3 ms·

The cost of generation is not the same thing as the net cost of transitioning an energy grid to a different energy source. Wind and solar have cheap generation,

The cost of generation is not the same thing as the net cost of transitioning an energy grid to a different energy source. Wind and solar have cheap generation, but have much greater infrastructure costs than nuclear. Storage is a huge one, we don't have effective means of storage besides hydroelectric dams which are geographically limited. Lithium ion batteries still take over $500/KWh to install (the <$200 figures are for the battery cells themselves, omitting the cost of installation, transformers, maintenance, etc.). And they're set to increase as raw materials are strained [1][2]. This is why plans to transition to a majority renewable grid typically assume that hydrogen, or some other form of storage will provide storage at a fraction of the cost of existing storage methods. Nobody has actually built commercial hydrogen storage, though, so this is a big assumption. There's also the cost of transmission. Energy dense sources like nuclear power can be placed close to electricity demand. But low-density sources by definition need to be spread out and distributed. Decentralized generation is not a good thing, as it requires more transmission infrastructure to support. It's not uncommon for renewable projects to be denied because the infrastructure can't handle the transmission requirements [3]. Nuclear avoids these issues. It's a non-intermittent source with the greatest capacity factor of any generation system. Downtime is usually scheduled. This eliminates the need for storage. It's also energy dense. It can be used in place of existing fossil fuel heat engines, avoided in the need to make large build-outs of transmission infrastructure. 1. https://www.aljazeera.com/economy/2022/7/7/lithiums-insane-costs-threaten-ev-fix-for-climate-change https://www.aljazeera.com/economy/2022/7/7/lithiums-insane-c... 2. https://tradingeconomics.com/commodity/lithium https://tradingeconomics.com/commodity/lithium 3. https://www.vox.com/videos/22685707/climate-change-clean-energy-high-voltage-transmission-lines https://www.vox.com/videos/22685707/climate-change-clean-ene...

dmitriid 4y ago
> but have much greater infrastructure costs than nuclear. And proponents of solar forget that they have to cover large swaths of land with solar to begin with. Some places don't have the luxury of converting land to solar farms
- spangry 4y ago
  The transmission and land costs also trade off against each other to an extent: land close to demand is more expensive, cheaper land far away from demand requires more transmission infrastructure.
- ben_w 4y ago
  Large swaths of land, but not absurdly large; the apartment towers two streets from me as I write, could generate about half their needs from PV cladding, and I'm in Berlin, slightly further north than the US-Canada border. Likewise, PV cladding on electric cars, even though this is absolutely not the optimal place for PV, can generate 50-80% of the mean demand for that vehicle in most places. But PV scales up and down, so you can also use it as a roofing material in a car park (one of the reasons cladding cars themselves in PV isn't the best idea), or on otherwise worthless land like the gap between the eastbound and westbound I-80 [0], or on top of reservoirs and rivers where we want to reduce evaporation [1]. Given where the sunlight mostly falls and where people mostly live, the biggest discrepancies are the UK (mostly as bad as the coastal mountains on the strip of Canada between Washington state and Alaska), and China (the people mostly live in the east which is relatively cloudy, the desert where they put some (but not all) of their PV is in the west). For the small settlements that can't use PV and which it would be uneconomical to connect to a grid, there's wind and geothermal, and the rest may be so few as to not actually matter for environmental purposes if they do keep using diesel or methane. But probably not nuclear, because you can get a lot of transmission cable for the price of a reactor. [0] I'm not sure how big that gap is, but a few random samples suggest 15 meters width, which means 20% efficient PV there has a nameplate capacity of 14 GW, and even assuming just 10% duty factor that's still more than a nuclear power plant, but the capacity factor is mostly going to be 18% in the area the I-80 goes through. The entire US interstate network, assuming 15 m direction separation and 20% efficiency, has room for 235 GW nameplate, adjust accordingly for capacity factor. There's much more length of railway line in the US, though I have no idea what the average width is over its length, and I don't understand the engineering constraints well enough to know how close PV could be placed to the rails (between the sleepers?) [1] Lake Mead by itself would be enough for 128 GW nameplate, looking at the solar potential in that area it probably has a capacity factor of 22%.
logifail 4y ago
> Nuclear avoids these issues. It's a non-intermittent source with the greatest capacity factor of any generation system. A nuclear plant is pumping out electricity 24x7, but the problem is, no electricity consumer needs that kind of supply, certanly not at the prices nuclear needs to charge. When the wind is blowing and/or the sun is shining, every single electricity consumer would prefer to pick a cheaper, renewable, option. The only way nuclear can even pretend to make the math work out is by going on and on about "baseload" and hoping that they can lock in buyers over several decades rather than having to actually look the spot price in the face.
- Manuel_D 4y ago
  Minimum energy demand is usually 80% of peak energy demand. Furthermore, the peak energy demand happens in the evening when solar tapers off. The time of day when energy is most scare is exactly when solar stops producing. Wind may or may not produce at the right time, as wind speeds through the day vary depending on location. There's no "pretending" about base load. The vast majority of electricity demand is base load, and the peaks beyond base load happen when solar stops working.
  - fulafel 4y ago
    It's more informative for judging value to examine peak vs min price. 80% production off peak just says there's some demand willing to exceed marginal production cost, which can be quite low.

dmitriid 4y ago

> but have much greater infrastructure costs than nuclear. And proponents of solar forget that they have to cover large swaths of land with solar to begin with. Some places don't have the luxury of converting land to solar farms

spangry 4y ago
The transmission and land costs also trade off against each other to an extent: land close to demand is more expensive, cheaper land far away from demand requires more transmission infrastructure.
ben_w 4y ago
Large swaths of land, but not absurdly large; the apartment towers two streets from me as I write, could generate about half their needs from PV cladding, and I'm in Berlin, slightly further north than the US-Canada border. Likewise, PV cladding on electric cars, even though this is absolutely not the optimal place for PV, can generate 50-80% of the mean demand for that vehicle in most places. But PV scales up and down, so you can also use it as a roofing material in a car park (one of the reasons cladding cars themselves in PV isn't the best idea), or on otherwise worthless land like the gap between the eastbound and westbound I-80 [0], or on top of reservoirs and rivers where we want to reduce evaporation [1]. Given where the sunlight mostly falls and where people mostly live, the biggest discrepancies are the UK (mostly as bad as the coastal mountains on the strip of Canada between Washington state and Alaska), and China (the people mostly live in the east which is relatively cloudy, the desert where they put some (but not all) of their PV is in the west). For the small settlements that can't use PV and which it would be uneconomical to connect to a grid, there's wind and geothermal, and the rest may be so few as to not actually matter for environmental purposes if they do keep using diesel or methane. But probably not nuclear, because you can get a lot of transmission cable for the price of a reactor. [0] I'm not sure how big that gap is, but a few random samples suggest 15 meters width, which means 20% efficient PV there has a nameplate capacity of 14 GW, and even assuming just 10% duty factor that's still more than a nuclear power plant, but the capacity factor is mostly going to be 18% in the area the I-80 goes through. The entire US interstate network, assuming 15 m direction separation and 20% efficiency, has room for 235 GW nameplate, adjust accordingly for capacity factor. There's much more length of railway line in the US, though I have no idea what the average width is over its length, and I don't understand the engineering constraints well enough to know how close PV could be placed to the rails (between the sleepers?) [1] Lake Mead by itself would be enough for 128 GW nameplate, looking at the solar potential in that area it probably has a capacity factor of 22%.

logifail 4y ago

> Nuclear avoids these issues. It's a non-intermittent source with the greatest capacity factor of any generation system. A nuclear plant is pumping out electricity 24x7, but the problem is, no electricity consumer needs that kind of supply, certanly not at the prices nuclear needs to charge. When the wind is blowing and/or the sun is shining, every single electricity consumer would prefer to pick a cheaper, renewable, option. The only way nuclear can even pretend to make the math work out is by going on and on about "baseload" and hoping that they can lock in buyers over several decades rather than having to actually look the spot price in the face.

Manuel_D 4y ago
Minimum energy demand is usually 80% of peak energy demand. Furthermore, the peak energy demand happens in the evening when solar tapers off. The time of day when energy is most scare is exactly when solar stops producing. Wind may or may not produce at the right time, as wind speeds through the day vary depending on location. There's no "pretending" about base load. The vast majority of electricity demand is base load, and the peaks beyond base load happen when solar stops working.
- fulafel 4y ago
  It's more informative for judging value to examine peak vs min price. 80% production off peak just says there's some demand willing to exceed marginal production cost, which can be quite low.