Hacker News Reader

Comment by LawTalkingGuy | Hacker News Reader

  "The total number of deaths already attributable to Chernobyl or expected in the future over the lifetime of emergency workers and local residents in the most contaminated areas is estimated to be about 4000. This includes some 50 emergency workers who died of acute radiation syndrome and nine children who died of thyroid cancer, and an estimated total of 3940 deaths [...] lifetime of about 600,000 people under consideration." - WHO https://apps.who.int/mediacentre/news/releases/2005/pr38/en/index1.html

The older Linear-No-Threshold model had predicted that small doses of radiation would have a small chance of causing harm. From watching Hiroshima, Chernobyl, and nuclear accidents, our new models have thresholds below which radiation has no observable effect over the lifetime.

Chernobyl was a disaster but its death toll is equivalent to a month of North America's opioid epidemic.

View on HN

That estimate is contested by many people including the scientists who did the work it was based on.

But it's more than enough to make my point. The hypervigilance around nuclear is the only thing stopping monthly chernobyls, as evidenced by the sheer malice and incompetence exhibited in every nexus of power and danger we are not that hypervigilant about. Bringing up the opioid epidemic makes that point pretty clearly.

It's like arguing that dimethyl mercury is perfectly safe and should be used to fuel passenger liners in spite of costing many times as much and requiring you to bring your exhaust with you because the one time someone tried to use it as a rocket fuel someone with sense told them to go f themselves and to never call the chemistry lab where it was made again.

> That estimate is contested

Find the deaths. We'd love to see better research.

> hypervigilance around nuclear is the only thing stopping monthly chernobyls

A few airliners caused thousands of deaths and billions of dollars of damage. Turns out that a lot of our infrastructure is dangerous. You should learn about dams!

But this is a reason to double-down on nuclear. Pretty much every existing reactor is bespoke because we didn't build all that many and technology was being developed the whole time. And they're still safer than everything else. Volume building would let us improve our design and perfect our execution.

> Bringing up the opioid epidemic makes that point pretty clearly.

Yes, virtue-signalers are the biggest problem in both cases. In nuclear it's the "green" fear mongering and in the opioid epidemic it's the "compassionate" free-drugs movement.

> Find the deaths. We'd love to see better research.

LNT is used at the low levels of nuclear accidents because the deaths are impossible to find (and because LNT is mechanistically plausible). They are lost in a sea of ordinary cancers, statistically undetectable by any practical experiment.

Note that there is no good evidence that LNT is incorrect. The NRC responded to a recent petition to abandon LNT with a polite slapdown of the petitioners, for this reason.

This does NOT mean cancers predicted by LNT can ignored. Regulation is not a game of technologies being innocent unless proven guilty beyond reasonable doubt.

If anything, nuclear stans should fear abandonment of LNT. Without it, the regulatory framework could become more conservative, with the assumption that the number of cancers caused is the maximum not ruled out by evidence. For low levels of radiation, this would result in LARGER estimates than from LNT.

> LNT is used at the low levels of nuclear accidents because the deaths are impossible to find [...] lost in a sea of ordinary cancers, statistically undetectable by any practical experiment.

No, this is the 'excess deaths' problem we just encountered estimating Covid fatalities. Either there are unexplained deaths that you can call low-dose deaths/illnesses, or there are not. If the influence per-person is so low it can't be detected that's the same as not causing many deaths/illnesses in the population.

> Note that there is no good evidence that LNT is incorrect.

Yes, there is. High natural background-radiation areas don't show an increase in cancers for residents.

> If anything, nuclear stans should fear abandonment of LNT

It's not a game and we aren't rooting for a team. We want the most accurate models.

> the assumption that the number of cancers caused is the maximum not ruled out by evidence.

This is the standard we should be using, but Chernobyl is the evidence you're talking about. We're actively developing the models based on observations.

> For low levels of radiation, this would result in LARGER estimates than from LNT.

No, that is the LNT model.

Wow, so much wrongness...

> excess deaths as in Covid

Deaths from a nuclear accident are much more diffuse than deaths from Covid were. Unlike the deaths from Covid, below a certain rate they are impossible to detect. The statistical noise in ordinary cancer deaths swamps them.

> Yes, there is. High natural background-radiation areas don't show an increase in cancers for residents.

This claim is directly rebutted by the NRC in their response to the LNT petition. The results cannot be disentangled from non-radiation effects on cancer rates. Epidemiology is a blunt instrument.

https://www.regulations.gov/document/NRC-2015-0057-0671

(look under "Comments Supporting the Petitions—Assertions That There Are No Observable Adverse Effects From Background Radiation")

> It's not a game and we aren't rooting for a team. We want the most accurate models.

No, what you want is something confirming your biases. This is apparent from this nonsense you are emitting, nonsense that is in direct contradiction to the non-cherry-picked and properly interpreted evidence.

> No, that is the LNT model.

Obviously not. There is some nonzero radiation level at which LNT is at the limit of statistical detectability. Below that radiation level, the effect by LNT becomes progressively weaker. So, at those lower radiation levels, the conservative estimate of the effect (as I defined it) will be larger than the prediction of LNT.

Linear No-Threshold doesn't account for bioaccumulation of radioisotopes which is the more likely source of radiation-related disease and death.

Reply on news.ycombinator.com