Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

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

beatle

Active Member
Aug 31, 2019
1,024
493
Springfield, VA
But you don't know if you're in a sub-population or not. You might be and not know it since Tesla isn't sharing failure rates among the different pack variants. In that case, the likelihood of failure may indeed be greater than if you're part of gen-pop and you have a normal random failure rate. If you assume the latter is true, yes, the failure rate does not go up, it remains the same. But you generally don't know whether you have a pack of duds or just a couple - that is unless you have a known problem-child pack, and then you have more confidence that its days are numbered.

The 85/90kwh reference is very relevant. Just because the failure mode was cell chemistry related doesn't mean that's the only way a pack can fail. It only means that not all failures are completely random anomalies; there are some packs that are predisposed to... be disposed.
 
  • Disagree
Reactions: MP3Mike and Rocky_H

No2DinosaurFuel

Active Member
Apr 16, 2015
1,383
712
San Diego, California
This way that you are thinking and talking about is the opposite of how statistics for rare things work, though. You are talking about this as if the fact that a rare thing happened here makes it MORE likely that another one will happen in the same place. But that's not how that works. Let's say it's some half a percent: 0.005 chance. So what are the chances that there will be two of them happen TO YOU? Well, it's the half a percent and times another half a percent again. So that is 0.005 squared, or 0.000025. Now what about the chances that you have three of them all happen to you in your pack? That is now to the third power, or 0.000000125 chance.

That's how the statistics work. The chances that you get hit with a rare thing once, are kinda rare. The chances that you will repeatedly keep getting hit with that same rare thing repeatedly get vanishingly small for those rare chances to keep compounding over and over on the same person.
While I complete agree with you on your analysis, you are incorrectly applying it in this situation. Remember how the individual cells work. There is a certain probability that ONE cell will fail and that probability increase as it ages. A good analogy to this is the lottery system. It is exceedingly rare for you to pick all the right numbers, but if you kept playing the lottery for millions of years. The probability of you winning once gets closer to 100%. So let's all agree on that.

Since Gurber isolates and fix that ONE cell, they do not replace any other cells with new ones. So the probability of any OTHER cells in the pack failing continues to increase individually. Whole pack failure probability is determined by ANY ONE cell failing. Hence the likelihood of another failure will continue to go up even after you isolate the previous failed cell. The model S have 7104 cells in the 85KWh pack according to google. Again analogy is you have 7104 people paying the lottery. It is exceedingly rare for anyone of them to win the lottery. But if they play long enough, eventually one will win. Assuming that one person leave (Gurber isolate and removed the offending cell). This has no effect on the other people's chances of winning. The more they play, the higher chances someone else will win.

I hope this helps.
 
Last edited:

Rocky_H

Well-Known Member
Feb 19, 2015
6,097
7,064
Boise, ID
But you don't know if you're in a sub-population or not .You might be and not know it since Tesla isn't sharing failure rates among the different pack variants.
I never said you had to know or not, but that doesn't matter. It's still incredibly unlikely.
In that case, the likelihood of failure may indeed be greater than if you're part of gen-pop and you have a normal random failure rate. If you assume the latter is true, yes, the failure rate does not go up, it remains the same. But you generally don't know whether you have a pack of duds or just a couple - that is unless you have a known problem-child pack, and then you have more confidence that its days are numbered.
But I already just gave examples with horrifically higher failure rates like 4X or 10X a reasonable level, and it STILL is incredibly unlikely for these repeated failure rates to hit the same pack. I don't know what you're going on about.
The 85/90kwh reference is very relevant. Just because the failure mode was cell chemistry related doesn't mean that's the only way a pack can fail. It only means that not all failures are completely random anomalies; there are some packs that are predisposed to... be disposed.
Sigh. Decreased performance is not a "fail". An actual cell failure is a "fail", where the car stops operating, and that is the kind of thing that would need to be fixed.
 

Rocky_H

Well-Known Member
Feb 19, 2015
6,097
7,064
Boise, ID
I hope this helps.
I know what you are saying, but I still don't think it has the effect you two are thinking it does. These aren't like 50% or 75% failure rates. These are like single digit. It is still going to be very unlikely for repeat failures to hit the same place.

You all keep beating on this, trying to convince me that this one thing is a higher chance than that other thing, so people should be terrified of it. And I'm pointing out that one thing being higher than another does not change the fact that they are still both incredibly low likelihood.
 

No2DinosaurFuel

Active Member
Apr 16, 2015
1,383
712
San Diego, California
Sure, and things may vary like that in sub-populations, where if the population as a whole has a half % failure rate, maybe some groups are more or less than that, but still, if you are computing this to the 2nd and 3rd and 4th power, it gets less and less likely. The chances do not go toward likely with 2nd and 3rd repeat events!

So let's look at an example like that. Let's say you had a totally awful "junk" batch that has 4X the normal failure rate of this half percent realistic level. That would be 2%. That's still a 0.0004 chance it would happen to the same person twice. What if they were insanely awful, with a failure rate 10X the norm, of 5% ? That's a 0.0025 chance it would happen to the same person twice. Unless you're talking of out of this world stuff, like well over half of things failing, the compounding just keeps it from being likely for repeat failures to happen to the same person.

Irrelevant. Those were issues with degraded capacity and performance because of cell chemistry issues, not as much outright failure rates of cells going dead.
The closet thing I can think of to your probability example is what is the probability of 2 cells within the pack failing at the SAME time. And yes that would grow exceedingly rare if you try to get the probability of 3 or 4 or more cells failing all at the SAME time.

Stat is the toughest Math subject IMO. The way I like to think about Stat problems it is if the answer doesn't make sense, then most likely you are not applying it right.
 

No2DinosaurFuel

Active Member
Apr 16, 2015
1,383
712
San Diego, California
I know what you are saying, but I still don't think it has the effect you two are thinking it does. These aren't like 50% or 75% failure rates. These are like single digit. It is still going to be very unlikely for repeat failures to hit the same place.

You all keep beating on this, trying to convince me that this one thing is a higher chance than that other thing, so people should be terrified of it. And I'm pointing out that one thing being higher than another does not change the fact that they are still both incredibly low likelihood.
Correct not the same cell. Because Gerber has isolate it and thus the probably of that cell happening is 0 because it doesn't affect the pack any longer.

On your 2nd point. All I am saying as the pack ages, you will logically get a higher chance of failure regardless if you repair the pack by isolating the cell. After repairs doesn't reset your probability of whole pack failure to when you got your car new. The only way to reset the probability is to get a NEW pack.
 

Rocky_H

Well-Known Member
Feb 19, 2015
6,097
7,064
Boise, ID
Correct not the same cell.
No, when I said "same place", I obviously did not mean that cell, which is disconnected. I meant like that same person's car's battery pack.
The closet thing I can think of to your probability example is what is the probability of 2 cells within the pack failing at the SAME time. And yes that would grow exceedingly rare if you try to get the probability of 3 or 4 or more cells failing all at the SAME time.
Yes! Exactly! And I get the statement that you're trying to make, that as time goes by, the chances will increase. Of course--that is a fact. But that does not mean it is becoming actually what people consider "likely", like near certain, by a 20 year point, that there will be repeated failures. That will still be very unlikely.
Stat is the toughest Math subject IMO. The way I like to think about Stat problems it is if the answer doesn't make sense, then most likely you are not applying it right.
Yeah, which is why it is confusing me why you all are pushing that perspective. It doesn't make any sense. These are like single digit failure percentages. Compounding it for a 2nd and 3rd event, even over time, produces extremely tiny chances of it happening. This idea that it's going to become more likely than not and is therefore something to fear doesn't make sense.
 
  • Like
Reactions: SmartElectric

beatle

Active Member
Aug 31, 2019
1,024
493
Springfield, VA
You all keep beating on this, trying to convince me that this one thing is a higher chance than that other thing, so people should be terrified of it. And I'm pointing out that one thing being higher than another does not change the fact that they are still both incredibly low likelihood.
You're putting words in people's mouths. Nobody here is saying that if a single cell fails that the rest of the pack is right behind it and we're better off pushing the car into the ocean because it's a lost cause. There is not enough data backing up those failures, though there are cases where that does play out. In an old system, the failure rate starts to go up, sometimes dramatically, but again, we don't have enough data to create a good curve yet.

However, to say that the likelihood of a second failure is less likely to occur because you've already had one failure is also false.
 

SO16

Active Member
Feb 25, 2016
2,864
8,791
MI
A Gruber repair is a gamble. Would love to know from Gruber if they have had any repeat failures.
 

CyberGus

Not Just a Member
May 5, 2020
820
1,771
Austin, TX
So much for "Building a Sustainable Future"...as least as far as externalized costs goes.
easy fix

Kenyas-diesel-generator-powered-car-1210x642.jpeg
 
  • Disagree
  • Funny
Reactions: rxlawdude and SO16

No2DinosaurFuel

Active Member
Apr 16, 2015
1,383
712
San Diego, California
No, when I said "same place", I obviously did not mean that cell, which is disconnected. I meant like that same person's car's battery pack.

Yes! Exactly! And I get the statement that you're trying to make, that as time goes by, the chances will increase. Of course--that is a fact. But that does not mean it is becoming actually what people consider "likely", like near certain, by a 20 year point, that there will be repeated failures. That will still be very unlikely.

Yeah, which is why it is confusing me why you all are pushing that perspective. It doesn't make any sense. These are like single digit failure percentages. Compounding it for a 2nd and 3rd event, even over time, produces extremely tiny chances of it happening. This idea that it's going to become more likely than not and is therefore something to fear doesn't make sense.
You are still not getting my point. When I say SAME I mean literally at the same point in time. i.e. what is the probability cell number 1200 AND cell number 5492 fails exactly on March 24th 2021 16:45GMT. That is when you can compound. The SAME you are thinking of is the same type of failure. They are 2 different things.

However, to say that the likelihood of a second failure is less likely to occur because you've already had one failure is also false.
Beatle hit it home. Just because it happened and you repaired it doesn't mean you are less likely to see another one. This explains why you see some people getting multiple pack replacements from Tesla. Tesla is probably doing a little better job at pack refurbishing because I think, but can't confirm, they are probably replacing the offending module vs just isolating the cells. If so, then you are getting some new cells there. But if they are doing what Gurber is doing, then yeah it doesn't really help people that much down the road.

A Gruber repair is a gamble. Would love to know from Gruber if they have had any repeat failures.
Agreed with this. It seems to be a temporary thing so people can keep driving, but in the end, I wouldn't be surprised if this method cost you more than a whole pack replacement. If you trust Gurber's numbers then you will be expecting to see them often. And each time it will be $5K. After the 4th, you would wonder why you just didn't get a whole new pack from Tesla to begin with. Don't get me wrong it is good service they are providing, but this is definitely not a long term solution. As of now, there is little data on this because the early model S are just out of warranty hence why people would turn to them providing them with a data point. We will need to wait a bit longer to see if there will be repeat failures in a different section of the pack.

Like I said before, the only long term via solution is cheaper batteries so the cost is not sticker shocking. Allowing DIY to fix things only pushes the can down the road and greatly increase complexity, logistics, and safety.

The other solution is battery pack rental like what some of the other Chinese/European EV companies are doing. However IMO it is only fair to the customer to subtract the cost of the battery pack from the initial cost of the vehicles. This way if the customer wants to continue to use their car 20 years in, they just need to rent the battery for 20 years. If it fails in the meantime, the car company is on the hook to provide a replacement. This way there is an incentive for the car company to make their battery last multiple decades to save on replacement cost. If the users stops paying, then the company just disable the pack and repo the pack just like repoing a car.
 
Last edited:

Rocky_H

Well-Known Member
Feb 19, 2015
6,097
7,064
Boise, ID
[/QUOTE]
You are still not getting my point.
I absolutely am. I just don't agree with you.
When I say SAME I mean literally at the same point in time. i.e. what is the probability cell number 1200 AND cell number 5492 fails exactly on March 24th 2021 16:45GMT. That is when you can compound.
That is not the only time you can compound. The numbers are not exactly, precisely identical because of a little more time shifting, but they will be extremely close. If that's what you are nitpicking on, sure, I shouldn't have said that it's just raising the exponent, but the math is nearly the same. For a person to have two of these same type of failures in different cells within a few years is still an extremely statistically unlikely thing, as I have been saying.
The SAME you are thinking of is the same type of failure. They are 2 different things.
Yes, they are rare things. And for someone to get double hit with these rare things is even more rare.
Beatle hit it home. Just because it happened and you repaired it doesn't mean you are less likely to see another one.
Yeah, I know that. That's a different psychological fallacy, where people think the odds go down and aren't the same as they were because they "got theirs done". And I have probably not been expressing this well. I do not at all mean that once people have one of these happen to them, that somehow changes the probabilities for them. And sorry if I have been confusing. But I mean that as a starting proposition, before any of it happens, the odds of this ever happening to the same person's pack more than once is extremely, outlandishly rare. And that is true, because that is where the compounding would be happening. You would have to hit that same roulette thing twice in a fairly short time, and that's just statistically very unlikely.
Agreed with this. It seems to be a temporary thing so people can keep driving, but in the end, I wouldn't be surprised if this method cost you more than a whole pack replacement. If you trust Gurber's numbers then you will be expecting to see them often. And each time it will be $5K. After the 4th, you would wonder why you just didn't get a whole new pack from Tesla to begin with.
Eh? If you're trying to keep the car going for a century, yeah... Eventually things are going to wear completely out. But the context of this was basically to get through from just past warranty ends to kind of a normal long life of a car, like maybe 20-25 years. But the fact you still think this is going to have many cell failures, like a 3rd and 4th in anyone's pack just after the warranty expires is just boggling my mind.
 
  • Like
Reactions: SmartElectric

SmartElectric

Active Member
Jul 9, 2014
2,426
2,060
Toronto,Canada
If only the bad cell could be auto-isolated.
Then you'd need to switch to a pack design with a fraction of the range such that individual cells (and few of them presumably) were packaged in larger number of modules which could auto-detect and isolate.

Hmm, the 2011 Nissan Leaf pack is described here:

48 modules <- many more modules,
192 cells (4 cells per module) <- less cells per module

^ 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.

Seems to me your "if only" translates to far more costly pack per kWh, far less range (lower pack density), similar complexity to ultimately fix, even if the pack cut off a portion of the capacity to isolate.

What IF : In the case of the Leaf, if a pack was designed to isolate a single module (due to a bad cell in the module) in a pack with ~50 modules loses you 2% of range for each "isolation".
 
  • Informative
Reactions: SO16

beatle

Active Member
Aug 31, 2019
1,024
493
Springfield, VA
The strange thing is that cells CAN self-isolate by blowing the built in fuse that links them to the bus. I'd bet that is the cause for some degradation. But in some cases that fuse doesn't blow for some reason and the car instead shorts through that cell and faults.
 

aerodyne

Active Member
Nov 19, 2018
2,186
1,628
Los Angeles
So, Gruber says a single cell can go resistive, and bring down the whole string.

I'm guessing that at some point the BMS will shut down the module or the car entirely.

But if it is resistive, does not that mean little current, and thereby little heat? Even if there was a severe voltage drop across the cell, how is that much different from an open circuit across that cell as far as the string goes?

(1/74) * 3.5vpc = 47 mv. I see that quite often driving.

I am sure Gruber will get things sorted out with time and increasing business, and be able to give more guidance on proactive pack repair
 

No2DinosaurFuel

Active Member
Apr 16, 2015
1,383
712
San Diego, California
So, Gruber says a single cell can go resistive, and bring down the whole string.

I'm guessing that at some point the BMS will shut down the module or the car entirely.

But if it is resistive, does not that mean little current, and thereby little heat? Even if there was a severe voltage drop across the cell, how is that much different from an open circuit across that cell as far as the string goes?

(1/74) * 3.5vpc = 47 mv. I see that quite often driving.

I am sure Gruber will get things sorted out with time and increasing business, and be able to give more guidance on proactive pack repair
Here is an experiment you can try to get around the idea of resistive element in parallel with the cells. Just get a high value resistor and put it in parallel with a 18650 cell. The battery will eventually drain. This is culprit of a resistive element cell. It is constantly draining the battery through it even if it's not used. And yes this will mean heat from the resistive element especially when you drain quite a bit and as we all know heat is no good for lithium cells. This will affect the cells around it so if the BMS detects it, I suspect it will try to prematurely shutdown the module to protect the other cells. This is different from open circuit in that a open circuit will not drain the cells around it. Open circuit by definition has no current through it so not battery drain. This is essentially what Gruber does to the pack. And what the internal individual fuse is suppose to do. But if there isn't enough current through the cells to heat up and blow the fuse, the battery continues to work until it doesn't. I guess for experimental purpose, someone with battery issue but still can drive it should heat up the pack via supercharging in the summer and then do multiple hard accelerations in hopes to blow the fuse on the offending cell and maybe return it back semi-normal operation.
 

David_Cary

Active Member
Dec 17, 2012
1,182
672
Cary, NC
I suspect the Gruber repair is something that you do before selling the car. No amount of statistical talk would convince me that something else wouldn't happen at some point in a relatively quick time. Whether it is a bad batch or a hot spot on the battery or whatever. And I have taken a few college statistical classes and make decisions with less emotion than the average person. For example, no extra warranty ever. This a minority opinion on TMC.

So what do you think average owner would do when faced with a car with a $20k value and a dead battery? It has a $5k repair that has zero warranty to get it up and running again. Doesn't really matter if the proverbial lightning doesn't strike twice. That car is sold for $20k within 3 months. Particularly given that Tesla may charge $35k to replace the battery (ie no core).

When that Gruber repaired car gets sold to an unsuspecting buyer, hopefully the battery lasts a good amount of time. But when stories come out that it doesn't, the $20k for an older Tesla is a pipe dream. Can you imagine the angst when a battery dies a year later and Tesla says - nope - we won't accept the core? After taking it apart and charging you for it.

Heck, I have 2 years left on my battery warranty but worry about resale value 2 years down the line. My car is only worth $25k now (to my estimation) so getting down to $15k in 2 years isn't bad. And then rolling the dice is probably right. But no way does my car get a $5k repair after the battery warranty, it gets salvaged. Now a $10k Tesla battery with warranty? Sure. But not a $20k one unless it was 8 years and that included newish charging speed and degradation warranty.

The fact that Tesla won't warranty a $20k battery beyond 4 years and presumably no charging speed and degradation warranty speaks volumes to what they think the lifespan will be after reman. And that Gruber doesn't warranty at all, also speaks volumes. Gruber should at least sell warranties at $1k a year or something reasonable.

Does anyone remember when Elon talked about buying a battery 8 years down the road for $8k? And then getting 8 more years of warranty? How far we have fallen.... (understand that I might have those details a bit off and it did involve buying the battery preemptively so time value of money and all that)
 
Last edited:
  • Like
Reactions: aerodyne

jpvdheijn

Member
Oct 30, 2018
514
391
Baarle-Nassau, NL
I suspect the Gruber repair is something that you do before selling the car. No amount of statistical talk would convince me that something else wouldn't happen at some point in a relatively quick time. Whether it is a bad batch or a hot spot on the battery or whatever. And I have taken a few college statistical classes and make decisions with less emotion than the average person. For example, no extra warranty ever. This a minority opinion on TMC.

So what do you think average owner would do when faced with a car with a $20k value and a dead battery? It has a $5k repair that has zero warranty to get it up and running again. Doesn't really matter if the proverbial lightning doesn't strike twice. That car is sold for $20k within 3 months. Particularly given that Tesla may charge $35k to replace the battery (ie no core).

When that Gruber repaired car gets sold to an unsuspecting buyer, hopefully the battery lasts a good amount of time. But when stories come out that it doesn't, the $20k for an older Tesla is a pipe dream. Can you imagine the angst when a battery dies a year later and Tesla says - nope - we won't accept the core? After taking it apart and charging you for it.

Heck, I have 2 years left on my battery warranty but worry about resale value 2 years down the line. My car is only worth $25k now (to my estimation) so getting down to $15k in 2 years isn't bad. And then rolling the dice is probably right. But no way does my car get a $5k repair after the battery warranty, it gets salvaged. Now a $10k Tesla battery with warranty? Sure. But not a $20k one unless it was 8 years and that included newish charging speed and degradation warranty.

The fact that Tesla won't warranty a $20k battery beyond 4 years and presumably no charging speed and degradation warranty speaks volumes to what they think the lifespan will be after reman. And that Gruber doesn't warranty at all, also speaks volumes. Gruber should at least sell warranties at $1k a year or something reasonable.

Does anyone remember when Elon talked about buying a battery 8 years down the road for $8k? And then getting 8 more years of warranty? How far we have fallen.... (understand that I might have those details a bit off and it did involve buying the battery preemptively so time value of money and all that)
 

About Us

Formed in 2006, Tesla Motors Club (TMC) was the first independent online Tesla community. Today it remains the largest and most dynamic community of Tesla enthusiasts. Learn more.

Do you value your experience at TMC? Consider becoming a Supporting Member of Tesla Motors Club. As a thank you for your contribution, you'll get nearly no ads in the Community and Groups sections. Additional perks are available depending on the level of contribution. Please visit the Account Upgrades page for more details.


SUPPORT TMC
Top