5 ms·

Once again, I am reminding HackerNews that the technology to build a battery capable of storing enough renewable electrical energy for the (world|nation) for ev

Once again, I am reminding HackerNews that the technology to build a battery capable of storing enough renewable electrical energy for the (world|nation) for even half a day *does not exist* at any reasonable cost. And if you want to store multiple days for a northerly nation with very cold winters, frequent high pressure anticyclones (so, no wind) that can last about a week, and you want to switch everyone to zero carbon heating, then the technology doubly doesn't exist. And the only retort to the above will be mumbling "yeah, but exponential improvement in batteries plus didn't someone say something about hydrogen?" which is essentially, wishful thinking. When you can build a zero carbon grid out of nuclear fission plants - and we've known how to do so since the 60s.

VaxWithSex 4y ago
Sure it exists, it is called compressed air. Even better with CO2. Close to me is the oldest one, built in 1972 and still operational today: https://de.wikipedia.org/wiki/Kraftwerk_Huntorf https://de.wikipedia.org/wiki/Kraftwerk_Huntorf
rapsey 4y ago
> Once again, I am reminding HackerNews that the technology to build a battery capable of storing enough renewable electrical energy for the (world|nation) for even half a day does not exist at any reasonable cost. But it is almost certainly closer to existence than fusion.
- gjsman-1000 4y ago
  Almost certainly not. The US alone generates 4,095 billion kWh yearly. For a half a day, you would need to store 5,600,000,000 kWh. Tesla Megapack can store 3916 kWh fully loaded. This means you would need 1,430,000 Megapacks to power the US for half a day. With Tesla only being capable of producing roughly 40,000,000 kWh of Megapacks annually, it would take 140 years to produce all the batteries. If Tesla created 100 times the factory capacity they have now (which, could the supply of raw materials even withstand the smallest fraction of that?), it would take 14 years, for batteries that have a warranty of 15 years. These are lithium-ion batteries which are the most space-efficient, unless you don't mind clearing hundreds of square miles of space for this project. Did I mention it costs about $1 million per Megapack right now, so this project would cost $1.4 TRILLION assuming all Lithium+Cobalt+Supplies+Labor cost the same as they do now despite demand being increased 100x, and ignoring all engineering costs, and factory scaling costs, which could multiply the cost exponentially. All to power the US for just half a day. Now consider how to add Europe, Asia, Africa, South America, the rest of North America... We're not close, and it's basically completely unfeasible. Fusion will be closer in 100 years than such a project.
  - rapsey 4y ago
    I have listed 5 different types of batteries than Tesla makes. A number more are much farther than the fundamental science stage of Fusion. Tesla primarily makes batteries for cars, grid storage is actually way more flexible in the type of battery that can be used. You are missing the forest for the trees.
  - pfdietz 4y ago
    Almost certainly yes. Consider pumped thermal energy storage. Use a thermal cycle to generate hot and cold (say, by compressing a gas, probably argon, extracting the heat, then reexpanding, and then storing the resulting "cold"), then reversing that cycle to generate power. This scales embarrassingly well. It can be made entirely from cheap materials available in essentially infinite supply. No component operates at a temperature above the creep limit of ordinary steel. Round trip efficiency could reasonably be 75%. This requires no technological breakthroughs -- it's 19th century technology.

VaxWithSex 4y ago

Sure it exists, it is called compressed air. Even better with CO2. Close to me is the oldest one, built in 1972 and still operational today: https://de.wikipedia.org/wiki/Kraftwerk_Huntorf https://de.wikipedia.org/wiki/Kraftwerk_Huntorf

rapsey 4y ago

> Once again, I am reminding HackerNews that the technology to build a battery capable of storing enough renewable electrical energy for the (world|nation) for even half a day does not exist at any reasonable cost. But it is almost certainly closer to existence than fusion.

gjsman-1000 4y ago
Almost certainly not. The US alone generates 4,095 billion kWh yearly. For a half a day, you would need to store 5,600,000,000 kWh. Tesla Megapack can store 3916 kWh fully loaded. This means you would need 1,430,000 Megapacks to power the US for half a day. With Tesla only being capable of producing roughly 40,000,000 kWh of Megapacks annually, it would take 140 years to produce all the batteries. If Tesla created 100 times the factory capacity they have now (which, could the supply of raw materials even withstand the smallest fraction of that?), it would take 14 years, for batteries that have a warranty of 15 years. These are lithium-ion batteries which are the most space-efficient, unless you don't mind clearing hundreds of square miles of space for this project. Did I mention it costs about $1 million per Megapack right now, so this project would cost $1.4 TRILLION assuming all Lithium+Cobalt+Supplies+Labor cost the same as they do now despite demand being increased 100x, and ignoring all engineering costs, and factory scaling costs, which could multiply the cost exponentially. All to power the US for just half a day. Now consider how to add Europe, Asia, Africa, South America, the rest of North America... We're not close, and it's basically completely unfeasible. Fusion will be closer in 100 years than such a project.
- rapsey 4y ago
  I have listed 5 different types of batteries than Tesla makes. A number more are much farther than the fundamental science stage of Fusion. Tesla primarily makes batteries for cars, grid storage is actually way more flexible in the type of battery that can be used. You are missing the forest for the trees.
- pfdietz 4y ago
  Almost certainly yes. Consider pumped thermal energy storage. Use a thermal cycle to generate hot and cold (say, by compressing a gas, probably argon, extracting the heat, then reexpanding, and then storing the resulting "cold"), then reversing that cycle to generate power. This scales embarrassingly well. It can be made entirely from cheap materials available in essentially infinite supply. No component operates at a temperature above the creep limit of ordinary steel. Round trip efficiency could reasonably be 75%. This requires no technological breakthroughs -- it's 19th century technology.