Non-agenda sources of radiation health risks?

Mar 20, 2011 03:25

So, I'm having a hard time parsing the health risks associated with various scenarios for the Japan reactor collapse. Most that I come across are either the admirably honest "I don't really know," or of the "No sweat, nuclear is teh awesome" or the "BAN ALL NUCLEAR POWER NOW" variety (last two slightly exaggerated). None of those are helpful ( Read more... )

Leave a comment

Comments 14

easwaran March 20 2011, 21:19:52 UTC
Why doesn't the XKCD graphic count as better? I suppose it doesn't tell you the chances of having all these different eventualities, but is there more that you're looking for?

I'm also not convinced that getting rid of nuclear is essential - this earthquake was a more massive one than anything that has happened anywhere near Japan in recorded history. Of course, we'd like to go for entirely solar/wind if we could, but hydro is really no better - in fact, hydro is likely to be much worse, because its failure mode in an earthquake could wipe out a major city pretty easily. (See St. Francis dam and the similar, and quite lovely Mulholland Dam that Matt and I walked around last weekend, which could easily flood central Hollywood in a bad earthquake.) The problem is that anything that releases large enough amounts of energy to provide our baseline power generation is going to be dangerous - large amounts of energy are inherently dangerous ( ... )

Reply

killtacular March 23 2011, 06:00:00 UTC
Oh, I'm not convinced that getting rid of nuclear is essential either. I'm still ambivalently in favor of nuclear. The xkcd graphic doesn't count because it doesn't total radiation exposure over time. Which is pretty crucial, and actually a little disappointing (for me, for the comic).

I suppose what I am looking for are honest probabilities for various catastrophic situations. What has happened in Japan (and the initial US response) has made me, I suppose, be pretty distrustful of the values nominated by the usual suspects.

Good point about the potential catastrophes resulting from a dam breaking, though.

And, again, I'm still in favor (I suppose) of nuclear over coal-fired plants. The question is whether we should go to, say, "wind/solar/natural gas" or "wind/solar/nuclear." This is making me rethink my commitment to the latter until I have numbers I am confident in. Which, at the present, I am not.

Reply


easwaran March 20 2011, 21:20:44 UTC
But yes, I agree - it's remarkably hard to get information about nuclear that isn't entirely pro or anti. The Union of Concerned Scientists has generally seemed good to me, but maybe it's because they've always been cautiously pro-nuclear, which is a more nuanced position than just about anyone else has ever taken.

Reply

killtacular March 23 2011, 06:01:39 UTC
Ya, the Union of Concerned Scientists has been about the best I've found, but I still kinda suspect something of an anti-nuclear bias there, and am not knowledgeable enough to know whether that is the case.

Reply

easwaran March 23 2011, 16:58:34 UTC
That's interesting - my worry with UCS is that it's a pro-nuclear bias, even if a relatively moderate one.

Reply

spoonless March 24 2011, 04:09:55 UTC
I'd expect they have a pro nuclear energy bias, and an anti nuclear weapons bias.

Reply


airstrip March 21 2011, 06:13:58 UTC
I think the problem is that there honestly aren't that many nuclear incidents and so there isn't much opportunity to study the health risks. Most of the data comes from a handful of events: a few first world nuclear accidents, the atomic bombs and people who become exposed while handling nuclear material (mostly for medical use). That's not a lot of data points to begin with and you're going to lose most of these because doctors aren't in the business of research.

I wouldn't suspect substantial health risks, however. Even people that were caught up in Chernobyl are largely without any symptoms. It also doesn't seem like we have breathless reports from Golden, CO where a major Superfund project was cleaning massive contamination from a nuclear weapons manufacturing facility.

Reply

killtacular March 23 2011, 06:03:33 UTC
Ya, you are definitely right about the limitations of what data we have. I think we could do better about data for the likelihoods of various types of accidents, but what that would actually mean does seem likely to be obscured.

I am also suspect of the "Chernobyl wasn't that bad" reports.

Reply


airstrip March 21 2011, 10:33:18 UTC
http://www.boingboing.net/dose%20table.png

Dose table put together by the Mayo Clinic. Read the footnotes carefully.

Reply

killtacular March 23 2011, 06:06:53 UTC
Ya, that makes sense. But looking at numbers like that also means that smoking is no big deal. Smoking a half a pack a day of cigarettes will give you similar results. People die of all sorts of things all the time. The problem is figuring out how much of an increase a Japan-catastrophe would result in.

Reply


Introductions anonymous March 22 2011, 10:03:04 UTC
Hello, nice to be here!
--------------------------------------------------
[url=EHALTHDIAATARY.INFO
GKIDERWAPKER.INFO
[/url]

Reply


Leave a comment

Up