Hacker News new | past | comments | ask | show | jobs | submit login

You wouldn’t, that’s not my argument. People talk about the cost of replacing the battery at today’s prices, but in reality you’d be paying the future price. Given the 8 year warranty, expected 200,000 - 600,000 mile* lifetime [0] and the rapid decline [1] of battery prices, most people would never need to replace the battery and if they did, it would be in a future where they are cheaper.

[0] https://zecar.com/resources/what-are-lfp-nmc-nca-batteries-i...

> A key advantage of LFP is its longer life cycle, resulting in less degradation concerns. LFP packs are capable of more than 3000 full recharge cycle counts compared to NMC at around 1000 to 2000 cycles.

* Assume 200 miles of range on one charge at 1,000 cycles for NMC, intentionally lower to avoid controversy.

[1] https://www.goldmansachs.com/intelligence/pages/electric-veh...

> Battery pack prices are now expected to fall by an average of 11% per year from 2023 to 2030




Ah, sorry I misunderstood and I think we agree nearly completely. However, while I'm a fan of LFP, especially for UPSes, for cars it's a delicate balance of lower energy density vs longevity. What I can say is the the Li-Ion in my car have held up much better than expected, but I'm also babying them to the extreme, only rarely charge beyond 80% and rarely supercharging. At this point I fully expect to keep my car for 15+ years and never replace the battery.


LFP batteries are definitely becoming more prevalent. The standard range Teslas in some regions come with them. As you said, NMC should limit the charging to 80% for daily use.

Interestingly for LFP the recommended is 100% so that actually means that a standard range LFP Model Y can get more miles at 100% than the long range can at 80% for daily driving. There’s not much in it but it still feels weird!


The reason LFPs are becoming more prevalent is because LFPs have worse range but are cheaper, and we've reached the tail-end of the cost-vs-range bell curve. Or to put it another way, buyers are much more sensitive to 100miles vs 200miles, than they are 300 miles vs 400 miles. OEMs just don't need the extra range, not if it increases the pricetag.

It also does wonders for the supply chain - the N in NMC stands for Nickel, and most Nickel comes from Russia. The C stands for Cobalt, for that matter.


And Cobalt comes from artisanal mines in Congo that utilize child labor. So definitely another plus for LFP there.


And (no joke) the word Cobalt comes from the belief that Kobolds were defending Saxon silver mines with toxic minerals. We've known for a long time that Cobalt compounds are pretty toxic to miners.


80% of cobalt mines in Congo are owned and operated by the Chinese battery companies, such as CATL, the largest producer of LFP.


>> LFP batteries are definitely becoming more prevalen. The standard range Teslas in some regions come with them. <<

LFP batteries are fine (but barely) for entry-level, low-range EVs, or stationary energy storage systems (aka, ESS).

>> As you said, NMC should limit the charging to 80% for daily use. <<

All lithium ion batteries, including LFP, suffers accelerated degradation under high SOC, high C-rates; or in LFP's case, extreme temperature.

>> Interestingly for LFP the recommended is 100% so that actually means that a standard range LFP Model Y can get more miles <<

That 100% SOC charging causes a lot of wear and tear in LFP batteries, but it is still required to recalibrate the BMS to avoid misreading of SOC/range and other associated problems such as cell imbalancing issues. According to Cleanerwatt reports Oct 22, 2022, as a result of 100% charging:

  Tesla LFP Battery 10% RANGE LOSS PROBLEM? | Model 3 RWD https://www.youtube.com/watch?v=suw20wPrbL0
  The brand new LFP batteries will degrade substantially quicker.  There's not long-term retention data for LFP batteries on the market yet, but the trend tends to be substantially faster degradation.  Trends show them stabilizing around that 10% degradation mark in about half the time as non-LFP batteries - around 50,000 miles instead of 100,000 miles."
There is no long-term data yet, though BYD has been making/using much smaller LFP battery packs in their EVs since 2009, but these LFP packs aren't likely to last anywhere near what many LFP faithfuls blindly believe they would last.

The main driver behind LFP's popularity is the cost, especially during the past 2-3 years of COVID-19 which wrecked havoc in the auto supply-chain and inflated prices. China under their massive state subsidy and protectionism (banning foreign battery competitors and cutting all subsidies to EVs with their batteries since 2015) was able to mass-produce them at scale as LFP's share of EV market went from 2% in 2020 to 30% in 2022. China is also very much behind LFP because their ability to export EVs is limited by the fact that other leading battery chemistries such as NCM/NCA are largely guarded by the Japanese and the Korean battery makers -- China ensured that they couldn't get ahead in their domestic market by banning their business very early on and cutting all subsidies to EV with foreign batteries or since 2015 under Made-In-China 2025.

Otherwise, LFP is really a cheap, inferior battery chemistry for EVs.


If this argument were true: (A) a Prius replacement battery with now dinosaur NiMH tech would have fallen from $4k new to nothing. (B) any existing hybrid / EV would be getting bonkers range improvements w/ updated cell technology.

The best analogy I can think of: a battery cell / chemistry (akin to standardized RAM) is not equal to a battery pack (similar to a motherboard with soldered RAM).

For a variety of reasons (consumer preference, automaker greed?), unlike a bus / plane, a modern automobile is engineered to last 5 - 10 years (200k - 400k miles). The entire industry is predicated on selling new vehicles (much like smartphones, or computers) = no mainstream automaker is going to invest in building, testing, and then certifying a new battery pack to take advantage of new chemistry. You might cite Tesla's work on the Roadster as an example, but I'd argue this is an exception (a sendoff for their original halo car) rather than the norm.

& until we have a vehicle with enough scale, there exists no incentive for aftermarket to invest in building a new pack for a specific car. You could also argue, (& I'll concede) that Model 3 / Model Y might be the first car at scale to attract such investment – although I have doubts given (A) the barrier to entry for a pack is higher than a rando ICE part (e.g. would said company be able to build a BMS that talks to the vehicle?!) & (B) would any aftermarket company even try, given manufacturers will surely go out of their way to prevent this (we've already seen OTA disabling of features on EVs & I don't have to cite all the right to repair fighting that's ongoing…).

Is this a good thing? Absolutely not. Is it how it should be? Also no. But a lot of us (present party included) enable this when we want nice new things.

Until that changes, we're not getting cheaper battery packs for a legacy vehicle.




Join us for AI Startup School this June 16-17 in San Francisco!

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

Search: