Hacker News Reader

> Finally, while I am not attacking nuclear power here, it is erroneous to reason about the relative safeties of technologies based purely on the historical track record.

I'm the one who made that argument.

That the safety systems failed points to the systemic flaw in Fukushima: it requires active safety. Better designs, such as those developed in the 40 years since Fukushima was started up, rely on the operation of physical laws to maintain safety in the event of a reactor emergency (edit: It's since been pointed out to me these only lengthen the window of reaction time, but do not remove the necessity to provide active cooling eventually).

It is not specious reasoning to point out that these designs would be even safer than a flawed design which has so far (with active management) survived an event far above its engineering specifications.

The problem with calculating unknown risks based on imaginable catastrophic consequences is that 1) the scope of imagination is large and 2) interested parties to the estimate can skew it up or down. The best known example of this problem is Pascal's Wager.

So.... your argument is that we should stop building nuclear plants using 50 year old designs? I think that problem is already well solved.

New nuclear plants would actually fare far better against these unprecedented natural disasters.

Edit: the problem is this, nuclear power is being put up to a far higher standard than any other power source because of irrational fears. There is risk in any activity, especially providing electrical power. People die from dams failing, from refinery explosions, etc. The question should be whether we can build nuclear plants that are at least as safe as competing power sources, and the answer, even in the wake of Fukushima, seems to be a resounding yes.

"the problem is this, nuclear power is being put up to a far higher standard than any other power source because of irrational fears"

That's probably correct. The question is, why? Perhaps it's because it is so much more difficult to understand. When a dam breaks or when parts of an oil refinery explode, the result may be catastrophic but at least we can understand what is happening, on an intuitive level. When something goes wrong with a nuclear reactor, it is so far removed from anything in our everyday world that it's difficult to understand on that same intuitive level. If something is on fire then we can see and hear it and we can use our experiences of other things being on fire to understand it. If something is radioactive then neither of the above apply.

So perhaps there is something about nuclear that makes it uniquely frightening (even if those fears are irrational). Unless everyone studies nuclear physics, those fears aren't likely to go away any time soon.

The basic reason is unfamiliarity. Unfamiliarity breeds fear and fear breeds irrationality.

Is it worse to die of radiation poisoning than, say, in an automobile fire? Both can be horrific ways to die, but most people today are likely to be more afraid of the former.

>So.... your argument is that we should stop building nuclear plants using 50 year old designs? I think that problem is already well solved.

No. My argument was that the reasoning in Monbiot's article was bad.

>New nuclear plants would actually fare far better against these unprecedented natural disasters.

Glad to hear it. I have a lot of faith in the power of modern design. People can do some amazing things.

> Edit: the problem is this, nuclear power is being put up to a far higher standard than any other power source because of irrational fears.

Yes, thats true. But it is also held to a higher standard, because the contents of a reactor (eg plutonium) are extremely toxic, in a way that no other power source is. The worst case scenario is thus very bad.

>There is risk in any activity, especially providing electrical power. People die from dams failing, from refinery explosions, etc. The question should be whether we can build nuclear plants that are at least as safe as competing power sources, and the answer, even in the wake of Fukushima, seems to be a resounding yes.

However, if out-of-control events had happened just slightly differently, the answer would have been a resounding 'No'.

And that's my problem with this sort of reasoning. Whether the answer was a resounding 'Yes' or 'No' seems to depend on whether there was (for example) another explosion, in an out of control situation, that had already had several explosions.

This is just like the shuttle scenario. Most of the safeguards, which were never designed to be defeated, have been defeated. That there was no release does not mean there was a margin for error.

We nearly - from an engineering point of view - had a level 7 event, in a highly populated, first world country. I am arguing that anyone that derives solace from this situation is being foolish.

I am not arguing the merits of nuclear power. (I think fission should probably play a large role in our medium term energy provision).

Thanks for reminding me about Feynman's comments on the first shuttle disaster. For those curious, here's a nice summary: http://en.wikipedia.org/wiki/Rogers_Commission_Report#Role_o...

It reminds me how very few of those involved in building things really, truly respect scientific rigour and make a serious attempt to actually know their shit. I know because I see it first hand, all the time, including in projects contracted by NASA. There's a ridiculous amount of cross-checking and testing that prevents it from being a serious problem most of the time, but as we can see it can cause oversights that can rear their ugly heads unexpectedly.

And this is why it's troubling to see people pushing so hard for nuclear with the battle-cry "we've figured it all out". We have to keep having it figured it out as long as the reactor as around, until they are fully decommissioned (which takes many years after they stop being productive). And there have to always be a handful of people who know their shit, always paying attention, and who can somehow resist the bullshit.

In 1998 one of the nuclear plants near Toronto, Ontario was taken out of service. During the Fukushima troubles, it sprang a coolant leak into Lake Ontario. Insignificant radiation, they say, and I'm inclined to believe them, but still... I have to worry about a nuclear plant 30km from me that was taken offline 13 years ago?

>Yes; there are few better alternatives. But we are going to have to be very careful about how we build them.

I don't know about that....even just careful would have sufficed in this case (using an intelligent definition of careful). Placing your backup generators at low elevation seems like an amazingly bad design choice. I mean really, how much does it cost to build a very tall tower to house your backup generators? In projects this capital intensive, it is a rounding error.

>What we should be thinking about is what the risk is in future; not what it has been in the past. If I told you I had an energy source that had a known 0.01% of ending the world, each year, but a clean 50 year track record, would you allow me to continue with it? Of course not!

Total hyperbole (ending the world)....but your general assertion is absolutely correct. Nuclear power is statistically by far the safest form of energy, but there is a very large inherent risk, that while not terribly hard to manage, must absolutley be managed properly, without exception. And I'd say in this case, was not handled properly, at all.

I am sadly disappointed in the nuclear industry in general, and also the Japanese...I expect more from them than I expect from other countries. I may be wrong about one or more things though in this assessment, but that's how I think about it at this point in time.

You can't design against a 7m tsunami wave following a 9.0 earthquake. What exactly does low elevation mean when the wave is 7m high? A tsunami wave that reaches 5km from the shore has so much water and pressure that designing against it is nearly impossible.

Your tower would have been nearly as high as the reactor itself http://www.spiegel.de/fotostrecke/fotostrecke-66018.html nobody sane would have proposed something like this before this event.

Looking at past events and errors it's always easy to say that it was a stupid to do it like it was done. Anticipating future events and their inherent risks and designing against them is nearly impossible. That's the danger of nuclear power, it's able to multiply the risks of every natural desaster. You design for a 5m wave - there comes the 7m wave - you design for a 20m wave - the wave was 23m high in some areas.

"Must be managed properly" is not an sentence we should have in anything we design that can have these catastrophic effect, because there's human error, and it will always be there, even in the best planned facilities on the globe.

Everything you essentially say is ridiculous, seems perfectly reasonable to me. Building a very tall tower, just a metal skeleton with skinny legs, to house your bakup power, seems perfectly reasonable to me. Or even simpler, just build that tower on top of the reactor, in this case.