4 ms·

Bear in mind that generation is only part of the holistic system of energy delivery. Intermittent sources require storage, which is separated out into a separat

Bear in mind that generation is only part of the holistic system of energy delivery. Intermittent sources require storage, which is separated out into a separate chart. Even just 4 hours of storage would bring solar's $30/MW to $210 to $350/MW. And 4 hours of storage is pretty thin - most plans call for 12 hours of storage. And storage costs are only projected to increase: https://www.cnbc.com/2022/05/18/ev-battery-costs-set-to-spike-as-raw-material-shortages-drags-on.html https://www.cnbc.com/2022/05/18/ev-battery-costs-set-to-spik... By comparison nuclear costs $131-$204/MW, so it's still cheaper after storage. The Lazard estimate also didn't include the transmission expansions necessary to support the distributed nature of renewable generation (explained further here: https://www.vox.com/videos/22685707/climate-change-clean-energy-high-voltage-transmission-lines https://www.vox.com/videos/22685707/climate-change-clean-ene...)

Schroedingersat 4y ago
You seem to be mixing up some units, but it doesn't change the point. Assuming your numbers are up to date (which I'm sceptical of at 4h storage): Firstly nuclear also requires storage otherwise you're paying $400-600/MWh on account of the idle capital. Secondly you don't need to lead with storage, you can build out up to 50% or so renewable without. Thirdly if the costs of battery were to halve or quarter, such as by the conversion of lithium ion to sodium ion production which is already happening, then the reasoning evaporates entirely.
- Manuel_D 4y ago
  The units aren't being mixed up, they're in both MW and MWh. 100 MW / 400 MWh refers to the storage capacity required to put out 100 megawatts for four hours. Say you have a 100 MW solar facility, how much does it cost to keep the plant running 4 hours after sundown? That's what 100 MW / 400 MWh means. So sum this number with the cost of generation up on the chart above, and that's the net cost with storage. Nuclear doesn't need storage, since it's power output is not intermittent. The amount of wasted capacity varies with season, but minimum demand is usually 80% of peak demand. In the summer, this increases and the peak demand occurs during the day. This is a good place for solar: rooftop solar mitigating A/C is a great application of solar. But to fulfill base load, which is the vast majority of electricity demand, it is an ineffective choice.
epistasis 4y ago
> Even just 4 hours of storage would bring solar's $30/MW to $210 to $350/MW (Assuming you mean MWh here and it's just a typo.). Where are you getting those numbers Lazard's estimates that I linked have solar at $130-$230/MWh when charging directly from the grid, and $85-$160/MWh when charging from attached solar which shares the same inverters. (And for additional duration, just add more batteries and discharge them at less power) So if you blend 50% stored electricity with 50% delivered from primary renewable generation at a fraction of the cost, it's way below nuclear's costs. Finally, we don't even really know that those are nuclear's costs. We haven't finished Vogtle, so we are at scientific-wild-ass-guess stage of how much it will be when it finishes. Go back 10 years and the estimated cost of nuclear is almost reasonable, but every single year that we have had Vogtle delays and cost increases, the general estimates of nuclear's cost goes up. We don't know when it will stop going up, until we have a few more reactors built. And when you add in the cost of potential construction failure, like at South Carolina's Summer addition, started at the same time as Vogtle, nuclear goes up even more. People always overestimate the potential of nuclear, for some reason, and reality bites them. In contrast, people always underestimate the potential of storage and renewables, and reality delights them. It's time to reset our expectations based on the copious amounts of data we have collected over the past decade.
- Manuel_D 4y ago
  No, the units are being mixed up. 100 MW / 100 MWh under "Levelized cost of Storage" means "storage of a capacity of 100 MWh, that can be released at a rate of 100 MW". So if I have a 100 MW solar plant, and I want it to run for 4 hours after sundown, I need 100 MW / 400 MWh of storage capacity. So it's that storage cost, plus the cost of 100 MW solar generation that yield the net cost. Why not just measure in MWh? Because storage is useless without means to put that energy back in the grid. The electrical transformers to do this are not insignificant. So 100 MW / 100 MWh is more than a quarter of the cost of 100 MW / 400 MWh because of this overhead. Current nuclear builds have indeed been expensive, the plants last a long time. As per your own source, it's still less than intermittent sources when you include the cost of storage. Furthermore, Lazard is not examining the transmission infrastructure needed o support low-density sources like solar and wind. So, yeah, despite the expense nuclear is still the most cost-effective option. Furthermore, Lazard's estimates about nuclear are actually quite conservative. Nuclear plants were cheaper when built at scale. During the 1960s through the early 1980s, nuclear plants were often built for less than a quarter of what present builds [1]. Nuclear doesn't need technological improvement to get cheaper, it needs the economy of scale that previous plants enjoyed. https://www.sciencedirect.com/science/article/pii/S0301421516300106 https://www.sciencedirect.com/science/article/pii/S030142151...

Schroedingersat 4y ago

You seem to be mixing up some units, but it doesn't change the point. Assuming your numbers are up to date (which I'm sceptical of at 4h storage): Firstly nuclear also requires storage otherwise you're paying $400-600/MWh on account of the idle capital. Secondly you don't need to lead with storage, you can build out up to 50% or so renewable without. Thirdly if the costs of battery were to halve or quarter, such as by the conversion of lithium ion to sodium ion production which is already happening, then the reasoning evaporates entirely.

Manuel_D 4y ago
The units aren't being mixed up, they're in both MW and MWh. 100 MW / 400 MWh refers to the storage capacity required to put out 100 megawatts for four hours. Say you have a 100 MW solar facility, how much does it cost to keep the plant running 4 hours after sundown? That's what 100 MW / 400 MWh means. So sum this number with the cost of generation up on the chart above, and that's the net cost with storage. Nuclear doesn't need storage, since it's power output is not intermittent. The amount of wasted capacity varies with season, but minimum demand is usually 80% of peak demand. In the summer, this increases and the peak demand occurs during the day. This is a good place for solar: rooftop solar mitigating A/C is a great application of solar. But to fulfill base load, which is the vast majority of electricity demand, it is an ineffective choice.

epistasis 4y ago

> Even just 4 hours of storage would bring solar's $30/MW to $210 to $350/MW (Assuming you mean MWh here and it's just a typo.). Where are you getting those numbers Lazard's estimates that I linked have solar at $130-$230/MWh when charging directly from the grid, and $85-$160/MWh when charging from attached solar which shares the same inverters. (And for additional duration, just add more batteries and discharge them at less power) So if you blend 50% stored electricity with 50% delivered from primary renewable generation at a fraction of the cost, it's way below nuclear's costs. Finally, we don't even really know that those are nuclear's costs. We haven't finished Vogtle, so we are at scientific-wild-ass-guess stage of how much it will be when it finishes. Go back 10 years and the estimated cost of nuclear is almost reasonable, but every single year that we have had Vogtle delays and cost increases, the general estimates of nuclear's cost goes up. We don't know when it will stop going up, until we have a few more reactors built. And when you add in the cost of potential construction failure, like at South Carolina's Summer addition, started at the same time as Vogtle, nuclear goes up even more. People always overestimate the potential of nuclear, for some reason, and reality bites them. In contrast, people always underestimate the potential of storage and renewables, and reality delights them. It's time to reset our expectations based on the copious amounts of data we have collected over the past decade.

Manuel_D 4y ago
No, the units are being mixed up. 100 MW / 100 MWh under "Levelized cost of Storage" means "storage of a capacity of 100 MWh, that can be released at a rate of 100 MW". So if I have a 100 MW solar plant, and I want it to run for 4 hours after sundown, I need 100 MW / 400 MWh of storage capacity. So it's that storage cost, plus the cost of 100 MW solar generation that yield the net cost. Why not just measure in MWh? Because storage is useless without means to put that energy back in the grid. The electrical transformers to do this are not insignificant. So 100 MW / 100 MWh is more than a quarter of the cost of 100 MW / 400 MWh because of this overhead. Current nuclear builds have indeed been expensive, the plants last a long time. As per your own source, it's still less than intermittent sources when you include the cost of storage. Furthermore, Lazard is not examining the transmission infrastructure needed o support low-density sources like solar and wind. So, yeah, despite the expense nuclear is still the most cost-effective option. Furthermore, Lazard's estimates about nuclear are actually quite conservative. Nuclear plants were cheaper when built at scale. During the 1960s through the early 1980s, nuclear plants were often built for less than a quarter of what present builds [1]. Nuclear doesn't need technological improvement to get cheaper, it needs the economy of scale that previous plants enjoyed. https://www.sciencedirect.com/science/article/pii/S0301421516300106 https://www.sciencedirect.com/science/article/pii/S030142151...