4 ms·

Even if you account for intermittence? As far as I know solar/wind alone are not a replacement e.g. I'm not aware of steel factories running on batteries given

Even if you account for intermittence? As far as I know solar/wind alone are not a replacement e.g. I'm not aware of steel factories running on batteries given the order of magnitude of energy consumption.

ben_w 3y ago
On paper a global 1 Ω power grid is only $265 billion of aluminium / 1.7 years of current global production. I don't actually expect the whole world to do this because politics, but I won't be surprised if China makes a proof of concept in the form of a 100 Ω connection to both Africa and South America. As for iron/steel specifically, I think people have only recently demonstrated that this can be electrified at scale, even without batteries. Battery capacity is also in the early stages yet; what we need for fully electric transportation — which is probably around a decade of exponential growth away — is more than enough for balancing current (no pun intended) electrical demand, but I don't know if it's also enough once we electrify steel. I think iron might be one of the cases where it makes sense to electrolyse hydrogen from water and use that as a direct chemical agent, but my chemistry knowledge is a mediocre GCSE from 24 years ago plus YouTube videos so I don't trust my own judgement on this regardless of what Wittgenstein says about 'to believe falsely' in the first person, present indicative.
- skissane 3y ago
  > On paper a global 1 Ω power grid is only $265 billion of aluminium / 1.7 years of current global production. Sorry, what does ohms mean in this context? I would get what gigawatts or gigawatt-hours would mean in this context, but don’t understand the relevance of resistance
  - ben_w 3y ago
    Ohms tells you how much power you lose to resistance given the voltage drop from one end to the other. The voltage you choose to put across it is a free parameter in the range +/- 1e6 V or so. As most people would consider "mildly heating the Pacific" to be pointless, you add the resistance of the line to the effective resistance of all the load on the grid and then use P=IV=I^2R=V^2/R to work out how many volts/amps/watts are going though the line. You actually have to have a number for grid load to turn this into "what percentage are we losing to the transmission line". (Also: I'm simplifying by assuming DC, because AC resistance depends on the environment and the frequency and that length is long enough to just be an antenna at normal grid frequencies).
    - skissane 3y ago
      Right, I understand now, you are talking about transmission losses. And using units of resistance is a way of talking about the loss independently of the amount of power transmitted. Although, I think your original statement I quoted, is omitting some information you are relying on – e.g. how much energy is "$265 billion of aluminium", and so how much energy are we talking about such that 1 ohm transmission loss would amount to that
      - ben_w 3y ago
        > how much energy is "$265 billion of aluminium", and so how much energy are we talking about such that 1 ohm transmission loss would amount to that I don't understand the question. $265 bn of aluminium is a square metre cross section going around the equator and back to the start, or any equivalent network. How much energy can be conveyed through that depends how long it lasts, which depends on the specifics, and how much power on average it is used to convey over that lifetime. Many grids are showing their age at 70-100 years. Perhaps it would last that long. Perhaps we'd choose to put the entire world electricity supply through it continuously so that Anchorage nighttime in winter is powered by the simultaneous New Zealand daytime summer. Or perhaps we won't. But these kinds of question are the domain of geopolitics and global economics, so the best I can do is give the numbers that show a very small electrical resistance is objectively quite cheap — significantly less than the US alone is planning on spending to modernise its grids.
        skissane 3y ago
        I misunderstood what you were originally saying. I thought you were talking about how much electricity was consumed in smelting that much aluminium. Not actually using that much aluminium to transmit power
patrickk 3y ago
I’m not sure if there’s a been a study done, but it’s interesting to see some potential solutions to the intermittency problem and how close they’d get the world to a completely green grid. Orientating solar panels differently than simply south facing to better match usage patterns: https://youtu.be/LqizLQDi9BM?si=4uqsRO3qEGfm4Yxr https://youtu.be/LqizLQDi9BM?si=4uqsRO3qEGfm4Yxr Building out a “supergrid”, like what’s already happening across Europe. Balance intermittent supply across a wide geographical area. There’s always wind blowing somewhere, Iceland has massive, 24/7 geothermal that’s not yet connected to mainland Europe, Southern Europe has huge solar potential, and so on. The UK has a project to get 8% of its electric demand from a massive solar/wind/battery project built in Morocco linked via undersea cable to the UK. Finally, batteries. Lifepo4 batteries are constantly dropping in price. We haven’t seen a huge deployment of Chinese sodium ion batteries for homes yet, but it’s coming. And that’s not touching on some of the other possible solutions.
AtlasBarfed 3y ago
LCOE from Lazards is starting to wrap in grid storage costs. The numbers show that armageddon, if it hasn't arrived for other power generation, is imminent: bundled unsubsidized solar/wind + storage is falling under natural gas turbine. That is without sodium ion batteries coming into mass production that are 40% of the cost of Lithium ion and 20-30% less than even LFP chemistries. Without mass scale perovskites. Sodium Ion and LFP are roadmapped for continued 20-30% density gains in the coming years. Coming down the pipe after that? Sulfur techs that might improve capacities 100-300% with no cobalt/nickel issues, solid state configurations, etc. Yes, who knows how many cloudy days or windless skies the storage holds compared to leveling, but what is apparent is that the intermittence concern is going to simply not be a major factor. Yes it requires a LOT of grid development (and IMO we should be heavily encouraging, financing, and subsidizing home solar generation and storage by a factor of ten than we currently do). This isn't a 10 year out trend. It's the here and now, and it's not the current invading horde (solar/wind being comprehensively cheaper than natural gas combined cycle with the issue of intermittency), there are waves of even stronger hordes coming to invade traditional power generation (solar/wind being cheaper and cheaper and cheaper, and storage being cheaper and cheaper and cheaper). Factories are currently not running on batteries and the like, but... costs are costs. If a factory needs a lot of power, will it pick a natural gas turbine that is 2x more expensive right now and 4-5x as expensive in ten years, or nuclear that is currently 6x as expensive (and no SMR availiability anyway) and 12-20x as expensive in ten years? Or will the factory adapt production to dirt dirt dirt cheap power production with some unreliability? Alternative energy is a HUGE economic battering ram. If a factory is paying 4x the cost for its energy, it's not going to compete. If a data center is paying 4x the cost for electricity, it will be shut down and moved to a place with more reliable alternative energy production (desert, some mountain valley with sustained winds, etc).

ben_w 3y ago

On paper a global 1 Ω power grid is only $265 billion of aluminium / 1.7 years of current global production. I don't actually expect the whole world to do this because politics, but I won't be surprised if China makes a proof of concept in the form of a 100 Ω connection to both Africa and South America. As for iron/steel specifically, I think people have only recently demonstrated that this can be electrified at scale, even without batteries. Battery capacity is also in the early stages yet; what we need for fully electric transportation — which is probably around a decade of exponential growth away — is more than enough for balancing current (no pun intended) electrical demand, but I don't know if it's also enough once we electrify steel. I think iron might be one of the cases where it makes sense to electrolyse hydrogen from water and use that as a direct chemical agent, but my chemistry knowledge is a mediocre GCSE from 24 years ago plus YouTube videos so I don't trust my own judgement on this regardless of what Wittgenstein says about 'to believe falsely' in the first person, present indicative.

skissane 3y ago
> On paper a global 1 Ω power grid is only $265 billion of aluminium / 1.7 years of current global production. Sorry, what does ohms mean in this context? I would get what gigawatts or gigawatt-hours would mean in this context, but don’t understand the relevance of resistance
- ben_w 3y ago
  Ohms tells you how much power you lose to resistance given the voltage drop from one end to the other. The voltage you choose to put across it is a free parameter in the range +/- 1e6 V or so. As most people would consider "mildly heating the Pacific" to be pointless, you add the resistance of the line to the effective resistance of all the load on the grid and then use P=IV=I^2R=V^2/R to work out how many volts/amps/watts are going though the line. You actually have to have a number for grid load to turn this into "what percentage are we losing to the transmission line". (Also: I'm simplifying by assuming DC, because AC resistance depends on the environment and the frequency and that length is long enough to just be an antenna at normal grid frequencies).
  - skissane 3y ago
    Right, I understand now, you are talking about transmission losses. And using units of resistance is a way of talking about the loss independently of the amount of power transmitted. Although, I think your original statement I quoted, is omitting some information you are relying on – e.g. how much energy is "$265 billion of aluminium", and so how much energy are we talking about such that 1 ohm transmission loss would amount to that
    - ben_w 3y ago
      > how much energy is "$265 billion of aluminium", and so how much energy are we talking about such that 1 ohm transmission loss would amount to that I don't understand the question. $265 bn of aluminium is a square metre cross section going around the equator and back to the start, or any equivalent network. How much energy can be conveyed through that depends how long it lasts, which depends on the specifics, and how much power on average it is used to convey over that lifetime. Many grids are showing their age at 70-100 years. Perhaps it would last that long. Perhaps we'd choose to put the entire world electricity supply through it continuously so that Anchorage nighttime in winter is powered by the simultaneous New Zealand daytime summer. Or perhaps we won't. But these kinds of question are the domain of geopolitics and global economics, so the best I can do is give the numbers that show a very small electrical resistance is objectively quite cheap — significantly less than the US alone is planning on spending to modernise its grids.
      - skissane 3y ago
        I misunderstood what you were originally saying. I thought you were talking about how much electricity was consumed in smelting that much aluminium. Not actually using that much aluminium to transmit power

patrickk 3y ago

I’m not sure if there’s a been a study done, but it’s interesting to see some potential solutions to the intermittency problem and how close they’d get the world to a completely green grid. Orientating solar panels differently than simply south facing to better match usage patterns: https://youtu.be/LqizLQDi9BM?si=4uqsRO3qEGfm4Yxr https://youtu.be/LqizLQDi9BM?si=4uqsRO3qEGfm4Yxr Building out a “supergrid”, like what’s already happening across Europe. Balance intermittent supply across a wide geographical area. There’s always wind blowing somewhere, Iceland has massive, 24/7 geothermal that’s not yet connected to mainland Europe, Southern Europe has huge solar potential, and so on. The UK has a project to get 8% of its electric demand from a massive solar/wind/battery project built in Morocco linked via undersea cable to the UK. Finally, batteries. Lifepo4 batteries are constantly dropping in price. We haven’t seen a huge deployment of Chinese sodium ion batteries for homes yet, but it’s coming. And that’s not touching on some of the other possible solutions.

AtlasBarfed 3y ago

LCOE from Lazards is starting to wrap in grid storage costs. The numbers show that armageddon, if it hasn't arrived for other power generation, is imminent: bundled unsubsidized solar/wind + storage is falling under natural gas turbine. That is without sodium ion batteries coming into mass production that are 40% of the cost of Lithium ion and 20-30% less than even LFP chemistries. Without mass scale perovskites. Sodium Ion and LFP are roadmapped for continued 20-30% density gains in the coming years. Coming down the pipe after that? Sulfur techs that might improve capacities 100-300% with no cobalt/nickel issues, solid state configurations, etc. Yes, who knows how many cloudy days or windless skies the storage holds compared to leveling, but what is apparent is that the intermittence concern is going to simply not be a major factor. Yes it requires a LOT of grid development (and IMO we should be heavily encouraging, financing, and subsidizing home solar generation and storage by a factor of ten than we currently do). This isn't a 10 year out trend. It's the here and now, and it's not the current invading horde (solar/wind being comprehensively cheaper than natural gas combined cycle with the issue of intermittency), there are waves of even stronger hordes coming to invade traditional power generation (solar/wind being cheaper and cheaper and cheaper, and storage being cheaper and cheaper and cheaper). Factories are currently not running on batteries and the like, but... costs are costs. If a factory needs a lot of power, will it pick a natural gas turbine that is 2x more expensive right now and 4-5x as expensive in ten years, or nuclear that is currently 6x as expensive (and no SMR availiability anyway) and 12-20x as expensive in ten years? Or will the factory adapt production to dirt dirt dirt cheap power production with some unreliability? Alternative energy is a HUGE economic battering ram. If a factory is paying 4x the cost for its energy, it's not going to compete. If a data center is paying 4x the cost for electricity, it will be shut down and moved to a place with more reliable alternative energy production (desert, some mountain valley with sustained winds, etc).