BMS-029 - Tesla Must Do Better

I am an early adopter of Tesla. I bought my Model S in Feb of 2015, laying out close to 90k for a car – which is something I would have never dreamed of doing previously. But I believed in the mission, I believed in the car – so I traded in my Kia Optima and subjected myself to this grand experiment. This was the days where the masses really didn’t know what Tesla was – I would get stopped in parking lots and get strange looks on the road – and I gave makeshift mini-presentations about how this was the future of transportation.

Fast forward 8 years and 3 months later…. After Supercharging I received an error on my screen that said “maximum battery charge level reduced” and gave the code BMS_029. After 8 years and 85k miles of very happy ownership – dealing with the usual door handle replacements, window regulator breaks, new MCU, new front dash screen etc – I now realized I was faced with something much more serious.

I planned on keeping my car indefinitely. I love the car. I have never loved a car, but I do love this car.

3 months after my battery warranty expired, I have this error that is going to limit me to about 35% charge, and from everything I read online it is basically a battery death sentence. The Tesla equivalent of the “blue screen of death”. I went to the Tesla Service App, and explained the error and a screenshot – and got back a 15k estimate. No phone call, no options, no offer of repair, no diagnostics - just give us 15k and we will fix it.

Thank goodness for the online community. I found a Facebook group dedicated to this, and lots of help at Teslamotorsclub. I am not an engineer. I am simply a normal consumer. I feel like that needs to be said because if not for the amateur Tesla engineers out there, and aftermarket technicians – I feel like there would be zero information about this because Tesla isn’t talking or explaining. They simply text you back an estimate in an app, with one option – pay us or your car is dead – bricked.

So after doing lots of reading online – and talking to several experts – these are the options:

• Error removal through software. There are people out there who will (for about $500), simply remove the error so that you can go back to where you were the day before this dreaded error showed up.
• Pay anywhere from 8k to 9.5K to ReCell or another 3rd party for a remanufactured battery. You will get a battery pack from a car that they previously replaced, and remanufactured for you. Your battery will then be remanufactured and sold to someone else. You will get a battery pack that is dated anywhere from 2012 to 2015 and a 2 year 25k warranty.
• Pay Tesla about 15k for exactly what ReCell does, but get a 4 year 50k warranty.
• Buy a brand new 90KWH battery from Tesla for about 19k, and get a 4 year 50k warranty.

Option 1 seems like the absolute worst option. It seems like this is widely advised against, as this simply removes the error but doesn’t fix the root cause – which could be catastrophic. This part seems obvious. But hiding under the surface is a very big problem for Tesla – and for Tesla owners – the resale market can never be trusted. When I got this error – overnight – my resale value went from 30k to 10k. If I can remove this error, it goes back up to 30k. So it is obvious that there will be lots of unsuspecting buyers who end up with a car that is going to get the error again – or a potential big problem with the battery – either from a dealer who buys it for 10k and removes the error and sells for 30k, or an individual. This seems like a PR disaster for Tesla – and a horrible situation for consumers. It has already happened multiple times.

Option 2 and 3 are very similar – really just warranty differences. But in the end, if you can get a brand new battery for 4k more, and you plan on keeping the car for a long time, ReCell and Tesla need to do a better job of educating the average consumer (like me) that a reman battery with 8-10 year old cells has a value proposition vs a brand new battery. I fully support ReCell and their mission, because they are doing what Tesla does and beating them on price – and for the right person – it is a great option.

I chose option 4. I hate that I am laying out 19k to basically get back to where I was before the error. But at the same time – with the limited information I have – especially from Tesla – and very limited options – it is the best decision for me. My car is at Tesla right now sitting waiting for the work to be done.

Tesla needs to do a much better job addressing this, and develop a program that has better education and options. Are they trying to get the early cars off the road? Are they trying to get the unlimited supercharging cars off the road? They are getting my battery as part of the 19k repair – and they will remanufacture that and sell it to someone else for 15k. How much work and cost goes in to the remanufacturing? What if it is a circuit board or a few cells or even a module on my battery – that costs them close to nothing in comparison to the 15k they will flip it for – is that fair that I pay 19k on a car that is only worth 30k, and they ALSO get my battery?

Tin foil hat time…. I don’t necessarily believe any of the following to be true – but as Elon likes to say on Twitter – “I am just asking the questions”. What if there was a company that could press a button and send an error to a car fresh out of warranty, and essentially brick it knowing that they then would charge between 15k and 19k to replace it, and in return get a battery that they will sell to the next person they send the error to?

It seems a lot of cars are getting this error just after 8 years. Tesla – isn’t it in your best interest to be more transparent about issues, education, and options? Do you not care that the people this is happening to are the same people who in part built the company to what it is today? I have probably sold 20 people over the years on buying cars, and I have bought a MY. I am not suggesting Tesla owes us anything – but it just seems like a smart business decision to better handle this.

There are lawsuits already out there. Who knows. One persons opinion… This experience has seriously diminished my faith and experience in Tesla. I am biting the bullet – spending 19k on a car that will only be worth 30k when done – but I will always wonder if the BMS_029 error was just a software glitch, a $50 circuit board, a real problem that just happened to occur at 8 years and 3 months – or something much more sinister.

Come on Tesla, you can and need to do better.

"2014 Tesla Model S" by harry_nl is licensed under CC BY-SA 2.0.
Image added for Blog Feed thumbnail

 

[MP3Mike]

With an EV battery, there are thousands of the exact same individual cells. I just refuse to believe that in the long term, someone won't figure out a solution to bypass a single cell instead of replacing the whole pack.

Perhaps the BMS could tell the techs which cell(s) are failing, the pop the bottom plate off the car, access those cells by coordinate and swap them with fresh ones. OR the BMS just learns to stop using them.

You can keep dreaming...

Whatever that takes - complex switching inside of each module perhaps?

That sounds cheap and reliable, NOT.

 

[tesluv108]

It would be in Tesla's best interest to address this problem. If the customer base believes that it's not practical to own a Tesla beyond the warranty period then it's going to have very negative consequences for Tesla.

Yes but Tesla is soooo into sustainability! I mean who would want to keep a car they paid 132,000 for in 2015 on the road! It’s just so much better for the environment to build a NEW Tesla !

 

[father_of_6]

You can keep dreaming...

Sure, you can say that. But here we are driving electric cars around the country. Dreams become reality.

Our battery packs are like a big tray with thousands of AA batteries, and when *one* goes bad we replace the *entire* tray of batteries for $15k. It's outrageous, and I choose to believe that this is solvable, by whatever means.

That sounds cheap and reliable, NOT.

Does it really seem impossible to sandwich the modules in PCBs and let the BMS disable cells or sectors of cells as necessary?

As mentioned, I'm not an electrical engineer... I just don't think the problem is as impossible to solve as it might seem.

 

[MP3Mike]

Does it really seem impossible to sandwich the modules in PCBs and let the BMS disable cells or sectors of cells as necessary?

Possible? Yes. Practical? No.

Something to remember is that in the Model S packs each cell in a group that get disconnected would reduce pack capacity by ~1.5%. (While putting more stress on the remaining cells in the group.)

 

[tesluv108]

The reality is this is a huge problem for Tesla. Because people that own these S cars are realizing Tesla doesn’t give a flying “f” about the customers that helped build the brand. Tesla giving up on the older cars and suggesting an expensive repair to keep it on the road is strategic on their part. They are betting most people will give up on the car and simply buy a new one. But why would a customer who felt wronged by Tesla actually agree to buy a new one? The answer is they won’t and other brands will benefit from their customer service ineptitude. Tesla has no interest in two things 1. Offering an extended battery warranty & 2. Warranting a new replacement pack for 8 years. Why would they want the customer to get perceived value out of a 19,000 expenditure? No one can answer why Tesla is offering the same warranty for a used pack versus a new one. If it’s new why are they even worried it would last even a day less than 8 years. Considering they have “claimed” to have corrected earlier issues yet you constantly see newer cars for sale that have had packs replaced. Tesla ransomwaring their batteries seems to be the complete opposite end of the spectrum from sustainability. No one in their right mind would pay 132,000 for a car and then scrap it after 8 years especially if it had low miles. Tesla needs to rethink this fast but I doubt they have the common sense to do so.

 

[father_of_6]

Something to remember is that in the Model S packs each cell in a group that get disconnected would reduce pack capacity by ~1.5%.

That sounds okay... $15k for new pack vs. lose ~1.5% of range? Easy decision.

(While putting more stress on the remaining cells in the group.)

This does *not* sound okay. Just to reiterate, I'm certainly not going to solve this problem... but would love to learn more.

Why would this put more stress on the other cells? Or can you point me at a lamen's resource for understanding the cell grouping / pack cell structure that would explain this?

Or is it as simple as... my TV remote needs 2AA batteries... and is wired either in series or parallel which gives it 3v (1.5v x 2) or 1.5 volts at twice the amperage respectively. Take one out and the remaining battery would have to provide more voltage than it's capable of, or would only provide half the capacity that the remote is intended to have.

Still (yes, I'm stubborn ), there has to be some solution that would allow a cell or group of cells to be disabled without compromising the other cells and/or long term health of the battery pack and vehicle.

With a gas engine, there are so many parts with specialized functions, but our battery packs are just rows and rows and rows of redundancy. There *must* be a way of creating a battery where each of the thousands of cells is not a single point of failure.

 

[MP3Mike]

Or is it as simple as... my TV remote needs 2AA batteries... and is wired either in series or parallel which gives it 3v (1.5v x 2) or 1.5 volts at twice the amperage respectively. Take one out and the remaining battery would have to provide more voltage than it's capable of, or would only provide half the capacity that the remote is intended to have.

You got it.

85/90 packs:
Group: 74 cells in parallel
Pack: 96 groups in series

 

[father_of_6]

You got it.

85/90 packs:
Group: 74 cells in parallel
Pack: 96 groups in series

Interesting. But still... a taser can issue a 300k volt shock with a 9-volt battery using a step-up transformer. My understanding is that the power source can provide any ratio of voltage to current through a bit of manipulation. It's power is still finite, but the voltage can be increased at the expense of the current or vice versa.

Why does one bad cell have to kill the whole pack? I won't give up on the notion that a bad cell can somehow be managed to allow the remainder of the pack to continue functioning.

Why would losing one cell create added stress on the remainder? This has to be a solvable problem.

A TV remote likely needs 2AA because of voltage, not current... otherwise they could just let you use it with 1 battery (which you would have to replace twice as often, but who cares?).

Why can't an EV battery just keep the voltage where it needs to be and let the amperage reduce as cells die off?

Or why can't individual cells be replaced? Certainly they can be. The issue is that this isn't a priority right now for Tesla or legacy auto... but the time will come when someone decides to focus on this problem and they'll solve it, and the rest of the industry will follow. The BMS will just "delete" dead cells when they happen from time to time. This will give battery packs *vastly* more longevity.

 

[brainhouston]

However, difference between reman and new pack from Tesla is $5000.

Its actually really amazing in a sense that new pack is only $5k more or in other words, reman pack is actually expensive compared to new..
Best guess is Tesla perfected manufacturing of cells but to open old packs n refurbish them (not actually doing anything to modules themselves) is quite expensive, cause its all the labor of opening/re-sealing it.
I do hope new packs will keep dropping in price as cells production/development keeps going...

With an EV battery, there are thousands of the exact same individual cells. I just refuse to believe that in the long term, someone won't figure out a solution to bypass a single cell instead of replacing the whole pack.

Perhaps the BMS could tell the techs which cell(s) are failing, the pop the bottom plate off the car, access those cells by coordinate and swap them with fresh ones. OR the BMS just learns to stop using them. Whatever that takes - complex switching inside of each module perhaps?

I'll explain more of technical side.
With 96s 74p, bms only sees data for 96 bricks n can't distinguish each individual cell.
For starters, every (all 7k+) cell will need to be somehow connected to bms to even remotely make it possible, so u can imagine the cost increase vs benefit.
Tesla decides current configuration is good enough since it roughly can last 8yrs, i don't blame them

Swapping cells don't work because unfortunately Li-ion cells change voltage vs SOC
Because Ohms law n 96 in series, current is the same so you'll be "forcing" one new cell to diff voltage/current that it naturally wants to be at
Only way to make series cells work if they all have same charging/discharging properties...

Yes switching would work to cut bad brick out but you'd have to have very big relays n buses (that can handle 1k amps) in between each brick to switch over.
As u can imagine, not practical or cost beneficial..
Hope this helps

Gas engines can be found at any scrap yard and most shops can replace it.
Batteries can be also found at scrap yards more and more, but how many shops can properly do a swap?

Any decent shop can actually replace it, not hard at all, its the software part thats steering ppl away i think
Only if u swap same exact pack u can re-deploy software yourself but if u downgrade/upgrade, u need access to gateway which is not easy...

 

[ucmndd]

Or why can't individual cells be replaced? Certainly they can be.

@brainhouston ’s explanation was quite good. Putting it slightly differently - cells gain resistance as they age. Putting a new cell in with 95 old ones in series and applying a current to charge them will result in that new cell charging more quickly than every other one in the string because of the lower resistance.

This is, obviously, a problem. Overcharging that one cell is a magic smoke releasing event.

People have tried to do module swaps, even going out of their way to painstakingly source modules with similar characteristics. It doesn’t work. The pack ultimately gets too unbalanced for the BMS to handle.

The issue is that this isn't a priority right now for Tesla or legacy auto...

The issue is chemistry and physics limitations of current battery technology. Those are hard problems with real physical limitations. This isn’t a “they just don’t care and would rather sell you a new battery” thing no matter how much we might all want it to be.

 

[walla2]

All the kerfuffle around firmware updates and such are noise. The firmware updates detect failures which earlier firmware did not. That's a safety improvement, and something which I personally would not mess with.

Consider it as a smoke alarm which has been improved to detect more types of smoke. You wouldn't ignore that alarm and say "oh, the old detector didn't notice it, so we'll just put it back", would you?

Microsoft, Apple, Google, Intel, Samsung, NVIDIA, and basically every software company on the planet has had bugs in their software that patches attempt to fix and updates that are periodically needed. It is in fact fact that patches sometimes fix things and sometimes break other things. Tesla isn't impervious to software errors and to totally discount this possibility is naive at best. It is in fact possible firmware updates could cause problems for some cars and is in in fact possible that Tesla can't and doesn't account for in every local firmware, hardware version, and build of every car that they have ever built especially when model year doesn't mean anything. They have built cars the same month that can have entirely different hardware in multiple parts of the car. Do you really believe that such variability can be patched along with one time releases and that every release is perfect?

 

[dmurphy]

Do you really believe that such variability can be patched along with one time releases and that every release is perfect?

Of course not. Even the very best of the very best software program (Great article BTW: https://www.fastcompany.com/28121/they-write-right-stuff) has had bugs.

But there's no conspiracy here either -- the folks running around, screaming that Tesla is generating those error messages intentionally "because moniez" or "they want to screw us into buying a new car" are just plain ignorant.

In this specific case, it's a safety change. And the folks here who do a great job reverse engineering these changes know what those features are, what the BMS is detecting, and why. @wk057 did a fantastic write up on it.

This isn't a bug, it's quite intentional, and there's a good reason for it.

Tesla may have learned a thing or two about failure scenarios over the last dozen years, and they're protecting the cars from them by updating the BMS.

I sympathize with the folks whose packs are having these error messages, but simply, it's a failing pack. Mostly in a way that wasn't originally accounted for, but is now.

 

[JVL]

Of course not. Even the very best of the very best software program (Great article BTW: https://www.fastcompany.com/28121/they-write-right-stuff) has had bugs.

But there's no conspiracy here either -- the folks running around, screaming that Tesla is generating those error messages intentionally "because moniez" or "they want to screw us into buying a new car" are just plain ignorant.

In this specific case, it's a safety change. And the folks here who do a great job reverse engineering these changes know what those features are, what the BMS is detecting, and why. @wk057 did a fantastic write up on it.

This isn't a bug, it's quite intentional, and there's a good reason for it.

Tesla may have learned a thing or two about failure scenarios over the last dozen years, and they're protecting the cars from them by updating the BMS.

I sympathize with the folks whose packs are having these error messages, but simply, it's a failing pack. Mostly in a way that wasn't originally accounted for, but is now.

Tesla's problem is not the reality. Tesla's problem is the perception.

Conspiracy theories grow in the dark. In this case, it is very easy for masses of people to think there is something sinister going on. These kinds of perceptions are battled on the corporate / consumer level every day. Tesla is sticking their head in the sand, while offering very little viable options for the consumers who built the brand.

Seems like bad business to me.

 

[dmurphy]

Tesla's problem is not the reality. Tesla's problem is the perception.

Conspiracy theories grow in the dark. In this case, it is very easy for masses of people to think there is something sinister going on. These kinds of perceptions are battled on the corporate / consumer level every day. Tesla is sticking their head in the sand, while offering very little viable options for the consumers who built the brand.

Seems like bad business to me.

Agree completely.

Could be that a PR department would solve for such things.

They have a communications problem that needs to be solved desperately.

They used to excel at that - honest, open discussions of the good stuff, and the bad.

Now it’s all twitspace, and that’s not healthy.

 

[father_of_6]

@brainhouston and @ucmndd - thank you both for those excellent posts.

Pushing forward in ignorance and stubbornness that this can be done, why not replace one of the parallel units instead?

If the series of 96 cells must remain congruent because of resistance, can't 74 units of them wired in parallel have just one unit replaced or disabled?

That would still bring the replacement cost down to $15k / 74, which is completely within the realm of acceptable.

I got rid of my last car because of leaking engine gaskets, and chose to trade that car in for my M3 instead of spending $2500 on an 8 year old, 110k mile Chevy (no regrets).

A few thousand dollars, in my opinion, is at the absolute upper limit that most people are willing to spend on a vehicle repair. Most people much lower even. Currently, replacing the battery of an EV far exceeds what is economically feasible for most people. That has to change.

Do you/we expect that the economics of pack replacement will change as the scale of battery production increases? For instance, in 10 years when no more ICE vehicles are being produced, will the cost of a replacement pack be down to $2500?

And how, then, is Tesla able to offer a "remanufactured" battery if cells can't be replaced?

 

[MP3Mike]

Pushing forward in ignorance and stubbornness that this can be done, why not replace one of the parallel units instead?

If the series of 96 cells must remain congruent because of resistance, can't 74 units of them wired in parallel have one unit replaced or just disabled?

Tesla used to replace modules to refurb packs, but they found out even they can't get them to match close enough for it to be reliable long term:

https://twitter.com/x/status/1547339152215130112

That would still bring the replacement cost down to $15k / 74, which is completely within the realm of acceptable.

Besides that doesn't work, cost doesn't work that way. There are fixed costs, say $3k just for opening, testing, and resealing the pack.

Do you/we expect that the economics of pack replacement will change as the scale of battery production increases? For instance, in 10 years when no more ICE vehicles are being produced, will the cost of a replacement pack be down to $2500?

No. Elon estimated that it would cost $5-7k to replace all of the modules in a Model 3 pack when they finally started wearing/aging out. (Which was probably referring to the 2026/2027 time-frame.)

And how, then, is Tesla able to offer a "remanufactured" battery if cells can't be replaced?

They can fix things other than cell failure. Say a BMB board failed, they could replace that. Or if contactors fail, etc. (There are some parts that can be replaced inside for some failure modes, but not with a BMS_w029 error.

 

[gt2690b]

.

As it stands currently a well designed and well maintained engine will outlast any EV battery. That has to change for EVs to be successful in the long term.

This is not been my experience.. even my Prius engine began burning oil and had water pump problems at 220k miles.. my S was better than new at 150k miles

 

[father_of_6]

Besides that doesn't work, cost doesn't work that way. There are fixed costs, say $3k just for opening, testing, and resealing the pack.

Understood, but those fixed costs can be significantly reduced over time. A gas engine is such a beast of mechanicals that it takes a cherry picker and 2 mechanics to replace (maybe not, but you get the point). We're already at the point where battery swapping stations can get an EV back on the road in 5 minutes.

A machine could be made to swap dead modules in very little time, if that were the goal. Or, once again, if we can just get the BMS to excommunicate a cell or group of cells, pack failure due to a single cell could be a thing of the past.

Elon estimated that it would cost $5-7k to replace all of the modules in a Model 3 pack when they finally started wearing/aging out. (Which was probably referring to the 2026/2027 time-frame.)

That's just too high, in my opinion.

 

[MP3Mike]

We're already at the point where battery swapping stations can get an EV back on the road in 5 minutes.

Only for vehicles designed to do that. Which at this point is just NIO. (And the swap seems to be close to 6-7 minutes, assuming there is nobody in front of you at the single swap station.)

A machine could be made to swap dead modules in very little time, if that were the goal. Or, once again, if we can just get the BMS to excommunicate a cell or group of cells, pack failure due to a single cell could be a thing of the past.

As I said before: you can keep dreaming. It might be possible in the future, but I think only if some completely new type of battery cell/formulation was invented. Technically someone could probably design a pack to do that, but would you pay ~$7-10k extra for that? And with that would come a lot more complexity, which would add additional new failure modes. So you would likely have to repair the pack much more often.

Doesn't it seem odd to you that none of the companies making EVs have designed a pack to withstand a single cell failure? (I'm sure they have all researched it, and likely determined that it wasn't worth the extra effort/expensive and additional failure modes.)

That's just too high, in my opinion.

You probably need to analyze your opinion to see why it is so far off from reality. Just look at what used modules from Model S packs sell for: generally more than $1,000 each, and there are 16 of them in a pack. So if you have a pack with two failed modules in it you can still sell the 14 remaining good ones for $14k+. Batteries are expensive, and while the costs have come down, and will continue to come down, we are still a long way from even that $5-7k module replacement prediction.

The majority of the cost of an EV is the battery pack, and I don't see anyone lining up to give stuff away for less than it costs. (Well, a lot of OEMs are selling whole EVs for less than it cost them to make them, but they are hoping to turn that around and actually make money at some point.)

 

[Crutonius]

This is not been my experience.. even my Prius engine began burning oil and had water pump problems at 220k miles.. my S was better than new at 150k miles

That sort of supports my point. A Google search tells me that replacing the water pump in a prius is about $900. The oil burning problem could potentially be something you might as well get a new engine to fix but it's also something you can live with. It's not going to leave you stranded.

Will an S reach 220k miles with one minor fix and one persistent annoyance? I wish I had the data to say for sure but I doubt it.