Hacker News Reader

Comment by baybal2 | Hacker News Reader

> But when batteries cost less than $100 per kWh

You were already able to get LFP batteries outside of China for a bit cheaper than that for 2 years at car maker wholesale quantities.

Battery cells plummeted as a part of vehicle cost.

The cheapest cost I ever heard quoted in China was $64 per kw/h. I don't preclude that battery makers with own car factories like BYD have real costs around $30-$40.

Taking cost at $40, BYD Dolphin had $1200 in it, comparable to costs economy class IC cars have in their whole powertrain.

But EV vehicles are on overall easier to engineer. Part counts of EVs are much smaller than IC cars.

More importantly, you can deliver satisfying driving experience with much smaller car if you use electric traction, while <999cc engines will have small powerbands no matter what amount of smarts is added to their ECUs.

I always tell that 1-to-1 IC, and EV comparisons make little sense. EVs are a whole different class of vehicles.

View on HN

> $64 per kw/h

Putting a slash in the unit there makes this read as "kilowatt per hour", which really doesn't make sense.

Think of it as the power the battery can continuously provide (in kilowatt) multiplied (not divided!) by the time it can sustain that power (in hours). Hence "kWh", "kilowatt hour".

In other words theoretically a 100 kWh battery could supply 100 kW for an hour, or 200 kW for half an hour, or 2 kW for 50 hours.

It's kW⋅h.

I think patent licensing is a bit of a concern. As I understand it, a lot of the big battery manufacturers in China have an agreement that the don't need to pay licensing fees for batteries used domestically, but if a major car manufacturer wanted to import LFP cells to, say, the U.S. they might have to pay the licensing fees or risk getting sued. (That might not be 100% correct, I'm kind of piecing things together from random Internet sources.) That may be moot soon as the relevant patents expire, which could change the battery landscape hugely (I hope).

Another thing is that not all LFPs are necessarily ideal for automotive use. For instance, batteries that can only safely discharge at 1C might be a bad idea to use unless the vehicle has a large enough battery pack that the load is spread out over a lot of cells. So, price isn't everything. Obviously, high energy density is also much desired for use in cars.

That said, I think LFPs are probably the ideal technology right now for replacing most of the world's combustion-engine ground transportation infrastructure with battery-electric vehicles. We just need to manufacture them at the necessary scale, and right now it's only China that's doing it. (It's like if Saudi Arabia discovered oil and the rest of the world said, "Well, I guess Saudi Arabia just owns the whole oil market. No reason to invest in it ourselves, because we can just get it from KSA." In the 20th century that would have obviously been a strategic blunder, but that's what the world is basically doing now with batteries. Whoever can make batteries cheaper than everyone else can use that to dominate the energy industry going forward. Fortunately there are fewer moats; anyone can make batteries if they can get the ingredients, and LFP ingredients are mostly not that rare except maybe lithium and copper.)

IIRC, global LFP patents are due to expire in the next year or two, allowing production everywhere.

LFP is great for cheaper vehicles and awesome for grid storage (which is not weight sensitive). The vast majority of batteries in the near-term transition will be LFP due to their common materials and super cheap cost.

Looks like the important ones have expired already:

https://www.youtube.com/watch?v=F_GtSA4Ig9s

"A Brief History of LFP, Patents, Licencing Costs, Pricing // and Tesla" by The Limiting Factor

According to that video, almost all of them have expired but there's one more that might be important that expires on April 27, 2022.

Meanwhile a 1kW/h battery for an e-bike can still retail for $2000 (thinking of the Stromer BQ983).

Surprises me how long it takes for these prices to tickle down into normal consumer goods. But they will, the true revolution of electric assisted urban transport has only just begun. Really exciting to see all the wacky new transport modes that are created left and right, I wonder which ones will stick

Consumer goods have all sorts of regulatory barriers to get through and you’re paying for it. If you get a spot welder and buy cylindrical cells wholesale, you can build dirt cheap li-ion packs. LFP really doesn’t buy you any safety for such a tiny (1kWh) pack.

Tiny branded cellphone batteries used to retail for $100+ 20 years ago.

Almost all e-bike batteries sold to consumers are ridiculously overpriced[1] relative to their components. Especially those from reputable brands like Bosch and Shimano which supplies most of the better e-bike brands on the street.

It’s a combination of having relatively well-off consumers, who lack the time/knowledge to start messing with their expensive machines and lack of generic spare parts due to the likes of Bosch locking down their components, and creating specialised computers (similar to how modern cars work) so that only authorised repair shops can work on them.

[1] https://ampedcycling.com/e-bike-battery-prices-what-can-i-ex...

But for anyone that can break the protection and manufacture a somewhat descent generic replacement battery the potential profit is enormous. I am sure some Chinese are well aware and hard at work on circumventing the DRM

It is unfortunately not that simple. The likes of Bosch have had a decade or two to learn how to lock up their automotive systems and they're using every trick in the book to do the same to e-bikes.

Perhaps, and then I hope they will be made obsolete as all the essential parts for making e-bikes are becoming commodities. 250W electric motors (max allowed in EU dor normal e-bikes) are not exactly rocket science. My jigsaw even has a 300W motor (though probably not rated for hours of continuous use)

Stromer is a Swiss e-bike company and one of the most expensive e-bike brands on the market.

A better example would be Lunacycle, which sells components for DIYing e-bikes. A 704Wh battery from them costs $550. $781/kWhr. Bafang is another well-known component supplier and sells a 754 Wh battery pack for $600. $795/kWhr.

Those are both 52v packs which are the most 'exotic' on the market at the moment, so there is a bit of a price premium.

In all cases, we're talking retail pricing for complete battery packs, with integrated battery management systems and packaging.

The article is talking about wholesale prices paid by companies buying these cells by the million unit quantity.

Until very recently Stromer was just about the only available e-Bike in the Netherlands that could reliably do to 45km/h,and thus quite popular. The retail prices for the 618/814/987 kWhr batteries are 1190/1650/1990, which to me indicates that the retail price is not at all related to any costs they make but that they are doing pure value pricing (whatever consumers are willing to pay). Their monopoly on proper speed pedelecs is rapidly coming to an end, so hopefully prices will soon get down to something more directly related to manufacturing costs.

It is already happening for normal e-bikes, where 3000+ used to be the starting range for a somewhat decent e-bike. This is currently about 2000+ but in a few years it will hopefully be possible to get a somewhat decent e-bike for 1000 - 1500.

Making batteries is like printing money right now. Everybody needs them and there is nowhere near the supply to meet the demand. Long term batteries will become a commodity and that's when purchase cost will drop dramatically.

The size of the car is in no way constrained by the engine size. People swap V8s into Miatas all the time. You can do a K24 swap into a Honda Fit without changing any bodywork. Vehicle size is entirely determined by market segment, design, and collision safety.

As for small engines, Fiat's 0.9 TwinAir is enjoyable once you get rid of the factory tune that is super-optimized for low emissions in the official tests at the expense of driveability. It even has extra-low rev limiters in the first three gears. Starting from the 105 bhp version you can get 120 bhp and 200 Nm of torque with only ECU tuning and removing baffles from the airbox, makes it a fun little thing.

> The size of the car is in no way constrained by the engine size.

It is. Nobody will buy a full-sized sedan with 800cc engine, unless you are in China.

There are a minimal size powertrains a car body need to be able to house for its size to be sellable.

The matter is you can put much smaller, and cheaper EV powertrains relative to the body size/weight for a much better driving experience.

You can use freed space to pile in extra niceties, which would've required bigger body size for an IC powered car, thus necessitating bigger, more expensive powertrain.

The attraction of hybrids was exactly that you can get a crap engine to provide driving quality of a much more expensive setup, like a good V6, or entry level V8 at lower cost, in a smaller car.

But "just better" is the worst enemy of good enough. You can throw away the whole IC altogether, and provide a 3L+ I6 experience in a micro car.

But why would you put an 800cc engine in a full-sized sedan? You have plenty of room to fit a decent turbocharged two-liter four-cylinder.

And there is no way EVs need less space for battery + motors in total. Have you seen the size of a Model X battery pack? Only difference is they typically use space below the floor of the car, rather than in front. They are typically a bit wider than ICE cars to compensate and give sufficient interior space.

> But why would you put an 800cc engine in a full-sized sedan? You have plenty of room to fit a decent turbocharged two-liter four-cylinder.

You will have to, and you say "The size of the car is in no way constrained by the engine size," which is a complete nonsense. On the lower end, it's constrained by sellability of a car, and on the upper end it's physically constrained by the size of car frame.

> Only difference is they typically use space below the floor of the car, rather than in front.

Yes, and that's a very big advantage. Even very tiny EVs can have batteries with hundreds of kilowatts of power output.

> On the lower end, it's constrained by sellability of a car

Yes, as already said, I completely agree.

> and on the upper end it's physically constrained by the size of car frame

In theory yes, but in practice you don't really feel this limitation, that's my point. As I said, you can fit a K24 engine into a Honda Fit (B-segment supermini car) without modifying the frame or bodywork. A turbocharged K24 can put out 300 kilowatts without breaking a sweat.

It is a limitation, i4 is an upper limit what you can put in it, and a 300kw i4 will vibrate itself to pieces in prolonged use.

Using bigger engines, for bigger cars really makes sense.

location and structure matter in system design. even if the battery takes up more total space than an ICE no one is going to notice the car being 1 inch taller overall for consumer vehicles.

Any idea what the depreciation on batteries is?

If you "used up" battery that's due to be sent to the recycler, How much of that 1200 could you hope to recoup? (I assume a power train can only be sold for scrap for a tiny fraction)

Bus batteries age very gracefully, but regular car often go to scrapper along with the vehicle. I don't know why.

I often see second hand cells from electric busses on 1688.

> second hand cells from electric busses on 1688

Can u give more details? Where is a good place to buy some 2nd hand LFPs for trial projects, ~50-500kwh quantities

1688.com

If it's an LFP battery, I think those are pretty easy to recycle but the materials weren't worth much to begin with. It's mostly just lithium, plastic, aluminum, fertilizer, iron, probably graphite, and copper.

Is it not the lithium that makes up the bulk of the 1200 price tag?

I don't think so, but I don't have specifics. There's some sources that a typical Tesla battery pack might have about 10-12 or so kg of lithium. LFP is a different chemistry, so it might have a different proportion of lithium but it's probably not hugely different. Lithium isn't particularly expensive (though it might become expensive in the future if demand goes up and production does not).

Absolutely not, and I have no idea why people think it does. It's like there's lithium in the name, so people instantly assume that it makes up a substantial portion of the cost. It does not.

Lithium is cheap, currently lithium hydroxide is at ~30$/kg. Of the total cost of a typical LFP battery, the lithium is ~2%. The rest of the materials round up to ~10%, the bulk of the cost is capital and labor.

I have bought shipments of 15kwh of LFP (Thundersky, SkyEnergy) for $0.5/kwh, delivered to USA, ten years ago. (EV conversions)

> $0.5/kwh

This seems to be wrong by an order of magnitude. $500/kWh is more plausible at that time.

Otherwise you are claiming that you paid $7.5 for a 15 kWh worth of cells, including the shipping, in ~2010.

Oh, yes, wrong units. $0.5/Wh => $500/kWh

Hmm, so $100/kWh is really good!

Reply on news.ycombinator.com