Hacker News new | past | comments | ask | show | jobs | submit login
Tesla’s readying a ‘million mile’ battery that could greatly lower cost of EVs (theverge.com)
89 points by caution on May 14, 2020 | hide | past | favorite | 27 comments



This article is pretty low quality. The claim that the new million mile battery technology is coming to the China Model 3 first is not supported by a source and contrary to what people who follow this stuff closely have been saying.

The partnership with CATL is rumored to be for providing Lithium Iron Phosphate (LFP) cells for made in china Model 3 & Y. The million mile battery technology is rumored to be a combination of technologies that Tesla will use to manufacture batteries themselves w/o a partner. The million mile battery will probably be reserved for their higher end models in the near term while they continue to use Panasonic and CATL for some time while they ramp up production of their own cells.

The best source for info on all this that I've found is YouTube channels like Hyperchange

[0] https://www.youtube.com/watch?v=KB-yc-lVExw


This article is pretty low quality.

Yes. First, who's doing this? It sounds like Contemporary Amperex Technology Ltd in Ningde, Fujian Province, is providing the batteries. But there's no press release from them.[1] Also, their stock went down a bit today,

CATL is the world's largest lithium battery manufacturer. They produce about 40GWh of battery capacity a year, compared to Tesla's 35GWh or so. CATL has been roughly doubling production each year for the last few years. Their output is not exclusive to Tesla. They've mostly targeted China and European car manufacturers. They're building a huge battery plant, bigger than Tesla's "gigafactory", in Germany.[2]

If this is real, it's all about CATL.

[1] https://www.catlbattery.com/en/web/index.php/news/lists

[2] https://www.electrive.com/2019/02/04/catl-plans-up-100-gwh-b...


[flagged]


If you read [2] you can see they're predicting 2025. Many such predictions in the past have come and gone.


TheVerge is sourcing Reuters but Reuters does not name their source, either.

https://www.reuters.com/article/us-autos-tesla-batteries-exc...



What kind of lifetime do people assign Tesla's current-day packs?

It certainly seems like, based on the 450k mile "Tesla Taxi"'s experience, assuming no faults, they're certainly good for ~300k but that's just one data point (and of course they had a faulty battery, too).

IIRC the warranty on my Model 3 pack is for >70% retention after 8 years/100k miles, but in real world experiences, they don't seem to degrade much past the first 10k miles. We're over 30k miles and the life is basically exactly what it was at 10k.

But I don't know what the commonly-expected "pack lifetime" is. I have seen the owner surveys out to 100k-150k miles and the lifetime still looks good at that point. And of course they've been through a lot of revisions over the years.


My limited understanding is that battery degradation is not linear. It's gradual for a long time and then falls off rapidly.



I'd like to know the strategy for maximizing lifetime. If typical use is about 30 miles per day that leaves a lot of room for strategizing. Should I charge it every night, every other night or every third night? Should I keep it charged to 80% all the time, or should I let it go down to some minimum level first before plugging in?


Keep it above 40% and below 80% and you should be good. The damage is mostly done during charging so setting your max level to 80% and leaving it plugged in is about the best way to go.

The main thing to look at is the floor. Keep the battery from going too far below 40%. (I believe past 20% is where damage really starts to happen).

Other tips to keep the battery happy are things like driving slower, using chill mode, etc. Fast discharges damage the battery so eco driving is usually good for the battery.

AFAIK, LiPoly batteries don't really suffer from any sort of memory effect. So there is no harm in not fully discharging. The one catch is that your BMS can get out of whack if it doesn't see 10% and 90% for a while. So occasional charges up to 90% or 100% are not a bad thing to recalibrate the BMS.


The worst thing to a LI battery is let stay at 100% in hot conditions. Teslas have an AC cooler though, so I think they might be okay for hot weather.


An estimate last year claimed Tesla's costs were ~$160/kWh[1]. This article suggests by end of next year, <$100/kWh. In percentage terms that's a big improvement, but even going by last year's estimates, only $8k out of the $40k price tag of a standard Model 3 is actually accounted for by the battery. So if that $8k drops to $4k, I can imagine the car price dropping to say $35k. But not more than that.

[1] https://insideevs.com/news/400529/tesla-battery-costs-droppi...


> This article suggests by end of next year, <$100/kWh. In percentage terms that's a big improvement, but even going by last year's estimates, only $8k out of the $40k price tag of a standard Model 3 is actually accounted for by the battery. So if that $8k drops to $4k, I can imagine the car price dropping to say $35k. But not more than that.

You're comparing manufacturing costs to retail prices. Presumably the battery has some kind of a markup on top of their production cost.

On top of that, you sell more cars if they cost $4000 less, right? So then you get more economies of scale, which lowers unit costs, which allows lower prices, which causes you to sell more cars, so then you get more economies of scale.


I thought Tesla made the claim last year that they hit the $100 per kwH target.


I know Tesla has the best batteries of an EV in mass production, but that price seems really high. I found this article last year talking about VW paying <$100/kWh.

https://www.electrive.com/2019/09/09/id-3-batteries-cost-vw-...


Batteries vs Battery Packs. Lots of extra stuff goes into the pack(cooling, wiring, frame, monitoring/charging electronics ext). Not sure exactly what Tesla's breakdown is, offhand don't think I've seen them release ever release the info.


This article's shadiness aside..

Battery tech just continues to amaze me. Right now the cheapest EVs are in the same price ballpark as mid-range gasoline cars, and range anxiety is fading fast.

I always thought that EVs would get "as good" as gas cars eventually but now I'm realizing that they may surpass them by a long shot. In the not-too-distant future an EV may cost way less, require far less maintenance, be cheaper (and greener!) per mile, be faster and safer, and have a longer usable lifespan, than a gas car. Amazing stuff.


> In the not-too-distant future an EV may cost way less, require far less maintenance, be cheaper (and greener!) per mile, be faster and safer, and have a longer usable lifespan, than a gas car.

While this sounds great I seriously doubt that this is going to happen as it is directly against interests of car manufacturers. As confirmation I like to bring Nissan Leaf example as it is on the market for a long time and is suffering from battery degradation - and guess what - despite fact that car is working fine battery replacement is not easily obtainable and not cheap, examples:

https://www.autonews.com/sales/nissan-leaf-buyers-dealers-wo...

https://www.reddit.com/r/leaf/comments/9l0pjo/first_gen_leaf...


What's interesting is look at the ratio of lithium to everything else that makes up a lion battery. There is in theory a lot of room for improvement. Then refigure current electric cars but with batteries with 50% more kwh/kg. The EV weight penalty disappears.


We don't really know all this stuff yet. Tesla will have a 'battery day' where we will get actual information. So far this is all speculation. I have heard so many different potential options and improvements, non of it is that well founded.

However Tesla does certainty have some strategy vertical integration and they are doing interesting things. They know that batteries is the competitive advantage and they are pushing that very hard.


Does that mean a single battery could be reused for a second car when the first car is worn out?


Based on their presentation they are building the cars to last that long. All the electronics and everything else is designed to last.

If you look at the videos of the Model Y breakdown you can see that they did a lot of stuff that is slightly more expansive but should improve lifetime.

They want to be able run these vehicles into the ground making profit for them, specially after coming of leases.


Same thing happens with ICE vehicles. Engine swaps are common on models where the engine outlasts the body- e.g. snowy states where cars rust out, and dirt bikes where the bike is often totaled in a crash long before it wears out and then parted out as spares for racers who blow engines.


I've seen transmission swaps more than engine swaps - they seem to fail more frequently in ways that they can be rebuilt. As for the frames themselves, I think modern car frames last a lot longer than the did in the 70s and 80s, even in snowy climates.


Probably, though I suspect as energy density continues improving you'd likely always want the latest mass-produced tech for mobility use even if your existing battery has plenty life left.

The used car batteries seem likely to transition into stationary uses like grid power, where density matters less.


Not necessarily, my Spark EV has older less dense batteries and has only 80 miles of range. It is still perfect for it’s role as a commuter and around town car, so there’s no reason to upgrade.


note - article is from the verge




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: