A fireman friend once asked me, "Know how you tell if the fire you're dealing with includes HazMat materials?" The answer is, "You look for the dead cop holding the flare."
Okay. So, again, I've been seeing the meme of about how nuclear energy is really safe with the added twist of how the Japanese disaster proves this case. Really. The argument is outlined over at Tobias Buckell's blog. Or you can look at this report from the World Nuclear Association. See, nuclear power is safe. No, really.
Except, really, it's not. There's a lot of statistics being thrown around. You need to know that you're being asked to compare apples to rockets with many of these.
As an example, from the World Nuclear Association article, "In over 14,000 cumulative reactor-years of commercial operation in 32 countries, there have been only two major accidents to nuclear power plants - Three Mile Island and Chernobyl, the latter being of little relevance outside the old Soviet bloc."
We'll forget that "the old Soviet block" he's discounting what was once an area larger than the US, Canada, and Western Europe combined. But that's a lot of time, isn't it? It's almost like we know what we're doing. "14,000 cumulative reactor-years" is just amazing. And understand, I'm not disputing that. As a mean the developed world gets 30% of it's power from nuclear sources (in the US it's 20%, in France it's 80%), Coal and natural gas generation account for many times that run time figure, and so has more chance to kill people. But let's look at the article Tobias links to (regarding death per TWH). This is a little more meaningful in that it smooths out the differences between actual total capacity. And in this circumstance, those numbers look pretty good.
Until you get to statements like this, "The (WHO) study in 2005 indicated that 50 people died… as a direct result of Chernobyl. 4000 people may eventually die earlier as a result of Chernobyl, but those deaths would be more than 20 years after the fact and the cause and effect becomes more tenuous. … there have been 4000 cases of thyroid cancer, mainly in children, but that except for nine deaths, all of them have recovered. 'Otherwise, the team of international experts found no evidence for (increased leukemia and cancer).'" First off, you don't "recover" from thyroid cancer, your survive it (with thyroid replacement therapy for the rest of your life). Chernobyl's population in 1986 was 12,000, just to give you some perspective. Next, understand these thyroid cancers are the cop with the flare types of reactions (the's not including the wider range studies). Also, you'll see that the researcher is discounting those "earlier" deaths because they're indirect (ie. not dead cops holding flares). However, for other statistics (such as coal) he's including indirect deaths (such as those from pollution from generation and lung issues in those who extract the coal, which also normally take years to manifest and are considered "early death"). Notice this article doesn't include deaths from people who both extract uranium, enrich it, and work with the end products, he's only counting those dead cops with the flares in the case of nuclear power.
Begin to see the issues? We can discount those 4000 people who may die early because of their exposure during the Chernobyl incident because 1) they may have gotten cancer anyway or from something else and 2) if we're lucky a bus will hit them before the cancer gets them. With nuclear power (and really every single other pollutant source) you get this "discounting" because there's "no direct correlation." This is another way of saying, "It could have been something else that got them." We can discount those who died earlier than normal because of Three Mile Island with the same excuse. And since we don't have any dead cops with flares in Japan, we all escaped without a scratch.
Except we didn't. Catherine Shaffer does a really good job explaining how radiation hurts you. I linked to that article before, but if you didn't read it, you might want to. There's a couple of things I'd like to clarify in her article, mostly about taking iodine pills to avoid thyroid cancer. There's pretty much only one organ in your body that uses iodine (your thyroid). And iodine can become radioactive pretty easily. What iodine pills are meant to do is flood your system with "clean" iodine and bind up all the receptors, which means that the radioactive forms will be excreted unused. This is the reverse of chemo which fills your body with poison with the hope of killing off more cancer cells than normal cells (odds are in that favor as cancer cells consume more and reproduce faster than your normal cells, all of which means they're big consumers). It's a percentage risk. Most times it works, sometimes it doesn't. Plus, until you excrete the iodine, you still have dangerous radioactive particles in your body doing damage. Also, please note the correlation between radiation exposure levels and time.
XKCD comes in with a goodly chart. And Phil Platt also has a good article covering this topic. Both come to the conclusion that you have nothing to worry about. Except that both expect that you're not looking closely (in both case, I don't believe this is intentional, but it is a scientific/statistics blind spot). In the paragraph (and the linked article) where Phil talks about living in Denver, about how much higher it is and how you'd expect more cancer. Colorado has higher radon concentrations and receives more solar radiation that most of the country. The article he links to discusses how skin cancer rates are higher, as you'd expect, but lung cancer rates are lower (which with higher radon concentrations, you wouldn't expect). Sounds like a wash, doesn't it. Except Colorado's radon exposure isn't that much higher that much of the northern US (and here's a place where you can find your state's radon information). Notice their disclaimers that these lower levels could be attributed to other sources. Well, yeah. They could be (CO is ranked 35th in number of smokers, compared to Ohio which is 13th).
Phil is also a scientist. Understand that limitation (I'm not saying it's a bad thing, but a limitation). If there isn't a study to prove the connection (with discounting all other factors, like the smoking), he won't say there is a connection. Most of the studies I've seen haven't been able to rule out all other factors, which means they're "inconclusive." Which means until it's proven, it's not proven and you go with the default "no connection."
Also, take another look at the chart XKCD put up. It makes you feel better, doesn't it? After all, the extra dosage right next to Fukushima is comparable to, say, a mammogram. Notice closer, that is a 1 day dose (and there were two days of that level). Also keep in mind, radiation damage is a function of dose over time (notice the descriptions on chart include time or instances). For those "yearly maximum" dosages allowed, those are cumulative numbers (the "normal yearly background dose" talks about that a little bit). So all the talk about "you get more radiation from the sun than you would from this reactor" you can see as throwing sand in the eyes. See, you've evolved to handle that dose from the sun and other environmental sources. And you have some tolerance for even higher exposure rates. How much over is still a matter of debate. Also, that background rate may (we say "may" because nobody has been able to show a link) help lead to "normal" cancer rates. That is, the lifetime accumulation of damage from background radiation can contribute to the norm of cancer rates (before you add in other factors like smoking, etc). Don't believe me? Go listen to longevity experts talk about reducing "free radicals."
Just a quick note here. We're all going to die (sorry to break this news to you). Our increasing life expectancy hasn't really lengthened our lives. If you make it to 65, you have the same chance to see 85 as your great grandparents. The difference is that more of us are making it to 65 (which pulls up the average). But if traumatic injury or infectious disease doesn't get us, organ failure or cancer (or the complications from these) will. That's really how we die. So as a matter of living on this planet, cancer is a possibility. And if you live long enough, it's a probability.
Also, if you listen, you'll hear about how radiation dosage limits are "set really low." This is a function of how the dosages have been set lower as more and more data comes in. Here's a little article on epidemiological studies on radiologists and radiologic technicians and their death rates from cancer. If you want, you can just read the abstract (it's the first paragraph). I'll note sentences 3 through 6. Basically it says the stats on hard cancers (organ cancers) aren't consistent (because lung cancer is not pancreatic cancer), however there is an increased risk of leukemia (blood cancer) in practitioners who started before 1950 (this was when acceptable radiation levels were higher). Also, there's a higher risk if you work longer hours early in your career. So, it looks like cancer rates are dropping. Right? Except now read the last sentence. Because we've continually dropped the dosage levels, we haven't been at the current levels long enough to see if there's a real effect (this is also a semi generic scientific statement that means, "we don't have enough data to reach a sound conclusion either way, so we're punting in case anybody calls us on it and our hunches are proven wrong later"). So, yeah, we don't know if those "acceptable dosage levels" are actually "acceptable" or not. So, no, those levels are not "low." Because higher levels have been shown to cause problems. To counter my argument, yes, the numbers are typically cut either in easy fractions (half, quarter) or by factors of ten. So they are big cuts. However, acceptable radiation exposure dosages, in my lifetime, have been cut 3 times that I can remember (the last in the late 90s). Each time because it was found what was a "way too big reduction" actually wasn't enough.
I think I need to wrap this up before it becomes a dissertation (too late!). I'm trying to stay positive. However it's too soon to wipe our brow and say, "Whew, dodged that bullet." Radiation is a slow bullet. Low level radiation sickness can take weeks to present, and months to die from. Even if we stay below those levels there are troublesome signs. Milk and spinach are taken off the market because they've shown up with contamination. Both of these are the result of fallout (notice how nobody is using that word?). Radioactive particles fall into the spinach leaves (remember that you have to wash spinach to keep it from being "granular" - that's dirt) and onto the grass the cows eat (it's how their milk gets radioactive). I'm hoping it's a surface dusting. However, if the fallout was dense enough, this problem isn't going away for decades (the other option for this is the radiation made it to the soil, which the spinach pulled up while forming nutrients, and the grass the cows ate did the same, instead of the particles just being on the surface). Again, my experience with Fernald comes in here. Two decades after the plant stopped all fission activity, there were 3 months (or more) out of the year the local farmers had to dump their milk and crops (because the radiation levels were too high, the rest of the year wasn't much better, but they were below EPA, NRC, and FDA guidelines and could be sold). Mostly this was in the spring and fall. So, I'm hoping it's just a dusting.
At this point you'd expect to see incoming radiation (fallout) hitting the west coast and we're not seeing as much as expected. It's a big ocean out there. It's also where we get a lot of food (especially the Japanese).
But, hopefully, yes, it will all turn out okay. But don't blindly believe everything being pushed in your direction.
edit A late afternoon update, looks like much of the radiation released is radioactive iodine which has a half life of only 8 days. So the resulting long term exposure risks won't be there. On the other hand, the WHO says the food contamination is more wide spread and the crisis is more serious that first thought. If the first report is to be believed, we might end up seeing a lot of kids with goiters in a few years and not much more, which is a good thing (compared to what it might have been).