4 ms·

> I’m continuously baffled by those that claim that nuclear power is safe. Nuclear power is relatively safe. Nothing is 100% safe. Rooftop solar actually has a

by Lazare 6y ago

> I’m continuously baffled by those that claim that nuclear power is safe. Nuclear power is relatively safe. Nothing is 100% safe. Rooftop solar actually has a surprisingly high fatality rate from installers falling off of roofs! Fossil fuels release horrible byproducts into the atmosphere, hydro sometimes causes floods...everything has drawbacks. > Analysis of nuclear power should take into account Black Swan events Quite right. Analysis of all power options should take into account black swan events (eg, a hydro dam failure taking out a city), but they should also do the reverse, and take into account the normal risks of operations. A coal power plant is subject to any particularly terrible catastrophic failure modes, but every day it operates it spews out a lot of carcinogens. Imagine a system that kills 5k people once every 20 years, versus a system that kills 1 person a day. You end up with scary headlines about the first one, but the second one kills significantly more people over time. > pro-nuclear arguments continuously give the tired argument that with new designs “this time is different.” The pro-nuclear arguments I have seen are much more conservative than that. They point out, correctly, that even counting Three Mile Island, Chernobyl, and Fukushima, nuclear energy is relatively safe. If the future ends up safer than the past that would be a bonus, but the past has been shockingly safe. Coal kills around 13k people every single year in the US alone. How many people have died due to nuclear power in the US in the entire history of nuclear energy?

bgorman 6y ago
How can we know that coal kills 13k people a year in the US? Sure, burning coal releases carcinogens but most of the population lives far away from coal plants.
- Lazare 6y ago
  This study says 13k: https://www.catf.us/wp-content/uploads/2010/09/CATF_Pub_TheTollFromCoal.pdf https://www.catf.us/wp-content/uploads/2010/09/CATF_Pub_TheT... Estimates vary; some reckon it's more like 50k: https://www.sciencedirect.com/science/article/abs/pii/S1352231013004548?via%3Dihub https://www.sciencedirect.com/science/article/abs/pii/S13522...
- corty 6y ago
  Similar to how we "know" how many people a low-level nuclear release killed: You know how much of a high dose of some carcinogen it takes for a certain cancer or death risk. You scale this down linearly to low doses. Then you estimate population doses from the release and a few multiplications later you arrive at a "death toll". You will note the amount of estimates necessary, and the pessimistic assumptions like linear-no-threshold for carcinogen exposure. Most probably, death-tolls estimated this way are upper bounds. Only for smog events, there is a possible calibration of such estimates through excess mortality, and even that is in some doubt, because smog hits the old-infirm-dying-anyways harder.
thehappypm 6y ago
The roof-falling problem is by far the more palatable one for me. The people directly working in the industry are suffering the casualties. There are no completely unrelated deaths or injuries. And it’s fundamentally easier to make solar roofing safer than it is to make nuclear reactions safer.
- NhanH 6y ago
  > And it’s fundamentally easier to make solar roofing safer than it is to make nuclear reactions safer. I do not agree with this statement. Although I'm still thinking on how to prove this one way or another. The closest analogy I can think of is between driving vs flying. It's easy to intuit that driving is fundamentally easier to be made safer than making flying safer. This is empirically not true.
  - Akronymus 6y ago
    Every rooftop is somewhat different. You can't just make all roofs safe. While with nuclear reactors, you can design them to be as safe as needed. (Altough, I do think they are overregulated in certain areas.)
- SamReidHughes 6y ago
  You should expect just the opposite. It’s totally plausible could make nuclear reactions fail safe if you do it right, and I think they’ve figured that out (supposedly). But there’s no way you can force thousands of different people to get up on a roof safely.
umvi 6y ago
> Imagine a system that kills 5k people once every 20 years, versus a system that kills 1 person a day. You end up with scary headlines about the first one, but the second one kills significantly more people over time. "Imagine a virus that kills 300,000 people once every 100 years, versus a virus that kills 20,000 people per year. You end up with scary headlines about the first one (and shut down economies, and political turmoil, and...), but the second one kills significantly more people over time."

by Lazare 6y ago

bgorman 6y ago

How can we know that coal kills 13k people a year in the US? Sure, burning coal releases carcinogens but most of the population lives far away from coal plants.

Lazare 6y ago
This study says 13k: https://www.catf.us/wp-content/uploads/2010/09/CATF_Pub_TheTollFromCoal.pdf https://www.catf.us/wp-content/uploads/2010/09/CATF_Pub_TheT... Estimates vary; some reckon it's more like 50k: https://www.sciencedirect.com/science/article/abs/pii/S1352231013004548?via%3Dihub https://www.sciencedirect.com/science/article/abs/pii/S13522...
corty 6y ago
Similar to how we "know" how many people a low-level nuclear release killed: You know how much of a high dose of some carcinogen it takes for a certain cancer or death risk. You scale this down linearly to low doses. Then you estimate population doses from the release and a few multiplications later you arrive at a "death toll". You will note the amount of estimates necessary, and the pessimistic assumptions like linear-no-threshold for carcinogen exposure. Most probably, death-tolls estimated this way are upper bounds. Only for smog events, there is a possible calibration of such estimates through excess mortality, and even that is in some doubt, because smog hits the old-infirm-dying-anyways harder.

thehappypm 6y ago

The roof-falling problem is by far the more palatable one for me. The people directly working in the industry are suffering the casualties. There are no completely unrelated deaths or injuries. And it’s fundamentally easier to make solar roofing safer than it is to make nuclear reactions safer.

NhanH 6y ago
> And it’s fundamentally easier to make solar roofing safer than it is to make nuclear reactions safer. I do not agree with this statement. Although I'm still thinking on how to prove this one way or another. The closest analogy I can think of is between driving vs flying. It's easy to intuit that driving is fundamentally easier to be made safer than making flying safer. This is empirically not true.
- Akronymus 6y ago
  Every rooftop is somewhat different. You can't just make all roofs safe. While with nuclear reactors, you can design them to be as safe as needed. (Altough, I do think they are overregulated in certain areas.)
SamReidHughes 6y ago
You should expect just the opposite. It’s totally plausible could make nuclear reactions fail safe if you do it right, and I think they’ve figured that out (supposedly). But there’s no way you can force thousands of different people to get up on a roof safely.

umvi 6y ago

> Imagine a system that kills 5k people once every 20 years, versus a system that kills 1 person a day. You end up with scary headlines about the first one, but the second one kills significantly more people over time. "Imagine a virus that kills 300,000 people once every 100 years, versus a virus that kills 20,000 people per year. You end up with scary headlines about the first one (and shut down economies, and political turmoil, and...), but the second one kills significantly more people over time."