I have been weighing the pros and cons of a future EV purchase, not in the immediate future, but maybe in the next few years. But it would probably need to be a fairly low (like under $10,000) budget. And there are some questions that I have where I can't seem to get a straight answer. Perhaps there aren't straight answers to be had; if nobody knows at all, that would also be worth knowing.
First, just how long do the batteries last? Some EV advocate sites have been claiming "Don't worry, the batteries should last 10 to 12 years!" Needless to say, with both my daily driver and my wife's coming up on 12 years, I read that as "Worry like crazy!" Does anyone have better numbers, ideally with a link to data?
Second, I've heard one report of a Ford Focus EV that had its battery pack wear out and the electronics simply refused to operate even though the battery could still be capable of driving the car with diminished range. How many cars have some sort of hard cut-off programmed in, where X% loss of battery capacity means it can't be driven? And if a car has this programmed in, what does it take to override this? Some sort of factory diagnostic tool, or Ghidra, major league coding skills, and a massive disregard for copyright laws?
Third, what EVs lend themselves to home battery swaps, and which ones don't? The Tesla Model 3 supposedly has them somehow built into the floor pan and difficult to remove.
Fourth, some sites claim that it's possible to repair EV battery packs by locating which battery cells are dead and just replacing the dead ones. How much does this extend a battery pack's lifespan, and what tools are needed to test the individual cells?
In reply to MadScientistMatt :
I don't know any answers 100%.
But for #3, do you mean getting a factory battery pack and replacing it? Or some kind of DIY battery? Although, I doubt either is going to be realistic- not sure that companies will want to take the risk of DIY battery swaps. Yea, I know they will let you do it all day long with a gas engine and all of that. But this isn't the same thing, and different training will be needed to deal with the battery.
For #4, I'm sure that's possible, if you can actually identify which cell is done. But I've tried some DIY battery building- welding cells together and whatnot. Not easy. I'd study this one a lot before you pulled the trigger on an EV that has range issues.
One thing that is common among 2-4 is getting into the car's EV software (controller, BMS, etc). Have you seen that people are reliably doing that? Some of the batteries I played with came with BMS's that had not been hacked, so it took a lot of physical work to take them apart.
On the other issues:
MadScientistMatt said:
Second, I've heard one report of a Ford Focus EV that had its battery pack wear out and the electronics simply refused to operate even though the battery could still be capable of driving the car with diminished range. How many cars have some sort of hard cut-off programmed in, where X% loss of battery capacity means it can't be driven? And if a car has this programmed in, what does it take to override this? Some sort of factory diagnostic tool, or Ghidra, major league coding skills, and a massive disregard for copyright laws?
Depends on the car, factory diagnostic tool is the best-case scenario, but Ghidra + l33t coding skills + wanton disregard for the sacred copyright laws might be needed.
MadScientistMatt said:
Third, what EVs lend themselves to home battery swaps, and which ones don't? The Tesla Model 3 supposedly has them somehow built into the floor pan and difficult to remove.
I know the Leaf and Tesla S have a pack that can be dropped out of the bottom of the car, and older Priuses have them as a box below the seats. Building the pack into the car is unusual from what I've seen so far, most EVs have a pack that unbolts from the floor.
MadScientistMatt said:
Fourth, some sites claim that it's possible to repair EV battery packs by locating which battery cells are dead and just replacing the dead ones. How much does this extend a battery pack's lifespan, and what tools are needed to test the individual cells?
This is true, it extends the pack's lifespan by the difference between when that one cell died and the average life of the remaining cells I suppose...or maybe to the life of the next outlier cell to die early. To test an individual cell you'd just need an ordinary multimeter, but a dedicated chemistry-specific battery tester you can get from AliExpress for $20 might be helpful. You'll also need spot-welding tools to reassemble a battery module with replacement cell, again they can be had for lunch money.
The Leaf and Volt debuted in 2011 and every other mass market EV is even younger--that's just not that long ago. EV tech is so different between manufacturers, and the changes in battery chemistry, pack construction, and controller programming from generation to generation of EVs make it hard to draw meaningful conclusions. Only Tesla has a large enough pool of consistent data to draw from--and those conclusions are not necessarily representative of EVs outside of Tesla.
Having just searched for and purchased a used EV in that price range, I'd say that cars currently under $10K are going to be first generation tech that is seriously compromised in terms of available range. These cars really only make sense for a niche set of use cases.
We paid a little more ($12K including taxes and registration) to get a second-generation Nissan Leaf. Its a great car and I enjoy running errands in it--but its real-world range of about 130 miles just isn't enough for most people.
Right now you need to move up to the $15,00+ range to get a Bolt or early Tesla in order to have over 200 miles of range. However if you don't plan to purchase for several more years, I bet that the used EV landscape at that time will have whole host of better and cheaper options as leases end for a bumper crop of current cars.
I think only the Gigapress Teslas have structural batteries. Not sure if they're replaceable or not, but they are for the earlier (much more common) cars. Tesla publishes their service manuals if you want to read up on the process.
Other than the Leaf (which does not use water cooling), EV battery lifespan looks to be on par with ICE lifespan at least. I suspect we'll also see the growth of refurbished battery packs like we see remanufactured engines.
The current incentives for buying EVs favor leasing. So in a couple of years, there will be a lot coming off lease and hitting the used lots.
MadScientistMatt said:
Second, ... Some sort of factory diagnostic tool, or Ghidra, major league coding skills, and a massive disregard for copyright laws?
Fourth, some sites claim that it's possible to repair EV battery packs by locating which battery cells are dead and just replacing the dead ones. How much does this extend a battery pack's lifespan, and what tools are needed to test the individual cells?
For your fourth item, which really overlaps your second, most controllers have a method for monitoring the individual cells in the pack. If you can "gain access" to the controller software or data feed, you would be able to check for cells that are on the way out. I think John Welsh has posted screen shots of battery cell state from some of his Prius threads, using the on car diagnostics. Its a pretty basic requirement for the system to know what is happening internally. Actually changing out individual cells may require a fair bit of work with cooling and packaging, but the cells themselves are almost all standardized shape/size formats.
Keith Tanner said:
The current incentives for buying EVs favor leasing. So in a couple of years, there will be a lot coming off lease and hitting the used lots.
ArsTechnica recently had an article on this topic: https://arstechnica.com/cars/2024/11/second-hand-evs-will-flood-the-market-in-2026-jd-power-says/
Going to be interesting to watch the pricing on used EV's in 2026+
so, we still have "Tom's Old Leaf" (a 1st gen 2012) that we bought from him in 2021 for $6k. It had a factory warranty battery pack replacement (24kWh) in 2017 (iirc). we bought it with 10 out of 12 battery life bars, and it was telling us ~74miles of range. We've just gone down to '8' battery life bars about 2 months ago. Generally it'll show between 65 and 70miles of range, temperature dependent (sleepywife tends to get better range driving it than I do).
It looks like the 2018+ Leafs with the 40kWh packs are just getting down to the $10k mark... and they'll have better range and better hvac (this matters with range). I think the 2018+s also had a ~60kWh pack option. Some of the things that hamper "full year Leaf driving" for us (30deg temps have a dramatic influence on a Leaf's range, and toe comfort) won't apply to you in GA as much. One nice thing about the Leaf is, that for $10k, they tend to have fewer miles on them compared to other EVs.
It looks like Bolts are starting to get down to that range... and if they've got the battery recall done, it'd probably be a good buy since their packs are liquid cooled... so they should have better life.
cars.com has started adding "EV Battery Health" via Recurrent. I dunno how they're coming up with that evaluation, or how accurate it is... but it's a nice that that's starting to be added to the used car information scape.
we've enjoyed the Leaf as a "3rd car", even though it's a "primary" car. all we've done the last 3 years is charge it, and put tires on it when we got it, and drive it ~20kmi apparently.
GameboyRMH said:
10 years is a minimum, from projected wear rates it looks like most EV batteries will last 20+ years:
https://thedriven.io/2024/09/19/new-study-finds-vast-majority-of-ev-batteries-will-outlast-the-vehicle/
Also here's a study on how they've been wearing so far IRL:
https://www.recurrentauto.com/research/how-long-do-ev-batteries-last
I had seen an earlier version of Recurrent's data, but now it seems to have a clear pattern:
Looks like the odds of an EV battery failing starts doubling every year after around 9 or 10 years. If newer batteries do not have the same abrupt degradation after this point, I would like to see what data shows that. Geotab (at the first link) does not show their data.
Keith - thanks, I didn't know Tesla put their FSM online. That's very handy to know, and it shows that even the structural batteries are under 6 hours flat rate. Unfortunately it looks like changing the battery would be a bear without a lift and a device that looks like a gigantic transmission jack. So something reasonable for an independent shop, but a nightmare to try in a carport with basic tools.
At this point it still looks like a cheap used EV is going to carry the threat of needing its own purchase price in battery replacement, with very little room to save money by doing a DIY repair. Not good, unless there's more data out there supporting the claim that newer batteries really can last 20+ years - maybe results of some sort of accelerated aging test?
In reply to MadScientistMatt :
Keep in mind 2011 era was pretty much just the Nissan Leaf which is just about the worst design for accelerated battery wear with air only cooling and very poor thermal management as a result coupled with taxing the batteries pretty hard. I wouldn't use that data as an indicator of general ev battery life at all. It's like looking at all car reliability and weighting the first few years with only Land Rover vehicles.
I've done an ev battery swap at home. I used my maxjax lift and a big caster wheeled cart I made to take out and reinstall the battery. I don't have time to go into details now but it was fun and interesting. Maybe later today I'll have time to type it out more.
If you want to do your own work Leaf is the way to go (because it has the worst battery), and Volt is getting there.
A lot of stuff out there, look up DIY electric car, endless-sphere, secondlife storage. Leaf batteries died because of lack of cooling, i3 batteries die from a coolant compressor design flaw.
Don't overlook i3, C-Max energi, e-Golf, and other less popular options. You do want a battery that hasn't been drained fully, or stored at full charge, and you don't want one from a hot area.
The only new EV in America I know of that should have a longer lasting battery is the cheapest Tesla Model 3. It is using Lithium Iron Phosphate batteries, which have a much longer lifespan than the older Lithium Ion chemistries. I belive some of the cheap Mach-E and Kia options are supposed to start using them.
Also remember, the battey failing is different from it losing capacity. If you can get by with shorter range/more frequent charging you will have a lot of options.
A lot of the older EVs are crap thrown into production to make EV sales legislations happy.
EVs with good battery and charging controls seem to have batteries that last WAY longer. Essentially immortal if you figure that the rest of the vehicle may be "used up" by then.
When the Leaf came out, Nissan also claimed its battery would last the life of the car. I can believe improved cooling can improve how long the batteries last, but it would be nice to have more reassurance about just how much longer newer batteries would last. Some sort of fitting data to a well established curve, torture test data, etc.
Any more notes on the second question - whether most EVs refuse to start after a certain percent of battery degradation, or keep running as long as there's enough battery capacity to move the car a few miles?
At least the third question on how much work it is to replace a battery and whether it can be DIY'd has a clear answer now.
There's some good crowd-sourced data on Tesla battery degradation. I posted it in my Tesla owner's thread a few years back, I'll have to see if I can find it again. Basically, there's an initial drop and then the loss levels out. Tesla has also published data based on their fleet but it's always nice to see something independent :)
Keep in mind that it's not age, it's discharge cycles you need to pay attention to. So a high mileage newer car will tell you a lot, and longer range cars are more likely to reach higher mileage. So there may be more data out there about newer batteries than you think.
IIRC lithium batteries don't gradually fall in voltage as they fail, they just run out of capacity. Tesla warranties their batteries to have at least 70% capacity during the warranty period - the length of the period depends on the model, but it's at least 8 years/120k.
Tesla has done a lot to demonstrate how many miles their batteries can last; there's one Model S out there with 1.2 million miles on the odometer on its fourth battery pack. So it seems you can get 22RE levels of distance out of them.
I'm still a bit worried that a time dependent degradation will show up before used Teslas reach the sub-$10,000 mark, as even well used examples still sell for significantly more. Also, would a Tesla with, say, 50% battery capacity still be drivable with half range, or does that trigger any sort of setting that makes it refuse to start?
I think it's actually the 12v battery that "starts" the car. The big battery powers the drivetrain and recharges the 12v, which runs everything else. Makes sense, there's a big automotive infrastructure for 12v.
If that's true, then a car with 50% degradation should act like a car that was built with 50% less battery - or one that's only charged to 50%. That's my understanding anyhow.
It looks like the accounts where a Focus EV would not start due to battery deterioration were due to a battery internal coolant leak, not diminished battery capacity. Some accounts of this in the news didn't look too closely at the failure mode and implied the car's firmware simply did not let it operate once the battery capacity dropped too far. So that is a bit less worrisome - an issue with a specific battery to avoid rather than a concern that there's an arbitrary, hard coded minimum battery capacity in order to drive an EV.
In reply to Keith Tanner :
The electronics run on 12V so the 12V is needed to bootstrap everything.
Some newer cars (hybrids) don't even have a traditional battery, they have a 12v lithium power source that is part of the HV battery assembly. It can be replaced separately. It doesn't need to be very big since it never sees high current loads, and it doesn't have to smooth out an electrically noisy alternator since it is "charged" by a step down voltage from the HV system.
In theory, one should be able to "jump start" an EV or hybrids 12V system from a USB cable.
I'll ask my industry friends if they know if there's a lower limit to battery degradation. I assume at some point there's a voltage drop. I do remember that a friend managed to get a Kia Soul with a toasted battery from Oregon to San Francisco. Car had a range of 30-something miles by that point, down from a rating of 93. Thta's a fair bit of battery degradation.
The 12v battery in my Model 3 is basically the same one as in a 1990 Miata. They use a little lithium unit now.
Turns out I had another friend who managed to get a Leaf battery to 99.9% degradation. 0 miles of range. It could still turn on and move back and forth (briefly) on flat ground.
Ok, so that makes me a lot less worried that the car will simply activate a "replace the battery now" failsafe for arbitrary reasons and that at least you'll get a progressive warning the battery is wearing out - much more progressive than any failure I've lost an engine to.
Our second gen prius went from working to paperweight suddenly due to a bad battery. Different chemistry than ev's so maybe that is the reason for the failure mode that would not allow the car to operate
Oapfu
HalfDork
11/10/24 6:09 p.m.
There are many different battery designs, and many different possible failure modes.
"How long does an IC engine last?" Depends on the design, depends if there are manufacturing defects, depends on how it was used/abused.
YT vid about fixing the battery in a FFE: https://www.youtube.com/watch?v=COsYIpy405Y
He finally does the simple thing and completely bypasses the battery cooling. It mostly has no effect on normal driving, with only occasional over-temperature warnings from fast charging or when ambient temp is extremely high.
The FFE battery uses pouch cells which are sandwiched between cooling plates. The cells swell as they age, and the swelling changes the spacing of the cooling plates. As the plates spread, the coolant lines between them start to leak. Coolant leaked inside the battery case triggers an isolation fault. The car sees an isolation fault as a very serious problem and it shuts down.
If you have a single battery cell sitting on a bench, you can fully charge and discharge the thing to figure out its capacity. Or, more accurately, its capacity at that specific temperature and that specific discharge rate.
An EV has a really large number of battery cells to deal with, and usually it can never fully charge and discharge them all while sitting under controlled conditions. Therefore, State of Health (SoH) is always an estimate, and State of Charge is never really exact, and the expected range display used to be called the "Guess-o-Meter".
dunno if this addresses any of the original questions, but it has some information I hadn't quite heard before, and is along the lines of what's been discussed here: