3 ms·

As long as the earth turns, we'll need power at night too (and on cloudy days), and current energy storage devices are not good enough to cover all of that usag

As long as the earth turns, we'll need power at night too (and on cloudy days), and current energy storage devices are not good enough to cover all of that usage.

ben_w 4y ago
A true fact about what we built yesterday isn't a necessary limit to what we can build tomorrow. The storage tech works (as would a global power grid), we "only" need to scale it up, and at current prices a global grid would cost about the same as current annual spending on fossil fuels, while the cost of batteries depends on your estimate for storage requirements, which depends on the diversity (location and energy sources) of the grid, but could be in a similar range. (Batteries are also the most expensive of the seriously considered solutions, other storage methods exist and work just fine but aren't so good for cars and we need the batteries for cars anyway). I'm not sure how long either approach would last before needing servicing/replacement, but I'd be surprised if it was less than a year. Anyone here know how long underwater HVDC cables last?
- snovv_crash 4y ago
  That's assuming that prices stay the same for the materials needed to build the grid. Have you looked into the amount of copper required to build this grid? It is many multiples of the current annual production, this would cause prices to rise enormously.
  - ben_w 4y ago
    > Have you looked into the amount of copper required to build this grid? It is many multiples of the current annual production Yes, I have, and it is, even including aluminium (surprisingly a better choice for various reasons). I think I'm one of the main proponents of this on Hacker News, so you might have even seen my previous comments calculating this. WRT economics, not my field; all I can say is that higher demand increasing prices will encourage (but not necessarily ultimately create) an increase in supply, which has a difficult to forecast feedback relationship when planning for a specific total output.
    - snovv_crash 4y ago
      At least the initial rampup would be expensive - the cheapest copper is being mined already, so prices would necessarily need to rise to mine the more expensive copper too. How much it would rise long term once the new mines are running is geologically dependant. Aluminium has different issues, in that it is very very very energy intensive to refine. It may require new coal/nuclear power generation buildout near the mine for smelting the bauxite to make it even feasible at a large scale, otherwise you run into a chicken/egg problem with your grid idea since the grid doesn't exist yet. This is because aluminium smelters can't be shut down at night, they have a single turn on/off in their lifetime otherwise they need a full disassembly and rebuild, more likely scrap. So it can't just be run intermittently on local solar. To some degree it can be ramped down, so maybe a smaller base generation assisted by solar during the day, but it still needs significant power at night somehow.
      - ncmncm 4y ago
        Aluminum ore is super-abundant. There is no possibility of any country anywhere on Earth running short. The only per-ton expense to produce it is for electricity, which price will be falling very, very fast. So, no, we will not run short of aluminum, even at many times current production rate. Aluminum production process from ore + electric -> metal is extremely mature. Factories to make the factories have already been running for most of a century. They will operate at night on wind and, during lulls, drawing down storage. They could keep hot by burning powdered aluminum. New ones will be built preferably where wind is reliable, or (as now) hydro power is cheap. Or they might be built near a cliff, and operate their own pumped hydro. So, no, intermittent sources are not a problem for them or really for anybody, because storage is easy and cheap.
        snovv_crash 4y ago
        Absolutely, no shortage of aluminium, as I said, copper is the one where the mining is the bottleneck, for aluminium it is energy for the smelting. Night time energy storage for the smelters is not a solved issue still and significantly increases the costs beyond just adding sufficient PV panels adjacent to the smelter. Storage is not "easy and cheap" as you assert, otherwise there would be no reason to build the longitudinal grid connections GP was talking about.
- ajsnigrutin 4y ago
  But why not just use nuclear, which produces power 24/7 with no need for storage?
  - ben_w 4y ago
    • Cost (as much as the most expensive serious storage systems) • Domestic politics (NIMBYing for both reactors and waste, and you don't get around that just by telling people it's fine) • International politics (extra costs from proving nobody is using them for proliferation) • Build time (too slow compared to renewables) • Depending on your particular perspective, you may also care that the fuel itself is in limited supply (given the assumption that at some point even the currently-poor will want energy-hogging lifestyles and that nuclear reserves, even with breeders are 2500 ZJ compared to 65 ZJ of fossil fuels, then if you give everyone a Qatar per capita use that runs out in only about 300 years).
ZeroGravitas 4y ago
I'm glad you say "cover all" because I think people underestimate by quite a lot how far PV and Lithium batteries can get you if you're not near the poles. So today's tech, just with more deployment mostly solves this. Yes there's room for Wind and heat batteries and all those other things, but even the brutally stupid approach of Solar PV and lithium batteries that we have today, without any new breakthroughs, is good enough to cover most of it. People use more energy during the day, when the sun is up. Solar PV is the cheapest energy in history and broadly matches this daily profile.

As long as the earth turns, we'll need power at night too (and on cloudy days), and current energy storage devices are not good enough to cover all of that usage.

ben_w 4y ago

A true fact about what we built yesterday isn't a necessary limit to what we can build tomorrow. The storage tech works (as would a global power grid), we "only" need to scale it up, and at current prices a global grid would cost about the same as current annual spending on fossil fuels, while the cost of batteries depends on your estimate for storage requirements, which depends on the diversity (location and energy sources) of the grid, but could be in a similar range. (Batteries are also the most expensive of the seriously considered solutions, other storage methods exist and work just fine but aren't so good for cars and we need the batteries for cars anyway). I'm not sure how long either approach would last before needing servicing/replacement, but I'd be surprised if it was less than a year. Anyone here know how long underwater HVDC cables last?

snovv_crash 4y ago
That's assuming that prices stay the same for the materials needed to build the grid. Have you looked into the amount of copper required to build this grid? It is many multiples of the current annual production, this would cause prices to rise enormously.
- ben_w 4y ago
  > Have you looked into the amount of copper required to build this grid? It is many multiples of the current annual production Yes, I have, and it is, even including aluminium (surprisingly a better choice for various reasons). I think I'm one of the main proponents of this on Hacker News, so you might have even seen my previous comments calculating this. WRT economics, not my field; all I can say is that higher demand increasing prices will encourage (but not necessarily ultimately create) an increase in supply, which has a difficult to forecast feedback relationship when planning for a specific total output.
  - snovv_crash 4y ago
    At least the initial rampup would be expensive - the cheapest copper is being mined already, so prices would necessarily need to rise to mine the more expensive copper too. How much it would rise long term once the new mines are running is geologically dependant. Aluminium has different issues, in that it is very very very energy intensive to refine. It may require new coal/nuclear power generation buildout near the mine for smelting the bauxite to make it even feasible at a large scale, otherwise you run into a chicken/egg problem with your grid idea since the grid doesn't exist yet. This is because aluminium smelters can't be shut down at night, they have a single turn on/off in their lifetime otherwise they need a full disassembly and rebuild, more likely scrap. So it can't just be run intermittently on local solar. To some degree it can be ramped down, so maybe a smaller base generation assisted by solar during the day, but it still needs significant power at night somehow.
    - ncmncm 4y ago
      Aluminum ore is super-abundant. There is no possibility of any country anywhere on Earth running short. The only per-ton expense to produce it is for electricity, which price will be falling very, very fast. So, no, we will not run short of aluminum, even at many times current production rate. Aluminum production process from ore + electric -> metal is extremely mature. Factories to make the factories have already been running for most of a century. They will operate at night on wind and, during lulls, drawing down storage. They could keep hot by burning powdered aluminum. New ones will be built preferably where wind is reliable, or (as now) hydro power is cheap. Or they might be built near a cliff, and operate their own pumped hydro. So, no, intermittent sources are not a problem for them or really for anybody, because storage is easy and cheap.
      - snovv_crash 4y ago
        Absolutely, no shortage of aluminium, as I said, copper is the one where the mining is the bottleneck, for aluminium it is energy for the smelting. Night time energy storage for the smelters is not a solved issue still and significantly increases the costs beyond just adding sufficient PV panels adjacent to the smelter. Storage is not "easy and cheap" as you assert, otherwise there would be no reason to build the longitudinal grid connections GP was talking about.
ajsnigrutin 4y ago
But why not just use nuclear, which produces power 24/7 with no need for storage?
- ben_w 4y ago
  • Cost (as much as the most expensive serious storage systems) • Domestic politics (NIMBYing for both reactors and waste, and you don't get around that just by telling people it's fine) • International politics (extra costs from proving nobody is using them for proliferation) • Build time (too slow compared to renewables) • Depending on your particular perspective, you may also care that the fuel itself is in limited supply (given the assumption that at some point even the currently-poor will want energy-hogging lifestyles and that nuclear reserves, even with breeders are 2500 ZJ compared to 65 ZJ of fossil fuels, then if you give everyone a Qatar per capita use that runs out in only about 300 years).

ZeroGravitas 4y ago

I'm glad you say "cover all" because I think people underestimate by quite a lot how far PV and Lithium batteries can get you if you're not near the poles. So today's tech, just with more deployment mostly solves this. Yes there's room for Wind and heat batteries and all those other things, but even the brutally stupid approach of Solar PV and lithium batteries that we have today, without any new breakthroughs, is good enough to cover most of it. People use more energy during the day, when the sun is up. Solar PV is the cheapest energy in history and broadly matches this daily profile.