8 ms·

Reducing carbon emissions means electrifying a lot of things that were not electric before. We are going to need a lot more base generation than we have now.

by mlyle 1y ago

Reducing carbon emissions means electrifying a lot of things that were not electric before. We are going to need a lot more base generation than we have now. Large grids, overbuilding renewables, diversity of renewables, short and medium term storage, and load shedding/dynamic pricing are all good starts but IMO won’t be enough— we should scale up nuclear too.

tialaramex 1y ago
More, but not as much more as people often naively expect because it turns out converting liquid fuel into motion by burning/ exploding the fuel isn't very efficient on a small scale whereas electric motors are very efficient, so 1TW year of "People driving to work" in ICE cars does not translate into needing 1TW year extra electricity generation if they have electric cars instead, let alone 1TW year of extra network capacity to deliver it. Where we're replacing fossil fuel heat with a heat pump we don't get that efficiency improvement from motors - burning fuel was 100% efficient per se, but the heat pump is > 100% efficient in those terms because it's not making heat just moving it. Nuclear is much less popular than almost any generation technology, so you're fighting a significant political battle to make that happen.
- mlyle 1y ago
  We need a lot more. Right now only about 25 to 33 pc of our energy consumption is electric. Some of the rest will get significant efficiency benefit like you mention — cars, building heating, etc. Others, much less so— high temperature industrial heat, long distance transport, etc. Reaching current nighttime use with storage and wind and existing hydro looks infeasible, and we need a minimum of twice as much. Power to gas (and back to power or to mix with natural gas for existing uses) is probably a part of this, but nuclear improves this (allowing there to be less of it and allowing the electrolysis cells to be used for a greater fraction of the day.
  - thfuran 1y ago
    Building heating won’t see much increase in efficiency from going electric compared to a high efficiency gas furnace.
    - dalyons 1y ago
      Hrm? Heat pumps are multiples more efficient than gas.
    - ZeroGravitas 1y ago
      A 4x efficiency bump is fairly easy (.9 efficient gas to 3.6 COP heat pump). Older non-condensing boilers (or modern condensing boilers run too hot) are more like .8 when new and messured at .6 in real life circumstances. Even at the high end efficiency, this is enough that you can burn the gas centrally in a generator, lose 40-60% of it as heat as is standard with fossil electricity generation, lose a few percent more electricity in transmission and still come out ahead overall. And of course, that's a lower bound as you'd ideally be generating electricity from other sources like solar and wind and battery and keeping the gas generators for when needed, making use of the giant scale gas storage most countries already have.
  - bryanlarsen 1y ago
    People have run the numbers. We need about 30% more. Which is a lot, but it's spread over 20-30 years, so it's not a lot each year.
    - AngryData 1y ago
      Does that also account for industrial chemical processes that don't have a simple power-energy exchange? Stuff like making fertilizer or solvents and the like do take a lot of electrical power currently, but will require even more rarely accounted for energy to create base reagents without fossil fuels. Like fertilizer already uses 1% of global electricity today, but if we want to create nitrogen fertilizers without fossil fuel sources, it takes up to a 10 times increase in energy requirements to synthesize from the air making it rise to near 10% of current electrical generation. Many oils are used in mechanical components are irreplaceable and have to be sourced, but to do it without fossil fuels and synthesize from organic materials also require a lot more energy than we use to purify or synthesize from fossil fuels. And the same is true of many solvents. Its usage is technically accounted for in fossil fuel extraction numbers, but generally ignored when people are accounting for total electrical generation and the usage of fuels as heat sources.
      - ben_w 1y ago
        Relevant question: fossil fuel dependency has two parts, the "peak oil" part, and the "global warming" part. As we don't have to solve these at the same time, are the things you raise more of a "peak oil" problem or a "global warming" problem?
        mlyle 1y ago
        They are related in that we use different petroleum products for different purposes and today extracting petroleum means burning most of it.
    - mlyle 1y ago
      US electric demand is 4 trillion kWh per year. Moving to EVs alone will be about 1 trillion kWh more. And that is leaving out transport, building heat, and industrial use. I suspect you are quoting an EV-only number. Alternatively, you might be looking at how much electricity demand is expected to increase if we maintain our current trajectory and don’t aggressively decarbonize.
pydry 1y ago
>we should scale up nuclear too. With a 5x higher LCOE and lead times of 15-20 years instead of 1-2 for solar/wind deployments, allocating money to scale up nuclear as well will just make the transition happen slower and at higher cost.
- mlyle 1y ago
  I don’t think we can scale up storage enough at any reasonable cost.
  - adrianN 1y ago
    We can start worrying about storage once we reach 60-80% renewable and just keep using fossil fuels as backup. Nuclear doesn't replace storage (at least not if you don't want to run your nuclear plants at like half capacity)
    - mlyle 1y ago
      In my market, we're already at fractions of renewable where prices go negative > 20% of daylight hours (doubled in last year), and still produce lots of CO2 in late afternoon. I think the time to start worrying seriously about storage is now or in the past.
  - bryanlarsen 1y ago
    We need about 30TWh of batteries to decarbonize the world's grid. China has 1TWh per year of capacity, increasing 50% per year. Cost is currently $35/kWh, dropping 20% per year.
    - mlyle 1y ago
      You’re again not considering electrification of current loads that burn fossil fuels. Unfortunately, a lot of these loads are closer to 24/7 and will require more storage. The IEA net zero scenario assumes 100TWh of storage and may not be enough. Total installed system costs— not batteries alone— are estimated at $300B/TWh. So that is on the order of $30T at current prices (some estimates reach to $100T). And of course, these investments don’t last forever— we can’t be kicking 3pc of GDP to storage. I expect this to improve, but having some clean, always-on generation greatly reduces the amount of storage and overprovisioned production of other types needed.
ben_w 1y ago
One of the bigger other sources of emissions is transport; transport requires some of the electricity is condensed into a portable form regardless of the specifics — batteries, hydrogen, chunks of purified metal to burn, whatever — and that condensation means it doesn't get any extra novel benefit from expensive-but-consistent nuclear over cheap-but-predictably-intermittent renewables. The scale is such that if we imagine a future with fully electrified cars, the batteries in those cars are more than enough to load-balance the current uses of the grid, and still are enough for the current uses of the grid when those batteries have been removed from the vehicles due to capacity wear making them no longer useful in a vehicle. The best time for more nuclear power was the 90s, the second best was 10 years ago; unless you have a cunning plan you've already shown to an investor about how to roll out reactors much much faster, I wouldn't hold your breath on them.
- mlyle 1y ago
  > and that condensation means it doesn't get any extra novel benefit from expensive-but-consistent nuclear over cheap-but-predictably-intermittent renewables This assumes you can do just the condensation during the day— E.g. you are amortizing the electrolyzers capital cost over just times when there is surplus power instead of something closer to 24/7.
  - nasmorn 1y ago
    Batteries to have solar 24/7 will be dirt cheap soon. The question is: is it cheaper to build nuclear and run the electrolyzers at the same rate the whole year or use solar and run them much less during the winter. Where much less is a function of latitude

by mlyle 1y ago

tialaramex 1y ago

More, but not as much more as people often naively expect because it turns out converting liquid fuel into motion by burning/ exploding the fuel isn't very efficient on a small scale whereas electric motors are very efficient, so 1TW year of "People driving to work" in ICE cars does not translate into needing 1TW year extra electricity generation if they have electric cars instead, let alone 1TW year of extra network capacity to deliver it. Where we're replacing fossil fuel heat with a heat pump we don't get that efficiency improvement from motors - burning fuel was 100% efficient per se, but the heat pump is > 100% efficient in those terms because it's not making heat just moving it. Nuclear is much less popular than almost any generation technology, so you're fighting a significant political battle to make that happen.

mlyle 1y ago
We need a lot more. Right now only about 25 to 33 pc of our energy consumption is electric. Some of the rest will get significant efficiency benefit like you mention — cars, building heating, etc. Others, much less so— high temperature industrial heat, long distance transport, etc. Reaching current nighttime use with storage and wind and existing hydro looks infeasible, and we need a minimum of twice as much. Power to gas (and back to power or to mix with natural gas for existing uses) is probably a part of this, but nuclear improves this (allowing there to be less of it and allowing the electrolysis cells to be used for a greater fraction of the day.
- thfuran 1y ago
  Building heating won’t see much increase in efficiency from going electric compared to a high efficiency gas furnace.
  - dalyons 1y ago
    Hrm? Heat pumps are multiples more efficient than gas.
  - ZeroGravitas 1y ago
    A 4x efficiency bump is fairly easy (.9 efficient gas to 3.6 COP heat pump). Older non-condensing boilers (or modern condensing boilers run too hot) are more like .8 when new and messured at .6 in real life circumstances. Even at the high end efficiency, this is enough that you can burn the gas centrally in a generator, lose 40-60% of it as heat as is standard with fossil electricity generation, lose a few percent more electricity in transmission and still come out ahead overall. And of course, that's a lower bound as you'd ideally be generating electricity from other sources like solar and wind and battery and keeping the gas generators for when needed, making use of the giant scale gas storage most countries already have.
- bryanlarsen 1y ago
  People have run the numbers. We need about 30% more. Which is a lot, but it's spread over 20-30 years, so it's not a lot each year.
  - AngryData 1y ago
    Does that also account for industrial chemical processes that don't have a simple power-energy exchange? Stuff like making fertilizer or solvents and the like do take a lot of electrical power currently, but will require even more rarely accounted for energy to create base reagents without fossil fuels. Like fertilizer already uses 1% of global electricity today, but if we want to create nitrogen fertilizers without fossil fuel sources, it takes up to a 10 times increase in energy requirements to synthesize from the air making it rise to near 10% of current electrical generation. Many oils are used in mechanical components are irreplaceable and have to be sourced, but to do it without fossil fuels and synthesize from organic materials also require a lot more energy than we use to purify or synthesize from fossil fuels. And the same is true of many solvents. Its usage is technically accounted for in fossil fuel extraction numbers, but generally ignored when people are accounting for total electrical generation and the usage of fuels as heat sources.
    - ben_w 1y ago
      Relevant question: fossil fuel dependency has two parts, the "peak oil" part, and the "global warming" part. As we don't have to solve these at the same time, are the things you raise more of a "peak oil" problem or a "global warming" problem?
      - mlyle 1y ago
        They are related in that we use different petroleum products for different purposes and today extracting petroleum means burning most of it.
  - mlyle 1y ago
    US electric demand is 4 trillion kWh per year. Moving to EVs alone will be about 1 trillion kWh more. And that is leaving out transport, building heat, and industrial use. I suspect you are quoting an EV-only number. Alternatively, you might be looking at how much electricity demand is expected to increase if we maintain our current trajectory and don’t aggressively decarbonize.

pydry 1y ago

>we should scale up nuclear too. With a 5x higher LCOE and lead times of 15-20 years instead of 1-2 for solar/wind deployments, allocating money to scale up nuclear as well will just make the transition happen slower and at higher cost.

mlyle 1y ago
I don’t think we can scale up storage enough at any reasonable cost.
- adrianN 1y ago
  We can start worrying about storage once we reach 60-80% renewable and just keep using fossil fuels as backup. Nuclear doesn't replace storage (at least not if you don't want to run your nuclear plants at like half capacity)
  - mlyle 1y ago
    In my market, we're already at fractions of renewable where prices go negative > 20% of daylight hours (doubled in last year), and still produce lots of CO2 in late afternoon. I think the time to start worrying seriously about storage is now or in the past.
- bryanlarsen 1y ago
  We need about 30TWh of batteries to decarbonize the world's grid. China has 1TWh per year of capacity, increasing 50% per year. Cost is currently $35/kWh, dropping 20% per year.
  - mlyle 1y ago
    You’re again not considering electrification of current loads that burn fossil fuels. Unfortunately, a lot of these loads are closer to 24/7 and will require more storage. The IEA net zero scenario assumes 100TWh of storage and may not be enough. Total installed system costs— not batteries alone— are estimated at $300B/TWh. So that is on the order of $30T at current prices (some estimates reach to $100T). And of course, these investments don’t last forever— we can’t be kicking 3pc of GDP to storage. I expect this to improve, but having some clean, always-on generation greatly reduces the amount of storage and overprovisioned production of other types needed.

ben_w 1y ago

One of the bigger other sources of emissions is transport; transport requires some of the electricity is condensed into a portable form regardless of the specifics — batteries, hydrogen, chunks of purified metal to burn, whatever — and that condensation means it doesn't get any extra novel benefit from expensive-but-consistent nuclear over cheap-but-predictably-intermittent renewables. The scale is such that if we imagine a future with fully electrified cars, the batteries in those cars are more than enough to load-balance the current uses of the grid, and still are enough for the current uses of the grid when those batteries have been removed from the vehicles due to capacity wear making them no longer useful in a vehicle. The best time for more nuclear power was the 90s, the second best was 10 years ago; unless you have a cunning plan you've already shown to an investor about how to roll out reactors much much faster, I wouldn't hold your breath on them.

mlyle 1y ago
> and that condensation means it doesn't get any extra novel benefit from expensive-but-consistent nuclear over cheap-but-predictably-intermittent renewables This assumes you can do just the condensation during the day— E.g. you are amortizing the electrolyzers capital cost over just times when there is surplus power instead of something closer to 24/7.
- nasmorn 1y ago
  Batteries to have solar 24/7 will be dirt cheap soon. The question is: is it cheaper to build nuclear and run the electrolyzers at the same rate the whole year or use solar and run them much less during the winter. Where much less is a function of latitude