empty". Tesla sets these Full/Empty parameters. The packs usable energy is between these Full and Empty thresholds, which could be determined by a single weak brick (same brick triggers both full and empty) or different bricks with unbalanced SOC levels (one module triggers full and a different module triggers empty). One of the BMS's jobs it to fix the latter, but it can only measure and report on the former. But neither should be marked as a failure unless there is a confirmed issue - i.e. the BMS is constantly chasing a short in one brick.

Now let's imagine a pack where we fully charge it and allow all bricks to balance at 4.200V, then we turn off balancing. Now we discharge the pack down to 3.900V but one brick is 3.890V, we then discharge to 3.600V but the lowest brick is now 3.580, next we go to 3.300V but the lowest brick is now 3.250V. Now when charging, the voltages rise in a similar pattern and by the time we get to 4.200V, all bricks are the same with no or minimal balancing being performed. This isn't a short, it's just a low capacity brick who's voltage depletes faster since it stores less energy. This pack would be failed today by Tesla's software - but the pack is just fine.

Not let's imagine a different pack. We do the same as above, but when we fully charge it after discharging, one brick is still 0.050V less. We make a note of this, balance it again, and repeat the test another time. Same thing, the same brick comes up 0.050V below the rest. Now we have identified a weak short and the pack should be repaired or replaced ASAP, as a weak short could become a "Not So Weak Short" in the future.

The problem is that Tesla took the approach that the pack should have balanced voltage at all SOC levels, which like I said before is nonsense. Pick a level that the car will see for a length of time regularly - maybe what the daily charge percent is set to? And test as above between multiple charge cycles. They should still watch the voltage delta for sudden deviations, but factor in the known delta.

 

[beatle]

New software is geared for overly aggressive safety measures? Planned obsolescence?

 

[ucmndd]

New software is geared for overly aggressive safety measures? Planned obsolescence?

I suspect it's more that they're in the phase of company/technology growth where they have a lot of incentive to be overly cautious and not many to push the boundaries.

Every EV fire or battery incident becomes front page news right now, and contrary to the old saying that IS indeed bad press. The great majority of cars are still in warranty, so if they make the conscious choice to be overly cautious and end up doing more pack replacements on their dime, they seem to have accepted that as a calculated risk for now.

Seems like the sort of thing they could relax over time after they have more data and analytics showing the real world safety, etc.

 

[David_Cary]

Except that every battery failure out of warranty diminishes the value of every car out there. And ruins the environmental creed of the car.
One of the arguments for Tesla, large battery and amazing BMS is that the battery would last a long time. Having an average life of 9 or 10 years is horrible given the upfront environmental cost. Really makes Nissan's approach seem better.

The battery replacement cost has always been one of the forefront questions for EVs and has the battery longevity question.

The customer service put me one the edge of never recommending Tesla to anyone but the battery problems have sealed the deal.

Super appreciate the BMS experts on here giving their input. But it only makes me more frustrated with Tesla's approach.

 

[henderrj]

My 2014 has gone the mean distance to the moon on its original battery! It's pretty hard for me to imagine that that could possibly be worse environmentally than my driving a diesel for that whole time. Let alone a gas vehicle. I think the handful of replacements that happened in the batteries would be much better to be free under warranty than $22,000 after warranty. The amount of bad press you get from that would be enormous!

 

[BCTS]

The most interesting aspect of what wk057 / Jason has indicated to me is the complete inability to repair/replace individual components of a pack.
Essentially the only options you have to repair a faulty/defective pack are to
- Replace Pack with New/Refurb from Tesla
- Replace Pack with Good Used Complete Pack from Wrecked Tesla (pending some workarounds as Jason has outlined)

Attempting to repair the pack with similarly dated/age cells just doesn't seem to work (from Jason's experience).

I've had similar experiences with NiMH Batteries in the Toyota/Lexus Sphere. Their early Hybrids such as 1st/2nd Gen Prius or 1st Gen GS450H had some quite high tolerances for charge/capacity levels in the BMS between Cells. This allowed you to remove/rebuild individual cells to get a pack operational again.
I did this personally to 3-4 Lexus GS450H Vehicles in Australia and had good results as long as donor cells were fairly close to originals.

However the 2nd Generation GS450H and Newer Hybrids such as 2010 onwards Toyota Camrys had much tighter tolerances on Cell variance, so it become pretty much impossible to rebuild the packs with replacement Cells etc.

This essentially means that a "defective pack" is unlikely to be recyclable into another Tesla at all.
Only recycling options are either
- Home Made Powerwall
- Home Made EV Battery
- Home Made etc etc projects.

Also leads me to wonder/suggest that there is probably little or no difference between a Tesla "New" and Tesla "Refurbished" Battery pack.
Surely the effort involved in testing/matching suitable cells would be too difficult. Tesla would potentially just fit New Cells and perhaps some refurbished internal components etc.

The Nissan Leaf appears to allow for much greater variances in Cell Capacities also. I've done some work on 24KWH to 40HWH Upgrades with some 3rd Party Canbus Adaptors and they work quite well. The interesting aspect of that is that you can fit an older 24KWH Battery to a Newer 40KWH Leaf without any issues.