4 ms·

> This is like saying SARS1 was as bad as a modern pandemic could be... of course it's not. Of course its not. Meaning "of course it's not like saying <insert

> This is like saying SARS1 was as bad as a modern pandemic could be... of course it's not. Of course its not. Meaning "of course it's not like saying <insert false analogy here>". > but with some bad luck, it could be substantially worse. Funny how bad operators + most powerful recorded earthquake in Japan's history + tsunami isn't bad luck worth considering because there's always worse luck around the corner. > that we should be clear-eyed about the actual tail risks while designing them. We are: the greatest disasters involving nuclear power plants have shown that they are incredibly safe. And yet, the perception is "this is just waiting for bad luck to be so much worse". Strangely enough when we look at other sources of energy by deaths from normal operation (coal plants) or historical disasters (Banqjao Dam, 26 000 dead from flood, 145 000 dead from subsequent famine and epidemics, 11 million homeless), somehow no one goes "oh you cannot look at those and state something".

mikeyouse 4y ago
Of course people look at those and state something... if you think that the Banqjao Dam failure is the worst dam failure that can happen, you haven't kept up on the news regarding the Mosul Dam (https://en.wikipedia.org/wiki/Mosul_Dam#Renewed_stability_concerns https://en.wikipedia.org/wiki/Mosul_Dam#Renewed_stability_co...) or Three Gorges (https://www.wsj.com/articles/worst-flooding-in-decades-raises-concerns-over-chinas-three-gorges-dam-11595337875 https://www.wsj.com/articles/worst-flooding-in-decades-raise...).. So new dam projects are built to avoid the obviously worse potential outcomes. I'm not sure how people are misunderstanding tail risks this much but NNT would be ashamed. "Regression to the tail" is super important when assessing long-term risk -- and likely points more in favor of nuclear plants due to the insanely disruptive tail risk that climate change represents. But "our repeated accidents didn't end up being too bad" isn't the way to assess them.. acknowledging that the accidents resulted in thousand-mile exclusion zones, but limited loss of life is important but that should directly inform the siting of new plants and the operation of existing ones (and why it's a good thing Indian Point is shutting down).
- dmitriid 4y ago
  And yet I don't hear anyone look at those and say "oh no, you shouldn't judge by the past, it can always be worse" :-\ > But "our repeated accidents didn't end up being too bad" isn't the way to assess them It's a valid way to assess them. > that should directly inform the siting of new plants and the operation of existing ones What makes you think the new constructions don't take that into consideration? What makes you keep saying "no you shouldn't say itself because earthquake+tsunami+bad operators isn't bad luck enough, there's always worse luck around the corner"?

mikeyouse 4y ago

Of course people look at those and state something... if you think that the Banqjao Dam failure is the worst dam failure that can happen, you haven't kept up on the news regarding the Mosul Dam (https://en.wikipedia.org/wiki/Mosul_Dam#Renewed_stability_concerns https://en.wikipedia.org/wiki/Mosul_Dam#Renewed_stability_co...) or Three Gorges (https://www.wsj.com/articles/worst-flooding-in-decades-raises-concerns-over-chinas-three-gorges-dam-11595337875 https://www.wsj.com/articles/worst-flooding-in-decades-raise...).. So new dam projects are built to avoid the obviously worse potential outcomes. I'm not sure how people are misunderstanding tail risks this much but NNT would be ashamed. "Regression to the tail" is super important when assessing long-term risk -- and likely points more in favor of nuclear plants due to the insanely disruptive tail risk that climate change represents. But "our repeated accidents didn't end up being too bad" isn't the way to assess them.. acknowledging that the accidents resulted in thousand-mile exclusion zones, but limited loss of life is important but that should directly inform the siting of new plants and the operation of existing ones (and why it's a good thing Indian Point is shutting down).

dmitriid 4y ago
And yet I don't hear anyone look at those and say "oh no, you shouldn't judge by the past, it can always be worse" :-\ > But "our repeated accidents didn't end up being too bad" isn't the way to assess them It's a valid way to assess them. > that should directly inform the siting of new plants and the operation of existing ones What makes you think the new constructions don't take that into consideration? What makes you keep saying "no you shouldn't say itself because earthquake+tsunami+bad operators isn't bad luck enough, there's always worse luck around the corner"?