That’s an interesting finding, but it makes sense. Lithium ion batteries are, in my observation, basically linear in their degradation. Discharging a pack by x% degrades it by p*x%, regardless of the pack size. So it terms of miles driven, the degradation is also in miles.
BTW, that 220 mi Tesla would degrade to 174 after 200k miles, not 154 (if the figures are correct).
Cars are certainly a lot better than they used to be. But I still gotta bring my car in for a smog check every year. All these little pains-in-the-ass add up after a while.
Also, don’t forget a gas cap every 90k miles :). (that’s how long mine lasted, at least…)
jz78817,
It would be hard to read what I have posted here and peg me as a “Tesla acolyte” and maybe the option of getting shuttled to work during those maybe twice a year oil changes is a choice that both exists and works for you. That shuttle option does not exist for many of the rest of us and would not work for many of us even if it did.
No question that today’s ICE vehicles are better, much more reliable, and longer lasting, than ones of a few decades past, and maintained reasonably well lasting 200K is not unreasonable to expect. And the maintenance required is not extremely onerous … but it is more than is required in an EV and again … I value my time.
Dr. Strangelove, thanks for catching my stupid math error. So yeah even more so. At 200K driven the average Tesla 3 would still be fine to keep driving. 300K and it’s fine. Maybe redo the upholstery and give it a paint job. Getting under 100 miles of range (still quite usable) would not occur until over 500K miles driven for most. The battery will in most cases outlast the rest of the vehicle.
I don’t think it stays linear. Think about it. As the capacity declines, the percentage of that capacity you are using is increasing. Using more of the capacity of a battery (charging it to nearly full or discharging it nearly complete) degrades it faster. And I think there’s other chemistry based effects that speed it up.
The plots I have seen look more like this.
And from common use, you know this. If you’ve ever had a cell phone or laptop long enough to wear out the battery, you’ve noticed it goes from somewhat diminished to an off the cliff degradation, eventually reaching a point where you cannot unplug a laptop from the wall without it instantly losing power.
This is what Prius owners who do have failed batteries experience, btw.
True, but note from that chart that the graphs are all remarkably similar early on–the charge rate affects when the dropoff happens, but before that point it’s quite linear.
Even the Superchargers stay below a 1C charge rate, and you can see from the chart that it’s still in the linear range at 500 cycles at 1C. Since most charging is at home at much lower rates, like under 0.1C, it’s liable to stay in that linear range for a very long time.
There are (obviously) several life-limited components that EVs lack. I drove my last car to a fully depreciated 185,000 miles. In addition to the lamps, tires, brakes, and suspension components common to any car my records say I went through the following.
An alternator, a starter, a voltage regulator, a fuel pump, a water pump, 4 half-shaft/CV joint assemblies (AWD), an accessory belt tensioner, two accessory belts, a couple radiator hoses, and a radiator tank - twice. (stupid Germans making hi-temp pressurized components out of eco-friendly almost-plastic. Not good.)
Other than filters and fluids, exactly zero maintenance was done on the engine proper, transmission, and differentials. Zero. And I was far from scrupulous about the fluid intervals.
None of my list will fail on an EV. The $64B question is how durable the stuff EVs have that ICEs lack will be.
There’s certainly the basic “solid state good, moving parts bad!” argument in favor of, say traction motor controller boards being more reliable than, say, water pumps. But consider how many threads we have on home white-good appliances where the standard refrain is “my old Kenmore washer with the mechanical timer dial lasted 20 years. My new one with the LCD screen and computer lasted 12 months before it needed a $400 computer part.” And that’s with bog-standard low power microcontroller & discrete parts used indoors.
Done properly, the EV-only components have this *potential *to be more reliable. Will they realize the potential? Actual in-use service times hundreds of thousands of units is a tough test. If you have a weak spot, that’ll find it. Another issue with electronics is they’re prone to bathtub curve failures. The stuff that, due to undetectable internal defects, will burn-in and factory test just fine, but fail at 5% of intended life.
If there’s a situation where 5% of some model of car fails the XYZ module and leave the motorist stranded by 15K miles but the other 95% sail through without a hitch to 200k miles, that’ll be bad. If this happens to an ordinary car by an ordinary manufacturer, Consumer Reports will tutt-tutt, the whiners on the internet will have a field day, and the engineers will need to release v2.0 of that part.
If this same thing happens to a company that’s already seen as gaffe-prone and struggling to climb the “we’re for real” curve of public perception, well … Tesla’s tightroping over a bigger canyon than is, say, GM.
Lastly consider the, say, alternator of an ICE. They’re conceptually simple and not too hard to make. They’re pretty highly optimized, but mostly optimized for least cost given their durability target. There is no bleeding edge tech in them.
To what degree are the various high power electronics (and batteries) pushing the states of their arts in terms of size, level of cooling, vibration resistance, etc? While still trying to fit under the $ cost limbo bar?
A lot of engineering is applying lessons from the school of hard knocks. It’s not all applied physics calcs. No organization has lots of experience with this stuff yet. The opportunities for an honest unforced error are larger with the bleeding edge tech than they are with established tech. Boeing had a lot more trouble getting the 787 out the door than they did the 777 and in their field they’re the preeminent engineering team on Earth. This stuff is hard.
Said another way, the expected value of durability for EVs vs ICE ought to be close. But the expected variance, how well any given model meets that goal, ought to be higher (at least at first) for EVs vs ICEs. IOW v(EV[sub]EV[/sub]) > v(EV[sub]ICE[/sub]) 
Gonna be fun to watch.
er, EVs currently have half shafts with CV joints, and water pumps and coolant hoses. Per the Model S owner’s manual, Tesla requires brake fluid changes every 25,000 miles and coolant changes every 50,000 miles. oh, and the gearbox oil needs to be changed at 12,500 miles.
OK. That’s what I get for accepting at face value the EV hype that so many moving parts simply don’t exist. :smack: Trust me when I say my overall POV is that greater fleet reliability is mostly hype now and may, in fact, come to pass after enough fleet experience. Eventually.
I also left one chunk off my last overlong post. Namely claims of increased brake durability.
The existence of regen braking means the friction brake system has to absorb less total energy in any given use and over its total life. So far, so true.
But if properly engineered, that will translate into the friction brakes being smaller & lighter & cheaper, resulting in the same durability. Why? Because the target is a certain durability. The engineers will (eventually) hit the target they’re given and there’s no a priori to assume EVs will have a higher brake durability target than ICEs.
Don’t forget the radiator. You ninjaed me, but where do you think the waste heat from supercharging an EV battery pack at 140 kilowatts goes? Charging is only about 95% efficient, meaning 5% of 140 kilowatts has to go somewhere. Similarly, on discharging, ends up being even more waste heat.
It’s a smaller radiator, but on the other hand, at least in the Tesla S, there is a cooling tube going by every individual battery cell. Hundreds of feet of total tubing, maybe miles. Going all through the car - the chargers are in the rear of the vehicle, but are being cooled by the radiator in the front. Not one but two water pumps. That are electric and variable speed, versus the chain driven things they put in economy cars. Allegedly there is even plumbing to let the air conditioner cool the coolant, though tbh that sounds ridiculous.
The main argument for increased EV reliability is actually one for autonomous vehicle reliability. Autonomous cars probably will be a ton more reliable - because their owners will want them to last a million miles as a taxi. So that engineer you mentioned above will do things differently. But that’s not an EV vs ICE thing - you could engineer ICE drivetrains to last a million miles. Big trucks do just fine. You might have to use diesel for the engine, but maybe not.
Agree with your overall points. Youch on the intricate cooling systems; every junction between components is a point of failure.
Ref the snip I’m not as sure. Or at least I’m going to posit another life-limiting factor: the interior. As a point of comparison we expect to replace much of the interiors in airliners several times over the economic life of the flying machine. Partly for wear and tear and partly for style and creeping feature-itis.
Riding in a modern USA taxi is a grungy experience akin to using a convenience store bathroom; one to be tolerated briefly and rarely only when otherwise unavoidable.
At least the early days riding in an Uber et al was different. The lack of grunge was a huge attraction point. That may still be the case but I don’t ride them often enough to have a reliable fresh sample.
A public shared, as opposed to individually owned, SDC fleet will be popular exactly to the degree it’s clean, reasonably new, non-smelly, and inviting *every *time you summon one. Otherwise fat chance getting the comfy suburban set to adopt them, and that’s where all the money and all the volume is/will be. At least after you get past the denizens of NYC who expect everything around them to be filthy & ratty. They’d jump at anything cleaner than the NYC subway.
My totally IMO bottom line: There’s no point in engineering the machine to last 5 years of intense use (like a Freightliner) if the interior looks or smells like crap in 4 months. Unless that interior is also a readily and affordably replaceable module.
let’s leave RVs out of this 
It’s also a completely sealed system. Most of that complexity is in a hermetically sealed box.
A car coolant system is only semi-closed, as evidenced by the overflow reservoir, pressure release cap, etc. Like most auto stuff, it’s pretty reliable these days but historically it’s been a problem area and definitely not perfect.
The Tesla cooling system (or any EV system, really) operates under milder conditions than an ICE and is about as maintenance free as one can get. You’ll never find bottles of battery radiator fluid at WalMart because if it ever did need to be filled, it would mean something very serious has gone wrong.
Incidentally, the battery heater on the Model 3 is just the motor. Instead of having a separate heater element like on the S, the 3 just runs the motor inefficiently if the battery needs more heat.
One day, maybe not–if wheel motors ever become popular. Too expensive, low performance, and too much unsprung mass at the moment, but maybe it’ll make sense in some applications eventually.
Here’s a point for Tesla over the potential competition - I clicked this article expecting to read about how Mercedes Benz was without much fanfare jumping into the electric delivery vehicle space in the way that Tesla was moving into the long haul space with Semi. Yup, offering "electric drive options on all its commercial van model lines. This will start with the mid-size eVito, with deliveries commencing in the second half of 2018. "
But the range only 93 miles and if fully loaded and not great weather maybe only 62 miles. Recharge 6 hours. 6 hours?! Is that range and recharge rate really something that makes an EV option attractive?
The range and charge time have to fit within the parameters of a trip. Put another way, those are the limiting factors (60-90 miles and a turn time of 6 hrs). I’m in the transportation business and that sounds REALLY limiting.
Maybe, just maybe, that works well for a city delivery van in a dense urban Euro-environment. It certainly amounts to 93 (at best) miles of travel per work day.
I wonder how many miles UPS puts on one of their “parcel cars” per day in a US urban or suburban environment? At 20 mph on surface streets the worst-case 62 miles is about 3 hours of travel. At the best case it’s 4+ hours of travel. Most delivery vehicles start out heavy and end their day much lighter. So even if it leaves base at max legal gross weight it won’t average that all day.
In an (assumed) 8 hour shift, how much time is the driver loading the truck at base, dashing in and out of buildings, up and down in elevators, deciding where to go next, finishing his/her work day, etc. vs actually driving from location to location? Hmmm.
It *might *not be as badly under-ranged as we think at first glance. As to slow charging …
For a vehicle that works one shift per day and sits the rest of the time, once you design the battery capacity to get through the 99%ile shift, there’s no upside to quick charging. As has been beaten to death upthread, quick charging is hard on battery life. Fleet operators are all about lifecycle cost, and a leisurely charge cycle that extends battery life by 10,20, 30% (?) would be a win, not a loss.
Finally, imagine installing and running the recharging yard where e.g. 40 of these things juice up overnight. You can either pay for 40 supercharger-like gizmos and the amperage feed to run them all at once, or you can pay for 40 sorta-trickle chargers that get the same job done over 6 or more hours. Which is cheaper to install and operate? IMO slow and steady wins that race.
Bottom line: ISTM those limitations are not driven by the bleeding edge of the tech. Instead they’re considered features flowing from the target mission. These things are evidently not intended as rural or over the road long haul machines.
I searched around a bit for figures and found this thread from a UPS driver’s forum:
Lots of numbers in there. There are certainly a few in the ~60 mile/day range, but somewhere around 120-150 miles/day seems way more common. As you say, maybe it makes more sense in a dense city environment.
Others have addressed the other points, but this is basically what an ICE is now. With the right tools, you can (and often do) replace transmissions, engines, 12VDC batteries, etc.
I just took my 14 year old truck in to have the steering shaft replaced (I couldn’t free the bolts myself). It doesn’t matter that it’s an ICE or PHEV or BEV. The universal joints were bad, and the shaft was replaced. I’ve not had any engine work done, yet. In the way of negative, my previous pride-and-joy blew a spark plug out of her cylinder head (probably due to negligence at 100K when I had the plugs replaced). That wouldn’t happen with a BEV, obviously, but damn, other than that, it was constant front-end and suspension work – something not unique to an ICE car.
Slightly more rapid charge even might even allow enough recharging at reloading times to get several delivery runs of 40 miles in.
Meanwhile sometime problems for Tesla that may be out of their control: cobalt supplies may get tight. Now as per that article such does not mean they are doomed, but it would require some new adjustments to plans. They don’t use lots of cobalt but supply demand is always relative. For now they need what they need and if what they need is more than is easily supplied it is immaterial that the absolute amount is not huge.
Not hundreds let alone thousands of vehicles but it is a start.
