Finally, a topic every man can relate to.
I had been under the impression that as an EV ages, it ages gracefully. It might slowly lose range, for sure, but otherwise, it was the same vehicle it always was. I didn’t read a lot of chatter about them losing capability, or reliability, as they age. So the story seems to be that EVs simply age into being reliable used cars, with more limited range than when new.
Maybe that’s true. Maybe that’s sometimes true. Maybe that’s true except for the bottom-of-the-market segment that I’m currently shopping. Maybe that’s pro-EV propaganda.
To cut to the chase: Size matters. That’s my conclusion, at any rate. All other things equal, performance and reliability issues should show up soonest in EVs with the smallest range. That’s because, for any given maneuver on the road, those EVs draw more power per unit of battery capacity. Unreliability shows up first during events that use (or produce) a lot of power.
My lessons so far from bottom-fishing used EVs, and the key question moving forward.
Everybody likes a shiny new car. But this series of posts is about well-used EVs. The bottom of the market.
To be clear, used cars constitute a self-selected subset of cars. And not in a good way. For example, while some (presumably) nationally-representative statistics suggest that Nissan Leafs lose about 4 percent of battery capacity per year, the used cars for sale in my area averaged about twice that rate of loss. Plausibly, that’s why some of them ended up for sale. So don’t take any of my conclusions about the used-car market as an indication of quality of cars in general.
But if you’re buying a used car, that’s pool of cars you have to choose from. People tend to sell their lemons, not their peaches.
First, not stated previously, there’s no publicly-available data showing actual on-the-road average lifetimes for EVs. For example, I stumbled across an estimate that Nissan sold 19,000 2011 and 2012 Leafs in the U.S. But nowhere can I find out how many of those 19,000 are still registered, and so presumed driveable.
That’s typical for cars in general. You can find data on the number of new cars sold from a wide range of sources. But we (the public) know virtually nothing about the number of each model scrapped, other than the aggregate number. There are proprietary “Vehicles In Operation” estimates, such as these from Experian. I’m pretty sure these are derived from state vehicle registration data. So I think they’re literally counts of actively registered (tags-still-good) vehicles. Close enough to “still on the road.” But as far as I can tell, there’s no data down to the make-and-model level available to the public.
So, when somebody tells you how long an EV is expected to last? They’re guessing. (With the possible exception of Tesla, which seems to track every car. And they’re not saying.)
Second, buying a used EV is all about the battery. Duh. Buying a used EV, purely to get from A to B reliably, is almost entirely about the state of health of the battery. As with any car, there are plenty of other wear-and-tear parts. But the death of the traction battery is typically the death of the car, absent the occasional owner willing to replace the traction battery on an elderly vehicle.
(As an aside, my wife replaced the traction battery on her 2005 Prius. The NiMH traction battery died around 175K miles, at which point the car was worth less than the roughly $3500 cost of replacing that battery at the dealer. But she chose to replace it, both because the car was otherwise in good shape, and because we loathe the hassle of buying a car. Plus, we knew the car’s history. It was a case of “better the devil you know than the devil you don’t.”)
Third, information on battery health, provided by used-car sellers, is more-or-less worthless. That’s my analysis of the health-and-range estimates provided by Recurrent.com, and featured in listings on Edmunds.com. I got interested in a Leaf that had 110 miles of range left, per Recurrent.com. It actually only had about 65 miles of range, per the car’s actual dashboard readouts.
Now that I know that, I would not consider shopping for a used Leaf without at least buying Leaf Spy and a compatible cheap ODB-II bluetooth dongle. This is a program that runs on your smartphone, reads the OBD-II data, and allows you to look in detail at the battery. In particular, it gives you some ability to spot weak battery modules, particularly if you can drive the car with Leaf Spy running.
Fourth, the big surprise was that some Nissan Leafs do not age gracefully. I saw quite a few anecdotes of older Leafs that were still driveable, but had become temperamental as they aged. For example, Leafs for which highway driving became chancy, either from incredibly rapid range loss, or from triggering “turtle” or “limp home” if freeway driving were attempted. Cars that would do the same on long uphills. Cars that had lost their regenerative braking.
The common thread to those events seems to be high power flow, out of or into the battery. The batteries could handle modest power demands just fine. But the battery — and so, the car — had lost the ability to handle high power demands correctly. Equivalently, because the battery pack runs at a roughly constant voltage, these older batteries had problems when faced with high current draws, or high regenerative braking currents, via the West Virginia Law:
Watts (of power) = Volts x Amps (of current).
Regenerative braking is a high-power event? Oh, yeah. Think about it. In a typical car, you can slow down faster than you can accelerate. Which means the power created by regenerative braking can easily be as high as your propulsion power. I put a Scangauge 3 in my wife’s Prius Prime, to try to avoid high-current events, and soon discovered that the brake pedal was a lot more dangerous than the gas pedal, in that regard. I rarely exceeded 75 amps on acceleration. Even a moderately hard stop would push the regen current over 125 amps.
The common explanation for this is uneven aging of the battery modules, leading to “weak” modules. These “weak” battery modules suffer unacceptably large voltage drops when subjected to high current draws. The car, in turn, senses that and reacts. Among other things, the car then limits the amount of electrical current allowed to flow into or out of the battery. Which means that it limits the amount of power that the car is allowed to produce. The result is a functional-but-temperamental Leaf. The car may drive fine if un-stressed, but may act wonky if asked to do (e.g.) sustained highway-speed driving, or to accept high regenerative braking current.
(FWIW, we got nothing like that when the traction battery died in my wife’s 2005 Prius hybrid. I would characterize that battery failure as abrupt total failure, preceded by a few warning signs. Mainly, about a year before total failure, the battery-cooling fan started to kick in on long downhills. That never occurred when the battery was new. But the car ran perfectly, right up to the day when I started it, got the “red triangle of doom”, and the car sputtered to a halt. I limped it home.
I think there was no loss of performance, up to date of battery death, because that’s a hybrid. It runs on a gas engine, the battery is only there to supplement power during acceleration, and otherwise even out demands on that gas engine. The weak traction battery did not prevent the gas engine from running properly, right up until the battery finally died entirely. At that point, it would not run normally, at all, under any circumstances.)
As I continue my search for a car, I must ask: Is that something unique to the Leaf? Or is that true, to some extent, of all EVs, as their batteries age? Maybe we just see it more starkly in the Leaf because a) those were the first EVs sold in large numbers, b) the earliest ones had a (plausibly) inferior battery chemistry, c) they had no thermal management system to protect the battery, and d) they just plain had small batteries.
In other words, there’s a notion out there that old EVs never die, they just fade away. They remain capable vehicles, and only lose range as they age. They age gracefully, so to speak.
With the exception of these low-end Leafs that I’ve been looking at, is that actually true? Or do all EVs lose both range and capability/reliability, once they’ve lost (say) 40% of initial battery capacity?
If (some) old EVs don’t fade gracefully, I need more caution when shopping for a low-end used EV. I may easily come across cars that run OK, but that will underperform or fail in a high-stress situation. I don’t need that. So I need to know, is loss of EV functionality with age, is that a Leaf thing, or does that generalize?
A clue from a temperamental 200K Tesla.
I stumbled across an anecdote from one Tesla owner (which, unfortunately, I cannot now find and cite), with more than 200K miles on his car. His car lost both range and performance as the battery aged. In his case, unlike the Leaf, there were no warnings triggered. But, in effect, when he floored it, he got less of a response than he used to. So much so that he began to notice it.
My guess is that this is a true anecdote. And that it probably shows that most, possibly all, EVs are subject to loss of performance as the battery ages. So older EVs probably do lose both range and performance.
But, interestingly, in the case of the Tesla, 200K miles was not enough to bring up outright failures during high-power maneuvers. As opposed to the early Leafs, where loss of performance was bad enough to (e.g.) trigger the car to severely restrict power draw (i.e., enter “turtle” or “limp home” mode). But, as with the oldest Leafs, the range gauge got wonky as well.
Does higher range mean a longer period of reliability?
Near as I can tell, there are no hard numbers whatsoever on “graceful aging” of EVs. Mostly, I think, because there are few EVs out there long enough to have suffered significant range loss. Plus, no hard definition of what that means.
So I’m going to have to take a guess.
I think these oldest Leafs with degraded range are showing higher rates of degraded performance mainly because the battery pack is small, relative to the size of the car.
The problem appears to be high power demand, causing voltage drops at weak modules, triggering a reaction from the car’s software. Something like a high-speed merge in an older Leaf, with a 24 KWH battery, is going to draw more current, from each cubic inch of battery, than an identical high-speed merge in (say) a Tesla with a 100 KWH battery.
In fact, all other things equal, it’s going to draw about (100/24 =~) four times as much current. Per cubic inch of battery. Assuming that both vehicles have batteries that are identical except for size.
(I have to use that awkward “per cubic inch” rather than “per cell” because the individual cells — what any normal person would call batteries — may vary widely in size. In just the same way that you can draw more current from a “D” cell than from a triple-A, you can probably draw more current from a bigger EV battery cells than from smaller ones.)
Restated, if battery pack voltages are equal, then those high-power events will result in equal flows of current coming out the battery packs. But this means a fourfold higher current density flowing through the cells within the battery pack, for the Leaf.
It’s entirely possible that these oldest Leafs have other issues as well. The battery chemistry used for the Leaf is not what modern EV makers use. Possibly, the software and other management systems are less sophisticated.
But I’m guessing the main culprit is simply small battery size. If the problem is due to high power draw, relative to the size of the battery, well, that’s going to show up first in the cars with the smallest batteries. All other things equal.
But what about PHEV, then?
By that logic, aren’t you in for a world of hurt if you buy a plug-in hybrid electric vehicle (PHEV)? Those cars operate part-time as EVs, but they have even smaller batteries than the original Leafs. My wife’s Prius Prime, for example, is just about the size and weight of a Leaf, but has a battery roughly one-third the size of the original Leaf battery (about 8 KWH, versus 24 KWH).
Here, I’m guessing that any loss of performance as an EV is going to be more subtle. That’s because these cars are still, basically, gas hybrids. In the case of a Prius Prime, the car will turn on the gas engine whenever it thinks it needs it. (Separately, it has an “EV preferred” mode, where you lower the threshold above which the car draws on the gas engine, rather than stay in pure EV mode.)
Well, we already own one, so it’s too late to worry about it. And so far, I’ve read no anecdotes about temperamental Prius Primes. Or traction battery problems with the Plug-in-Prius, an earlier version of the Prime with an even smaller battery.
Only time will tell.
But it does give me pause about buying a well-used PHEV, such as a mid-2010s Volt. As discussed in earlier posts, PHEV should just plain beat the hell out of a battery, relative to EV use, owing to the vastly smaller size of the battery. But I have no evidence that’s even starting to show up in the used PHEV market.
Conclusion
For now, the only used EVs for which I read many anecdotes of “driveable but temperamental” behavior are older Leafs. But I suspect this same issue affects other older EVs, to a lesser degree owing to larger battery size.
I think my only conclusion is that I probably want to stay away from used EVs with small batteries. It’s not that they are all bad, per se. It’s that, as I read it, the smaller the battery (relative to the size of the vehicle), the greater the risk of running into current-limiting behavior by the car as the battery ages. You might end up with a car that seems to run fine, until you floor it on the interstate on-ramp on a hot summer day. Then is not the time to find out that what you actually own is a reliable Neighborhood Electric Vehicle, but no longer a reliable car.
Alternatively, if I opt for one, I want to have the battery assessed in detail, in some fashion. I’m going to look around for the equivalent of Leaf Spy, for use on multiple models, if such a thing exists. Otherwise, I’m going to see if there is a battery-assessment service available in my area. Other than national firms catering to commercial customers, I’m not seeing one. Plausibly, I’d have to take any prospective used car, to that car brand’s local dealership, to get a battery test. If I can find one that will do that. Which just adds to the overall hassle factor of buying a used car. But is, in principle, no different from having a mechanic check over a conventional used car before you buy it.
Pictures in this post are from freepik.com ai, and Gencraft.com.
