How long will our batteries last?

[ExCivilian]

Where do you get the facts that that only 1/2 are on the road. There are many Tesla's on the road with 200,000+ miles and the batteries are fine if you take care of them.

The discussion was about Roadsters specifically. If you want to expand the conversation to all Teslas generally that's fine but it's a different conversation. I don't think you'd want to do that, though, because those data don't support your point. While there may exist "many" Teslas on the road with over 200,000 miles it seems like you haven't looked up how many there are or how many there are relative to the Teslas that have experienced failures and can't make any other conclusions other than your personal opinion that it seems like a lot. The research on this that is emerging indicates somewhere around a ten-year failure rate at higher than expected numbers, which makes sense since everyone here can hopefully understand that manufacturers aren't going to spend extra resources to ensure batteries remain viable much longer than their warrantied period.

Also, Ford designed the battery in separate packs and they replace individual modules all the time with no issues. The older Tesla’s have to be replaced as a full battery and the new Tesla’s with the battery as part of the frame, good luck.

If Ford is replacing battery modules "all the time with no issues" that's a huge problem for the "batteries have long lives no worries here" crowd for what should be obvious reasons...regardless, the modules they are replacing and the packs they are placing them in are relatively close to one another in age and few of us in this thread know the process the techs use to install them. What little I know about the process leads me to conclude it isn't as simple as opening the pack and dropping new cells in--that would lead to cell imbalance, which anyone interested in the facts would have looked up by now instead of arguing with me. That said, it's important to note we aren't discussing in-warranty failures since those EVs were deliberately excluded from the cited studies. If we include them, the failure rate becomes much greater than 30%.

The greatest damage to batteries occurs when folks charge to 100% all the time and leave it sit charged at 100%. Also charging at superchargers is very hard on the batteries due to how fast they charge and the heat generated. Even Tesla recommends folks to not use superchargers as their only charging method but unfortunately their sales folks sell them to folks who don't have access to home chargers, IE: apartments and etc. Even my 6 month old Samsung phone instructions recommends charging to 80% in order to make the batteries last.

While the greatest damage to batteries occurs because of heat, in a conversation about using batteries over forty years the more relevant factor becomes time because one can control all the variables you described and the battery will still degrade over time at a roughly consistent rate and isn't damaged much more or less than a battery that didn't control those variables over the same time frame. This conversation took place in the context of an in-service battery over 40 years, which would make the time variable much more salient than the state of charge variable. Additionally, the reason those Teslas you referenced with viable batteries after 200,000 miles were in the news was because they had utilized the fast charging network frequently and they were used to demonstrate that fast charging, done correctly, does not damage batteries as initially feared. Again, this technology is not esoteric and the principles are well-known and documented.

Again, I suggest you conduct some rudimentary research before disputing this point. Now, I've done my research before I posted my points and my opinion is that you are answering off-the-cuff because the points I've raised are well-documented and generally well-known already. If you want to hold your opinions in spite of these data that's fine but you and I both know you haven't researched your opinion before disputing what I wrote.

Sponsored

 

[husky10101]

The discussion was about Roadsters specifically. If you want to expand the conversation to all Teslas generally that's fine but it's a different conversation. I also don't think you'd want to do that because those data don't support your point. While there may exist "many" Teslas on the road with over 200,000 miles you haven't looked up how many there are or how many relative to the Teslas that have experienced failures and can't make any other conclusions other than your personal opinion that there it seems like a lot. The research on this that is emerging indicates somewhere around a ten-year failure rate at higher than expected numbers, which makes sense since everyone here can hopefully understand that manufacturers aren't going to spend extra resources to ensure batteries remain viable much longer than their warrantied period.


Well, if Ford is replacing battery modules "all the time with no issues" that's a huge problem for the "batteries have long lives no worries here" crowd for what should be obvious reasons...regardless, the modules they are replacing and the packs they are placing them in are relatively close to one another in age and few of us in this thread know the process the techs use to install them. What little I know about the process leads me to conclude it isn't as simple as opening the pack and dropping new cells in--that would lead to cell imbalance, which anyone interested in the facts would have looked up by now instead of arguing with me.


While the greatest damage to batteries occurs because of heat, in a conversation about using batteries over forty years the more relevant factor becomes time because one can control all the variables you described and the battery will still degrade over time at a roughly consistent rate and isn't damaged much more or less than a battery that didn't control those variables over the same time frame. This conversation took place in the context of an in-service battery over 40 years, which would make the time variable much more salient than the state of charge variable, for example. Additionally, the reason those Teslas were in the news with viable batteries after 200,000 miles was because they had utilized the fast charging network frequently and the examples were used to demonstrate that fast charging, done correctly, does not damage batteries as was feared. Again, this technology is not esoteric and the principles are well-known and documented.

Again, I suggest you conduct some rudimentary research before disputing this point. Now, I've done my research before I posted my points and my opinion is that you are answering off-the-cuff because the points I've raised are well-documented and generally well-known already. If you want to hold your opinions in spite of these data that's fine but you and I both know you haven't researched your opinion before disputing what I wrote.

By the way you aren't the only one here doing their own research and I NEVER accused you of not doing yours so please don't accuse me of not doing mine. Also the lithium batteries and the advanced battery management software (BMS) have not been around even close to the 40 years you quoted. You have some very good points, no argument there, but you should open your mind and listen to other peoples points too. So lets let the rest here.

 

[Texas Dan]

I am an expert in statistical analysis and research methodology--but as it pertains to crime and crime rates--not batteries as I already clearly stated.

But generally speaking, without knowing the methodology of the research or the variables involved in the research cells (an area of a spreadsheet where data are entered, not a battery "cell") it's not informative to simply take all of the EVs on the road, divide them by an overall failure rate, and then project a completely different vehicle's failure rate out into the future like you attempted to do.

Just to give some people an example of why this doesn't work, the study you cited actually states that EVs from 2011 to 2015 had a 13% failure rate, not 1.5% average as you claimed in this thread. Additionally, looking at just the 2011s they found the failure rate was over 30%.

One of the problems with this study, and the one conducted last year by the same group, is that last year the failure rate was 1.5% whereas this year it's 2.5%. We could take more at length about why that might be but the bottom line for us, since we're using rough generalities I suppose, is that the failure rate nearly doubled in the span of a single year. Extrapolating into the future like you did would indicate to us that next year this same group using the same methodology would "discover" that EVs actually fail closer to a 4% rate. Also, keep in mind all of these data are done without accounting for warrantied replacements, which I would argue provides an under-reporting of the true failure rate.

There is also a logical problem inherent in how you're presenting these data. On the one hand, you're arguing we can use these data from older vehicles and older battery technology to extrapolate what our current EVs might do in the future based on the premise that newer technology has extended the life of battery packs (the cited study makes this argument explicitly). That, however, would mean that older EV technology is not predictive of what our current batteries might/might not do in the future.

The oldest packs in our vehicles are less than 1.5 years old. They haven't reached a point of wear that would introduce the problems I was referring to in my post. Additionally, neither you nor I have any knowledge about what the techs do when they install a "new" module into the pack. The modules are likely the relative age as the installed pack and the techs probably conduct some kind of wear leveling when they install them.

Like I said, what I wrote is well-known and well-researched already and if you were interested in learning about it instead of arguing your point you'd have looked it up already.

I don’t see life any life expectancy numbers in your response so I must assume that you’re not going to respond to my request. As I suspected, you have no answers and you’re unwilling to analyze statistics. Please do your trolling somewhere else.

 

[ExCivilian]

By the way you aren't the only one here doing their own research and I NEVER accused you of not doing yours so please don't accuse me of not doing mine. Also the lithium batteries and the advanced battery management software (BMS) have not been around even close to the 40 years you quoted. You have some very good points, no argument there, but you should open your mind and listen to other peoples points too. So lets let the rest here.

If you researched the numbers of Teslas on the road relative to out of and in warranty failures, why didn't you post those numbers instead of stating "many are on the road" and that it seems like a "lot" to you?

That's a peculiar way for a researcher to discuss data they've looked up and are citing.

Lithium-Ion technology has been around for over 30 years and battery management systems aren't esoteric--the principles of what it takes to maintain a battery's life are well-known and documented. Vehicle manufacturers aren't doing anything new here...but that's not the point because the 40 years was in reference to the OPs point about the Lightning's longevity after 40 years and nothing to do with anything scientific. In fact, I was disputing that they could make such projections--not using them myself.

And no, this is precisely the problem: some EV proponents make grandiose claims that are unsupported and disputed by the known science and present unrealistic and inaccurate expectations and then try and shut the conversation down (or leave it, I guess) when they're called out on it. It's the same coin as the people who concoct unrealistic negatives about EVs--both sides of the discussion are guilty of this kind of crap analysis and, as I tell my students, garbage in > garage out. Presenting garbage results that are easily researched and disputed simply calls into question the veracity of the rest of what else may be true, which is a disservice to EV adoption.

 

[ExCivilian]

you have no answers and you’re unwilling to analyze statistics. Please do your trolling somewhere else.

I disputed your singular formula and explained why it was incorrect and I also described the ways in which you misrepresented the articles you presented in support of your position.

Also, I did provide a realistic expectation for the Lightning battery pack based on longitudinal studies of average degradation so you can scroll back through my responses to review that if you're genuinely interested in my opinion of that. Calling me a troll is just more of the same bombastic language you've already demonstrated, unprovoked by me I'm going to add, to me so if you really feel I'm trolling and you can't control your tonality please stop responding to me.

 

[husky10101]

If you researched the numbers of Teslas on the road relative to out of and in warranty failures, why didn't you post those numbers instead of stating "many are on the road" and that it seems like a "lot" to you?

That's a peculiar way for a researcher to discuss data they've looked up and are citing.

Lithium-Ion technology has been around for over 30 years and battery management systems aren't esoteric--the principles of what it takes to maintain a battery's life are well-known and documented. Vehicle manufacturers aren't doing anything new here...but that's not the point because the 40 years was in reference to the OPs point about the Lightning's longevity after 40 years and nothing to do with anything scientific. In fact, I was disputing that they could make such projections--not using them myself.

And no, this is precisely the problem: some EV proponents make grandiose claims that are unsupported and disputed by the known science and present unrealistic and inaccurate expectations and then try and shut the conversation down (or leave it, I guess) when they're called out on it. It's the same coin as the people who concoct unrealistic negatives about EVs--both sides of the discussion are guilty of this kind of crap analysis and, as I tell my students, garbage in > garage out. Presenting garbage results that are easily researched and disputed simply calls into question the veracity of the rest of what else may be true, which is a disservice to EV adoption.

If you are so anti EV why are you here? Also where are your facts you are so proud of. If you would open your mind and actually listen to other folks, you might actually learn something rather than sounding like a know it all because none of us know it all. Fact

 

[husky10101]

I disputed your singular formula and explained why it was incorrect and I also described the ways in which you misrepresented the articles you presented in support of your position.

Also, I did provide a realistic expectation for the Lightning battery pack based on longitudinal studies of average degradation so you can scroll back through my responses to review that if you're genuinely interested in my opinion of that. Calling me a troll is just more of the same bombastic language you've already demonstrated, unprovoked by me I'm going to add, to me so if you really feel I'm trolling and you can't control your tonality please stop responding to me.

Agree

 

[lightspeed]

It will be nice in the future when battery modules become standardized. Standard capacities plus electrical and cooling/BMS connections.

Then anyone can source a module from any provider and swapping them out should be relatively easy too.

It will take some time, but it's something the industry needs to move towards.

By the time our trucks need batteries, they should be less than $4K at the OEM level. And there are commercial batteries coming online now that have double the energy for the same weight, so maybe we can get 200kWh upgrades in the future (probably from a 3rd party).

 

[ExCivilian]

If you are so anti EV why are you here? Also where are your facts you are so proud of. If you would open your mind and actually listen to other folks, you might actually learn something rather than sounding like a know it all because none of us know it all. Fact

I'm not anti-EV and I'm here because I own a 2022 Lightning.

I am disputing the misuse and mischaracterization of statistical data to promote EV battery longevity that is in contrast to what is already known and non disputed elsewhere.

The source of these data I've been citing is contained within the articles the OP cited and misrepresented to the board. I've mentioned several times that anyone who is curious enough to read it themselves should do so and I'm confused what you think I would learn from listening to who exactly?

The OP stated his formula predicts that our Lightning batteries should last more than 40 years. I dispute that and then you, for some reason, simultaneously told me that I'm wrong for claiming the batteries would last 40 years (which I didn't say) and then I should be listening to the person who made that inaccurate claim.

The only other claim I made was that "most" Roadsters are not, in fact, still in service and you're welcome to look that up yourself. You then generalized that to all Teslas, which wasn't a claim I made or even addressed until you brought it up, and then asked you to look up your numbers substantiating the claim that "many," "most," or whatever the numbers are, which you refused to do so that's certainly not an example of me not listening to you.

 

[lightspeed]

I'm not anti-EV and I'm here because I own a 2022 Lightning.

I am disputing the misuse and mischaracterization of statistical data to promote EV battery longevity that is in contrast to what is already known and non disputed elsewhere.

The source of these data I've been citing is contained within the articles the OP cited and misrepresented to the board. I've mentioned several times that anyone who is curious enough to read it themselves should do so and I'm confused what you think I would learn from listening to who exactly?

The OP stated his formula predicts that our Lightning batteries should last more than 40 years. I dispute that and then you, for some reason, simultaneously told me that I'm wrong for claiming the batteries would last 40 years (which I didn't say) and then I should be listening to the person who made that inaccurate claim.

The only other claim I made was that "most" Roadsters are not, in fact, still in service and you're welcome to look that up yourself. You then generalized that to all Teslas, which wasn't a claim I made or even addressed until you brought it up, and then asked you to look up your numbers substantiating the claim that "many," "most," or whatever the numbers are, which you refused to do so that's certainly not an example of me not listening to you.

40 years does seem unlikely, plus what duty cycle?

They should be good for 350K+ miles and over time they will become more susceptible to individual cell shorts/issues.

But what difference does it make? Nobody is going to keep their Lightning for 40 years.

 

[husky10101]

I'm not anti-EV and I'm here because I own a 2022 Lightning.

I am disputing the misuse and mischaracterization of statistical data to promote EV battery longevity that is in contrast to what is already known and non disputed elsewhere.

The source of these data I've been citing is contained within the articles the OP cited and misrepresented to the board. I've mentioned several times that anyone who is curious enough to read it themselves should do so and I'm confused what you think I would learn from listening to who exactly?

The OP stated his formula predicts that our Lightning batteries should last more than 40 years. I dispute that and then you, for some reason, simultaneously told me that I'm wrong for claiming the batteries would last 40 years (which I didn't say) and then I should be listening to the person who made that inaccurate claim.

The only other claim I made was that "most" Roadsters are not, in fact, still in service and you're welcome to look that up yourself. You then generalized that to all Teslas, which wasn't a claim I made or even addressed until you brought it up, and then asked you to look up your numbers substantiating the claim that "many," "most," or whatever the numbers are, which you refused to do so that's certainly not an example of me not listening to you.

Also I suggest you call Ford engineers and set them straight regarding all the batteries that they have already replaced one module or so in and have them recall all of those trucks because it sounds like you are much more of an expert than the guys that designed the trucks and batteries. I will put my money and faith in the engineers, not some guy on the internet. Again where are your facts, where are the sources and links to all your info? I am done with you and just like the previous post, don't reply to me and I ask this pose be locked.

 

[ExCivilian]

Also I suggest you call Ford engineers and set them straight regarding all the batteries that they have already replaced one module or so in and have them recall all of those trucks because it sounds like you are much more of an expert than the guys that designed the trucks and batteries. I will put my money and faith in the engineers, not some guy on the internet.

I've already responded to this several times:

The oldest packs in our vehicles are less than 1.5 years old. They haven't reached a point of wear that would introduce the problems I was referring to in my post. Additionally, neither you nor I have any knowledge about what the techs do when they install a "new" module into the pack. The modules are likely the relative age as the installed pack and the techs probably conduct some kind of wear leveling when they install them.

Again where are your facts, where are the sources and links to all your info?

Citations are required for novel or controversial information. It's not novel or controversial of me to point out that wear leveling and voltage levels of lithium-ion cells have to be within spec of one another. I don't know why you're being so argumentative and rude about my point because it should be obvious why that's less of a concern now, with only 1.5 years of Lightnings on the road, than it will be in 8 years after the warranty expires. Furthermore, it's not controversial of me suggest that neither you nor I know how the techs implement new modules into old packs and that it may not be as simple as you're assuming/claiming it to be.

If you have trouble believing any of those points you need to conduct your own research because it's not my responsibility to dig up decades old information and present it to you, who it appears isn't particularly interested in having a civil discussion anyway.

I am done with you and just like the previous post, don't reply to me and I ask this pose be locked.

If you're done with interacting with me, that's your prerogative. I'm not sure why you think the thread should be locked or why you would have the audacity to close the conversation down for others when you can simply disengage. If you review your responses to me they've been snide and belligerent for seemingly no reason--I certainly haven't been addressing you in that manner.

 

[Texas Dan]

I intentionally tried to be vague in my life expectancy analysis at the beginning of this thread. Predicting life expectancy of EV is exceeding complicated with a lot of different variables, so I choose to just use some very simplistic statistical values for life projections. I do not intend to retract my projections now, but I would like to make a few clarifications.

First, there is the phenomenon in EV battery degradation called “Calendar Aging”. Calendar Aging affects EV batteries even if the EV never gets driven. I could not find any information on when Calendar Aging would make an EV battery unusable, but I did find one report that stated Calendar Aging results in 10% capacity loss over a 15-year period.

Another important phenomenon to consider in EV battery degradation is cycling. The Lightning battery packs are good for about 2,000 EFC (equivalent full cycles) which would equate to maybe 600,000 miles for the extended range battery pack and maybe 400,000 miles for the standard range battery pack. And of course, we all know about the other phenomenon that impact battery pack degradation such as battery temperature, fast charging and charging outside of the 20-80 percent range.

The Recurrent Auto report originally cited has test data on over 15,000 EVs that go up to 3500 days or ten years. They have degradation charts that go up to 3500 days. They also have a hand drawn degradation graph that shows an “end of life” point but it’s not clear what is defined as “end of life” and that point is not supported by the test data.

Getting back to my life expectancy analysis, we know that because of Calendar Aging at some point the battery pack is going to unusable but when? In 60 years, Calendar Aging may make the battery pack degrade 40%, will that make the battery pack unusable? Of course, this will mean that the EV never gets driven and what’s the point of having an EV that never gets driven?

We know that Calendar Aging is the outside limiting factor but what about all the other factors? Larger battery packs are going to last longer than smaller battery packs due to EFC. Charging habits will also have an impact on battery pack longevity.

So, considering all these factors I think my life expectancy analysis is reasonable. I mean there is no expiration date on battery packs. Battery packs in EVs that get a lot of use will fail early, before life expectancy and battery packs that in EVs that don’t get used much will fail late, after life expectancy.

What is an EVs battery packs life expectancy or, in other words, when will at least 50% of a model year’s battery packs need to be replaced? Of course, we are just guessing based on historical test data, but researchers agree that it’s a lot longer than most people predicted a few years ago.

Here is the link to the original Recurrent Auto report.

Here is link to another Recurrent Auto report on EV battery degradation.

Here is another link to a good report on EV battery degradation.

 

[ExCivilian]

we are just guessing based on historical test data, but researchers agree that it’s a lot longer than most people predicted a few years ago.

The strength (and weakness) of a statistical model hinges on the assumptions of the researcher and how the variables are input into the model.

First, we deconstruct the historical data presented in the cited articles:

Across all years and models, outside of big recalls, only 2.5% have been replaced. This increase from last year is entirely due to older cars. For cars older than 2015, replacement rates are 13%, but under 1% for cars from 2016 and newer.

Those data make the following claims:

1. Cars that are nine years or older have a higher than average failure rate.
2. Cars that are eight years or newer have a lower than average failure rate.
3. Combined, these two rates result in an approximate failure rate of 2.5% across the 20,000 EVs in the Recurrent community (not the general public).

The way I interpreted your analysis was you took the average total EV failure rate over time and extrapolated from that how many EVs would drop out from the fleet each year and then calculated how long that would take for the total drop-outs to exceed 50%.

Was that a correct take on your analysis?

The authors of those articles hypothesize that the reason for newer vehicles experiencing fewer failures is due to improvements in technology and they imply future improvements will result in even fewer failures over time. Their analysis fails to take into account that we have no evidence to support the contention that 2016 and newer vehicles won't adhere to a similar higher than average failure rate once they age beyond nine years--because we aren't in 2025, yet. That is, the high failure rate of early model year EVs could be related to their age rather than the use of older technology in and of itself.

If the independent variable was age we'd expect to see failure rates increasing as the vehicles aged. If it was due to the technology we'd expect to see similar failure rates across eras of EVs (since the technology did not shift substantively year to year). In fact, we see the first expectation:

We saw increased numbers of non-recall battery replacements in:

• 2011 - 30.00%
• 2012 - 15.94%
• 2013 - 9.81%
• 2014 - 6.81%
• 2015 - 3.90%

Now we can add some additional premises to our earlier three:

1. As an EV ages past its warranty period (read: expected, or federally regulated, lifespan) the failure rate accelerates.
2. In addition to that, based on these data, we would not expect to see significant failure rates of the 2016 and newer models until at least 2028.

We'd also want to compare that to average failure rates of early and late model ICE vehicles to provide a holistic understanding of the relative lifespan of an EV before determining it's better or worse than the general public's expectation of how much life they could get from an average vehicle.

In the article, the author portrays an EV as having an average mileage expectation similar to that of a modern ICE vehicle:

A modern battery pack that is only built to last 500 full charges and discharges can net a driver 150,000 miles. For comparison, it is common for a cell phone battery to degrade enough within two years for folks to consider getting a new one. Cell phone batteries are often fully charged and discharged once each day. Over 2 years, that would be approximately 730 cycles. On average, EV batteries are built to last over 1000 cycles.

Our ER F150L batteries have a nominal 131kWh capacity. That's supposed to result in a 320 mile range, which is an approximate 2.44 miles/kwh (I experience a more conservative 2.3 mile/kWh long-term estimate in my truck). That's how the authors derive the greater than 150,000 miles per 500 cycles. The problem with that, however, is those averages don't reflect how people drive their vehicles. For example, if someone exclusively towed with their Lightning they could experience a lifespan as short as 65,000 miles (500 cycles * 131 @1.0 mile/kWh). It also ignores all the other things that catastrophically fail on an EV aside from the cells themselves and doesn't calculate the potential lifespan according to their own charging suggestions (resulting in a 60% usable portion of the pack, or 79 kWh, rather than 100% of the 131kWh), which would result in 96,000 miles instead of 150,000.

But is our only concern absolute failure?

Using their own estimate of average battery degradation (1-2%/year) based on time alone, our ER trucks could experience a range reduction between 30-65 miles by the time they are ten years old. If someone is keeping 20% in the "tank," only charging to 80%, and has lost 10-20% of their battery capacity over ten years, they would be looking at an effective range of 150 miles. That may be good enough for many owners here but it's only about half of the industries' own predicted range that was necessary for broad adoption by the general public.

 

[Piquette]

This truck will celebrate its 100th birthday in July. I won’t be around in 40 years, but hopefully someone will still be driving it.

Oh, it’s a Ford too.

Sponsored