4 ms·

The money has been on the table for a long time (considering how many places lithium batteries are used in) yet still no considerable improvement. I really wish

by dharma1 11y ago

The money has been on the table for a long time (considering how many places lithium batteries are used in) yet still no considerable improvement. I really wish 10x energy density would be possible in the next few years, but 10x is definitely in the lithium/aluminium-air territory, and don't see any indication of that being commercialised within 2-3 years

mikeash 11y ago
If the last graph on this page is correct, then lithium batteries have seen at least a 2x improvement in all the important metrics since they've been introduced: http://batteryuniversity.com/learn/article/battery_statistics http://batteryuniversity.com/learn/article/battery_statistic... (I'd be interested to see a graph that includes the last ten years....) I'd call that a considerable improvement, albeit nowhere near 10x.
- dharma1 11y ago
  You are right, on average doubling every 10 years or so - I suppose that is considerable improvement, though still frustratingly slow compared to advances in transfer speeds, computing power, disk/memory capacity etc. It would be fantastic if this was a problem where solutions are accelerated significantly (ie. 10x jump in a few years) by throwing more money at it, but I am not sure it's the case. Here is a relevant Quora thread: https://www.quora.com/Is-it-true-that-battery-energy-density-improves-5-8-per-year https://www.quora.com/Is-it-true-that-battery-energy-density...
  - mikeash 11y ago
    Frustratingly slow considering how important batteries are, too. It's interesting to think about what would happen if you had some sort of magic battery, say a 1kg 10cm^3 cube that cost a dollar to make and held 1000kWh of energy. The changes would be huge. The electric grid would become vastly more efficient, wind and solar would suddenly become vastly more practical, virtually all cars would become electric, etc. Compare to a similarly magic CPU. That would be cool, but it wouldn't change the world. (Until somebody figures out how to run superhuman AI on it, anyway.)
    - masklinn 11y ago
      > a 1kg 10cm^3 cube that cost a dollar to make and held 1000kWh of energy 1MWh is 3600000000 or 3.6GJ. 3.6GJ/L is twice the energy density of Uranium, and 3.6GJ/kg is 40 times the specific energy (energy/mass). We wouldn't need a power grid anymore because we could just charge batteries on the moon and lob them to earth. If we're in dreamland, we can just think about what'd happen with instantaneous subspace energy transmission. We could get rid of batteries and electrical grids altogether and we'd just turn the moon, mars and mercury into giant solar arrays. I'd wager that's more likely to happen than a 1L/1kg 3.6GJ battery too.
      - mikeash 11y ago
        Turning the moon into giant solar arrays isn't useful because it's really expensive to put stuff up there and hard to maintain it, not because of the difficulty of transmitting the resulting power. I mean, it is difficult, but not that difficult. Even with magic power transmission technology, you'd be better off covering parts of the Earth with solar panels, rather than celestial bodies.
        masklinn 11y ago
        If you can essentially store unlimited energy, there is no need to "put stuff up there", you can just go and shuffle it around from the rest of the solar system. Putting it anywhere other than earth avoids further damaging the planet's ecosystems, and is a step towards the direct exploitation of the sun itself.
      - trhway 11y ago
        >> a 1kg 10cm^3 cube that cost a dollar to make and held 1000kWh of energy >1MWh is 3600000000 or 3.6GJ. 3.6GJ/L is twice the energy density of Uranium, and 3.6GJ/kg is 40 times the specific energy (energy/mass). We wouldn't need a power grid anymore because we could just charge batteries on the moon and lob them to earth. well, these densities is fusion energy. No need to charge at all. Just scoop up some He-3 on the moon :)

by dharma1 11y ago

mikeash 11y ago

If the last graph on this page is correct, then lithium batteries have seen at least a 2x improvement in all the important metrics since they've been introduced: http://batteryuniversity.com/learn/article/battery_statistics http://batteryuniversity.com/learn/article/battery_statistic... (I'd be interested to see a graph that includes the last ten years....) I'd call that a considerable improvement, albeit nowhere near 10x.

dharma1 11y ago
You are right, on average doubling every 10 years or so - I suppose that is considerable improvement, though still frustratingly slow compared to advances in transfer speeds, computing power, disk/memory capacity etc. It would be fantastic if this was a problem where solutions are accelerated significantly (ie. 10x jump in a few years) by throwing more money at it, but I am not sure it's the case. Here is a relevant Quora thread: https://www.quora.com/Is-it-true-that-battery-energy-density-improves-5-8-per-year https://www.quora.com/Is-it-true-that-battery-energy-density...
- mikeash 11y ago
  Frustratingly slow considering how important batteries are, too. It's interesting to think about what would happen if you had some sort of magic battery, say a 1kg 10cm^3 cube that cost a dollar to make and held 1000kWh of energy. The changes would be huge. The electric grid would become vastly more efficient, wind and solar would suddenly become vastly more practical, virtually all cars would become electric, etc. Compare to a similarly magic CPU. That would be cool, but it wouldn't change the world. (Until somebody figures out how to run superhuman AI on it, anyway.)
  - masklinn 11y ago
    > a 1kg 10cm^3 cube that cost a dollar to make and held 1000kWh of energy 1MWh is 3600000000 or 3.6GJ. 3.6GJ/L is twice the energy density of Uranium, and 3.6GJ/kg is 40 times the specific energy (energy/mass). We wouldn't need a power grid anymore because we could just charge batteries on the moon and lob them to earth. If we're in dreamland, we can just think about what'd happen with instantaneous subspace energy transmission. We could get rid of batteries and electrical grids altogether and we'd just turn the moon, mars and mercury into giant solar arrays. I'd wager that's more likely to happen than a 1L/1kg 3.6GJ battery too.
    - mikeash 11y ago
      Turning the moon into giant solar arrays isn't useful because it's really expensive to put stuff up there and hard to maintain it, not because of the difficulty of transmitting the resulting power. I mean, it is difficult, but not that difficult. Even with magic power transmission technology, you'd be better off covering parts of the Earth with solar panels, rather than celestial bodies.
      - masklinn 11y ago
        If you can essentially store unlimited energy, there is no need to "put stuff up there", you can just go and shuffle it around from the rest of the solar system. Putting it anywhere other than earth avoids further damaging the planet's ecosystems, and is a step towards the direct exploitation of the sun itself.
    - trhway 11y ago
      >> a 1kg 10cm^3 cube that cost a dollar to make and held 1000kWh of energy >1MWh is 3600000000 or 3.6GJ. 3.6GJ/L is twice the energy density of Uranium, and 3.6GJ/kg is 40 times the specific energy (energy/mass). We wouldn't need a power grid anymore because we could just charge batteries on the moon and lob them to earth. well, these densities is fusion energy. No need to charge at all. Just scoop up some He-3 on the moon :)