Life and Death of Li Ion batteries
Discussion
A little study shows Tesla battery capacity typically heading for the 80% watershed around 10 years:
https://www.nimblefins.co.uk/study-real-life-tesla...
Some science about Lithium batteries:
https://iopscience.iop.org/article/10.1149/1945-71...
Summary for shed drivers:
Once a battery is getting down to ~80% of new performance, failure mechanisms are accelerating and it's rapidly downhill.
https://www.nimblefins.co.uk/study-real-life-tesla...
Some science about Lithium batteries:
https://iopscience.iop.org/article/10.1149/1945-71...
Summary for shed drivers:
Once a battery is getting down to ~80% of new performance, failure mechanisms are accelerating and it's rapidly downhill.
SWoll said:
Why do you consider 80% a watershed?
A current Model 3 LR at 80% battery life would still be capable of 250 miles between charges. Hardly at the end of its useful life and still considerably more than many other EV's when new?
Because the decay (for want of a better word) is not linear.A current Model 3 LR at 80% battery life would still be capable of 250 miles between charges. Hardly at the end of its useful life and still considerably more than many other EV's when new?
As described in the second paper.
Once a cell is down to the 80% region, all the decay processes are accelerating and it's relatively short haul to 50% and then more quickly still to 10% and time for recycling.
This isn't about Teslas specifically, it's just that Teslas and Leaves have useful amounts of data.
80% is a vague consensus or average of the 'knee point' in the graph where decay steepens.
Different studies with different conditions and methods put the knee anywhere from 95 to 70%
Merry said:
wormus said:
It’s commonly accepted that an EV battery needs replacing every 10 years or 100k miles.
Is it? For which chemistry? LFP is capable of many more cycles than NMC, for example. So it would hardly be a hard and fast rule. The answer on battery replacement is more like 'it depends'. Batteries do degrade, no doubt about it. We had a Mk1 Leaf so know all about that - but it strikes me as becoming less and less of a problem going forward, particularly taking into account the average miles someone in the UK would do.
As someone pointed out due to the size of some of these packs even 80% remaining capacity would still give you a pretty useful vehicle.
Clearly long range cars like Teslas are managing more than 100k miles.
LFP batteries are indeed good for many more cycles than most other chemistries.
But not many cars have LFP batteries.
With most cars it is going to be a combination of calendar aging and cycle aging.
LG make a battery or two.
Measurable degradation after only 40 weeks.
https://xebike.com/wp-content/uploads/2019/12/lg-e...
Nomme de Plum said:
If the battery warranty is 8 years then manufactures will know that useable life will be much, much longer.
I've spec'd batteries in other industries.Depending on the consequences of a warranty 'failure', the average life might be anything from 20% to 100% more than the warranted life.
Statistics and all that.
If the failure consequence is somebody's 'uninterruptable power supply' only lasts 55 minutes instead of an hour and the have to fire up a generator, the battery maker faces a minor expense of giving the customer a nice discount on a new battery.
That will have a small margin on the life time spec.
If we're talking life support equipment and the consequence of falling short is death, law suits and corporate manslaughter charges, margins will be on the big side.
I think if your Ford EV loses too much capacity inside 8 years, they warrant to fix it back to 80% of nominal range or so, or in the limit that might have to buy back a handful of 6 year old cars or something. That side of it is easy, but the reputation damage of early failures would be taken seriously.
So, no doubt there is some margin in there. The average battery won't die before the mileage or time.
They can afford to be nice to 1% of failures or something and that sounds better than having a Range Rover.
But if the margins are as much as 50%, you won't want to be paying too much for the car once it's out of warranty.
Nomme de Plum said:
....
A decade ago battery management systems were not quite as sophisticated as they are now yet my friend's BMW i3 is showing minimal battery degradation at 9 years. I'll ask her the actual though.
.....
What the car chooses to tell her, and what might be measured under proper lab conditions might be two very different things.A decade ago battery management systems were not quite as sophisticated as they are now yet my friend's BMW i3 is showing minimal battery degradation at 9 years. I'll ask her the actual though.
.....
SWoll said:
OutInTheShed said:
What the car chooses to tell her, and what might be measured under proper lab conditions might be two very different things.
That's a rather lazy deflection. More likely it just doesn't suit your argument if we're being honest?It's not even an 'open' secret that many battery devices start off at something like 105% of nominal to give some margin, but only report 100%.
Also is the i3 one of the cars which when new doesn't fully charge, because that's best for battery life?
I think with the i3, you can delve into it and get calculated capacities in kWh?
But these have a fair error margin, because apparently the indication you get varies up and down several percent from day to day.
What a random 'merkin i3 owner said:
New user here. Love my 14 REX, aside from this battery degradation thing. Here's my story:
I got my car around this time last year. I'd get around 75 miles of range on battery-only power for the first 7-8 months of ownership. Then I never got more than 65 miles. Chalked it up to cooler temps in the fall and winter, no big deal. I always charge the car to full on my 220 charger. When the car is ta home, it's usually on the charger.
In January I started seeing 60 miles max per charge, and the weather was warmer. Read about "batt. kappa max" here on the forum and started tracking it. First entry in my spreadsheet is January 27th - 48 degrees Fahrenheit, 52 miles of range, batt kappa max at 13.8. By April by April 26th - 3 months later- I saw my range drop to around 52 miles (83 degrees, 50 miles range, 13.2 batt. kappa max). My low for a single charge was 44 miles. You can see as you red down the spreadsheet - 13.8, 13.6, 13.4, 13.3, 13.2. It's a pretty smooth decline..
According to my calculations I was at 70.2% battery capacity. Granted, batt. kappa max is an estimate, etc.. but it was trending down over a fairly short period of time. Falling off a cliff, really. To go from a consistent 75 miles to 55 miles TOPS in a year is unacceptable. Time to call the only people I know of that have the ability to look into this problem for me - the dealer.
This is where the story gets weird. The dealer is cagey about what info they'll give. I ask what they do in this situation, how they check capacity? They tell me they plug the car in and BMW looks at the car and tells them how to proceed. They then measure the capacity of the battery with their diagnostic software. Questions are answered with rambling non-answers. No real info for me. 6 days later I get the car back and magically I'm at 16.2 batt kappa max. The car says it has roughly the same range, but after a couple drives it goes up to 64 miles. Dealer says I had 77% capacity. Also says they'd be happy if I was at the threshold for a warranty claim as they'd love the work. Right.
Now, I'm happy they've returned some range to me. What I'm afraid of is owning a failing battery pack which my local dealer (or BMW corporate) has moved the "safety stops" on. I had a pack which according to the car (batt. kappa max) was at 70%. I have the same pack, which is now miraculously at 86%. Not only did the dealership not tell me that they got me more capacity, they didn't tell me how they test capacity or what parameters they check. Doesn't seem right (legal?) to engineer and sell a car with certain parameters built into how the main power source functions, and when it stops functioning in that way the manufacturer reengineers it.
I'm going to call the dealer and try to get some info. Just thought I'd share some real world experience here. There doesn't seem to be much out there.
TheDeuce said:
It's completely different - you charge your iPhone every day, a typical driver in an EV will charge once per week = two years for your phone is 14 years for an EV.
Y....
Your phone may get charged more often, but a year is a year is a year.Y....
Also although I charge my phone most nights, it's rarely discharged very deeply.
It does take a kicking when I run GPS based apps, which can flatten it in a day's hiking.
As it happens, my current phone is rather more than two years old.
A battery, or each cell, is a bit of wet chemistry. As well as wearing out from use, the 'wrong' chemical reactions happen over time.
That's pretty true for every flavour of battery, from Alkaline through lead acid to zinc carbon.
The LiFePO4 batteries which are favourite for home power storage are promising a lot more cycles than 'traditional' lithium, but feel free to point out any offering a decent warranty more than 10 years.
Evanivitch said:
OutInTheShed said:
LFP batteries are indeed good for many more cycles than most other chemistries.
But not many cars have LFP batteries.
Only about 20,000 standard range MG5/ZS, another 15,000 later Model 3 SR, and 20,000 Model Y sold on the UK!But not many cars have LFP batteries.
About 6% of BEVs on UK roads?
It's possible in due course, the LFP cars will hammer the used value of cars with other chemistry.
I wasn't planning to wait that long though.
SpeckledJim said:
Standard Range Plus gets better with age?
That throws the whole thing into question.
I suspect it shows that simply reading the 'range' off the dash doesn't tell you the whole story.That throws the whole thing into question.
Maybe higher mileage cars get driven in such a way that they do better mi/kWh?
Maybe the Muskware is compensating for the natural aging by allowing the battery to be flattened to a lower voltage?
wormus said:
GT9 said:
I was referring specifically to the shape of the degradation curve.
Sorry, it’s a Tesla sales brochure and it’s absurd to suggest it’s still got over 90% capacity after 100k miles. 80% might be realistic I’d guess but then these batteries are considered “dead” by 70% so the question is, if you buy a 100k, 10 year old Tesla, with unknown charging/usage history, how much life does it still have? These are the useful answers for skeptics like me to ever consider buying one, not the naive “EVs are amazing!!!!!” bilge which seems to permeate these threads.It will also become important for the 2nd user in the middle of the market, because his ownerships costs having an EV from say 4 to 8 years, will be hugely influenced by what the 3rd user is prepared to pay.
GT9 said:
I was referring specifically to the shape of the degradation curve.
There are several documented cases of ultra-high milage taxis, here's one.
https://electrek.co/2020/05/11/tesla-model-x-extre...
No sign of the main battery on the maintenance list???
It's interesting the way a car used as a taxi can do galactic miles.There are several documented cases of ultra-high milage taxis, here's one.
https://electrek.co/2020/05/11/tesla-model-x-extre...
No sign of the main battery on the maintenance list???
In diesels, it's partly because the engine rarely gets cold.
With an EV, and more specifically the battery, maybe it's quite significant that the battery spends little time resting at a steady charge level?
This kind of links in with some of the V2G studies which suggested that 'exercising' a battery could actually be good for it.
Maybe it supports the concept that time kills batteries just as much as use? Or even more so?
The taxi in question did the miles in just a few years.
Doing over 300 miles a day average, it can't have been idle much.
Nomme de Plum said:
......
Surely data can only be available when sufficient EVs have reached 10 or more years and/or mileage over 100k.
Anything else is a prediction or some sort of modelling.
Modern EVs did not go into production until 2010. Nissan Leaf.
Quite a lot of early leaves are having pretty tired batteries now, but how much of that is age, and how much is caused or exacerbated by the relatively low battery capacity, which means the battery gets cycled a lot harder?Surely data can only be available when sufficient EVs have reached 10 or more years and/or mileage over 100k.
Anything else is a prediction or some sort of modelling.
Modern EVs did not go into production until 2010. Nissan Leaf.
It's perhaps unfortunate that the two ranges of cars we have most data for are Leaf and Tesla, which lie at extremes of nominal range.
A lot of the time based aging will probably follow (broadly) the norm for chemical reactions, i.e. faster in warmer conditions, so perhaps data from the warmer bits of the US might extrapolate to UK conditions?
It's not hard to find anecdotal forum stuff about i3's suffering early battery decline in the US, no doubt BMW has gathered a lot of data from that.
BMW of course will get info from the cars which are doing well, as well as those whose owners are whining on the internet.
It would be nice to think that such data would encourage carmakers to stand behind their batteries for longer.
220 capacity fade scenarios are analysed, based on GPS driving data from 16,263 vehicles.
•
41 out of 220 scenarios give a capacity drop below 80% of the nominal battery capacity in less than 5 years.
From
https://www.sciencedirect.com/science/article/pii/...
•
41 out of 220 scenarios give a capacity drop below 80% of the nominal battery capacity in less than 5 years.
From
https://www.sciencedirect.com/science/article/pii/...
Andeh1 said:
You are wrong, quite simply. If you were an expert/consultant you would recognise that report is on Chemistry the industry no longer uses & is 2 generations old. Whatever they concluded fair enough, but the industry has moved on a long way & that report is a moot point.
My point still stands. Buy a newly launched EV from the last year or two and it'll see you through 250k without breaking a sweat. The reliability over that duration (excluding first month or two of manufacturing defects) would also be significantly better then an ICE vehicle.
So you are saying that cars previous to about 2021 have inferior battery technology?My point still stands. Buy a newly launched EV from the last year or two and it'll see you through 250k without breaking a sweat. The reliability over that duration (excluding first month or two of manufacturing defects) would also be significantly better then an ICE vehicle.
We're not seeing that writ large from automakers TBH.
The viability of the 500,000 or so pre 2021 cars is not a 'moot point' in the used market.
Those cars are out there and people want to know what the prognosis for them is.
You seem to be saying it's somewhat poor compared to the latest offerings?
On that basis, someone considering a used EV this year might be better off either waiting or paying the extra for a new car?
Feel free to link to published science or something carmakers will stand behind to back up your assertions.
That was what I was looking for when I started this thread, I wasn't really wanting the defensive ad hominem stuff from the uninformed EVangelists.
If you actually look at what I've written it's actually mostly open questions and links to reputable sources.
off_again said:
There is a Toyota Rav4 EV around where I live!
https://en.wikipedia.org/wiki/Toyota_RAV4_EV
We dont have age related plates here, so I cant tell you how old it is, but they were made between 97-03 - so not quite 25 years old, but most likely 20(ish) years old....
Yikes, that makes me feel old! 2003 was only a few years ago. But I digress - it is perfectly fine and working well. Dont know the owner but its not broken, dead or anything like that. I guess this is the exception that proves they dont exist.
They were NiMh battery, so all good fun and nice to hear about, but roughly as relevant to modern EVs as the lead-acid fork lift truck I drove 40 years ago?https://en.wikipedia.org/wiki/Toyota_RAV4_EV
We dont have age related plates here, so I cant tell you how old it is, but they were made between 97-03 - so not quite 25 years old, but most likely 20(ish) years old....
Yikes, that makes me feel old! 2003 was only a few years ago. But I digress - it is perfectly fine and working well. Dont know the owner but its not broken, dead or anything like that. I guess this is the exception that proves they dont exist.
Weren't the early Prius NiCd?
Back in the early 2000's I recall the early adopters of Lithium batteries, Sony Laptops, having some inflammation issues.
Feel likes a lifetime ago, but in other ways no time at all.
SpeckledJim said:
You are right. So show us the thousands of dead EVs with knackered batteries to prove it. Autotrader. Ebay. Copart. Wherever.
There will be thousands of early Model S going for £5- 15k because their battery is goosed. Just like you said.
(Why am I bothering?)
Take a look at Howmanyleft for those 'thousands of early Teslas'.There will be thousands of early Model S going for £5- 15k because their battery is goosed. Just like you said.
(Why am I bothering?)
They didn't sell very many.
An interesting number of older ones are SORN or 'missing'.
dvs_dave said:
You’re surprised that current Gen tech is better than previous Gen tech? Against the backdrop of previous Gen tech still being capable of similar lifespans proven by real world consumer usage.
Seems your concerns are actually moot.
Mostly 'Gen Tech' tends to be marketing bSeems your concerns are actually moot.
ks.Reality tends to be small incremental improvements or economies.
SWoll said:
Otispunkmeyer said:
This will likely improve
Euro 7 is basically going to ask that BEVs subscribe to a 90% SOH or better at 10 years as part of in service conformity testing. So at 10 years (and presumably there'll be some caveat on average miles per year) you'll still have at least 90% of the battery health left.
I think that is sufficient.
Going to be interesting to see how they provide that?Euro 7 is basically going to ask that BEVs subscribe to a 90% SOH or better at 10 years as part of in service conformity testing. So at 10 years (and presumably there'll be some caveat on average miles per year) you'll still have at least 90% of the battery health left.
I think that is sufficient.
Battery recharge cycles? At what charging speed? At what temperature range?
Does time not affect battery efficiency? If it does how do they simulate that?
If it's not at 90% after 10 years what is the penalty? How is it confirmed and enforced? Does the owner get compensated?
Seems a ridiculous rule that's there purely to try and bolster confidence for the buying public who have concerns like those discussed at length in this thread? Posturing basically.
Edited by SWoll on Friday 10th March 19:37
For many things we would up the temperature because chemical processes often double in speed with 10degC temperature increase.
Not sure that's valid for batteries.
A 10 year old car is going to be a banger anyway, but on the other hand we have millions of older IC cars which people are using every day.
The concern is that Euro 7 compliance might exclude affordable cars from the market. BMW/VW/Steel-itis are probably quite keen to do that.
Euroland is the new iron curtain, wanting to exclude good from other continents.
Will the UK follow euro 7 with £30k cars while the commonwealth are tooling around in £10k asian city cars?
Gassing Station | EV and Alternative Fuels | Top of Page | What's New | My Stuff