Hacker News new | ask | show | jobs
by blake1 1208 days ago
The stats are not crazy for an air battery.

But the catch is two-fold. First, are they weighing the battery before or after discharging? Oxidizing will change the weight significantly. The most honest result would be the average weight during the cycle.

The other catch with these air batteries is usually the purity requirements on the intake. I recall reading about earlier experiments that could not tolerate pollen, dust, and smog, and required an energy intensive purification step (maybe involving cryogenics) that was a nontrivial power draw.

Those caveats aside, a back of the envelope estimate for the energy density would be something like 600-700 Wh/kg.

2 comments

The other issue is that metal-air batteries go by another name: fuel cells. You are power-gated by how much oxygen you can deliver to the battery. As a result, metal-air batteries/fuel cells are either very slow to discharge, or have big air pumps to have decent performance.

And of course, they will release oxygen if you try to charge them, which implies a way of rapidly expelling air when charging up quickly. Many past attempts avoid this problem by "mechanically" charge up the battery, meaning literally swapping out the spent chemicals with new ones. This of course require an auxiliary battery if you want regenerative braking or the ability to electrically charge.

And of course, the real catch is that we've already invented the metal-air battery in a practical way: hydrogen fuel cells. The big advantage with them is that mechanically recharging is very straightforward compared to other mechanisms. All other attempts are basically reinventing the wheel or have a very specific niche in mind.

A problem with fuel cells is that if they're acid, they need expensive platinum group element electrodes. If they're alkaline, they can use nickel, but then they need to have the CO2 scrubbed from the air or they clog up with carbonate.
But there's something to be said for a fuel that's solid under ambient conditions rather than being literally the least-dense substance on Earth.
Methanol fuel cells exist too.
> a back of the envelope estimate for the energy density would be something like 600-700 Wh/kg.

https://www.google.com/search?q=current+tesla+battery+energy... says Tesla batteries are currently in the range of 270-290Wh/kg

Is that accurate to you?

As in this is roughly at least twice as good as current technology? Seems too good to be true. When can we expect to see it hit consumer cars? 5 years? 10 years?

Except tesla is expecting and has been achieving 10% density increases in the last few years. Do this for 8 years, and you get double, same as the new battery tech that's gonna take 10 years.

1st gen: 276 Wh/kg (2022)

2nd gen: 305 Wh/kg (2023)

3rd gen: 333 Wh/kg (2024)[1]

Here is a cool article on all the tech that is scheduled or went into these new 4680 batteries and getting the energy density up well past 300.

https://cleantechnica.com/2020/09/22/everything-you-need-to-...

[1]https://insideevs.com/news/598656/tesla-4680-battery-cell-sp...

Your info is out of date. The 4680 actually ended up being 244Wh/kg[1], which is lower than the Panasonic 2170 at 269 Wh/kg that they were already using.

[1] https://www.youtube.com/watch?v=4XOHetABrag

Still, Tesla with 4860 is 9kg lighter than Tesla with 2170.

https://insideevs.com/news/600297/tesla-model-y-4680-is-just...

I assumed Tesla was at 300, which is pretty close. This might not be the tech for consumer cars though, but one thing I have seen is the enormous proliferation of applications for Li batteries as the costs have come down, and these will certainly find an application. That would be my bet.