Perhaps the lesson is that reactors are just about the most complicated construction projects that humans build, they continuously operate on the edge of disaster, and they require a huge number of highly educated operators. There are too many factors to assure they will be safe, and their failure mode can reach catastrophic levels.
As you obviously know, the Fukushima disaster was not because of quake or tsunami damage to the reactors. It was because of water damage to the diesel generators that circulated the coolant to remove decay heat. Then failures in 3 (4?) redundant backup measure to provide that power occured, too.
What would happen, hypothetically in the event of an human generated EMP (or general infrastructure because of war) or a multi-stdev CME from the sun? What if a sufficiently dedicated group was able to infiltrate a reactor and intentionally damage it? There's always a way to imagine a failure scenario that is possible, even if unlikely.
Thankfully Fukushima wasn't nearly as bad as Chernobyl. Had the fuel breached the containment unit and triggered a steam explosion with groundwater. it would have been worse.
> they continuously operate on the edge of disaster
No, they don't.
> they require a huge number of highly educated operators.
Older designs, like Fukushima, do require operators to take certain actions in the event of failure in order to prevent damage to the reactor. (Newer designs don't even require that, which is certainly an improvement; but that doesn't mean we should misrepresent the features of older designs.) But they do not require operators to take certain actions in the event of failure to prevent radiation release to the outside world. They just require backup power. (Again, newer designs don't even require backup power, which, again, is certainly an improvement. But that doesn't mean we should misrepresent the features of older designs. As another poster noted, a similar reactor even closer to the tsunami had no problems--because the backup power generators were sited properly.)
> There are too many factors to assure they will be safe, and their failure mode can reach catastrophic levels.
This is not true either. Even if we grant that Fukushima has harmed many people, the main reasons for that harm were unrelated to the reactor itself: they were the bad siting of the backup power, and the fact that TEPCO and the Japanese government refused to admit what was really going on and take proper steps to contain it, including asking for help. The same sorts of human errors have caused much more harm in connection with other technologies--for example, consider the people harmed due to coal mining (several orders of magnitude more than those harmed due to nuclear energy).
> There's always a way to imagine a failure scenario that is possible, even if unlikely.
Sure, and this is true of any technology. But just imagining the failure scenario doesn't quantify the risk, and it's the quantitative risk that you need in order to make a fair assessment of the technology.
I'm no friend of coal and would prefer nuclear to it. I prefer solar to both.
I can't go on at length responding to you - too much time on this thread and only so many hours in life^H^H^H^Hthe day. However, I think you really know the issues but are approaching it from a slightly contrarian tack.
I think it's very fair to say reactors operate on the edge of disaster. But I say that in the same sense that motorcycles operate on the edge of disaster (for the rider). It's the controlled fission of a super critical amount of enriched uranium.
Even if there is a hypothetically perfect reactor, imperfect people with build it, run it, make decisions about it, and design/implement the risk control protocols around it. Just look at a few of the almost infinite examples of human factors mistakes in complex systems: Colgan 3407, Air France 447, Chernobyl, Challenger... I would even be kinder to Fukushim Daiichi because 1) the risk at the site was not as well known at the time of construction 2) there were backup power sources beyond the diesel generators that failed 3) most damage was not preventable after the wave hit whether help was requested or not.
People are the weak link and we're pretty much intrinsic to the process.
If solar gets to the point where it can provide reliable base load power, I will too. I keep watching for solar thermal plants to take off.
> It's the controlled fission of a super critical amount of enriched uranium.
No, it's the controlled fission of a critical amount of enriched uranium. Supercritical means the reaction is not controlled--that's a bomb, not a reactor.
As far as operating "on the edge of danger", that depends on how the reaction responds to various changes in conditions, particularly those caused by failures in other parts of the plant. The key is that you want reactivity to go down in response to a failure condition (i.e., you want the reactor to become subcritical, so the chain reaction dies out). Early on, we did not know how to design reactors that had this property under all failure conditions; but now we do, and we have for quite some time (and, as I noted in an earlier post, nobody since the Soviet Union has been insane enough not to make use of that knowledge when designing a reactor).
The problem with the Fukushima reactor was something different: decay heat removal, i.e., once the reactor shuts down, you still have fission products inside the core that are producing heat because they're highly radioactive. At the time the Fukushima reactor was built, we did not know how to design a reactor that could remove that decay heat in a controlled fashion without backup electrical power available. Now we do know how to do that; in fact there are at least two different ways to do it (either design the cooling system to operate using natural convection, or redesign the reactor core so that fission products can be removed and reprocessed while the reactor is operating).
So while I'll agree that early reactors were more like motorcycles, recent reactor designs are more like family sedans, with all kinds of safety features that protect the driver even from his own mistakes. (The Fukushima and Three Mile Island designs were somewhere in between.)
> People are the weak link
Agreed; we humans are almost always the least reliable component of any system that includes us.
> and we're pretty much intrinsic to the process.
Not necessarily; it's often possible to design passive safety features (i.e., ones that work automatically without human intervention), and in systems where reliability is paramount, this should be done whenever possible.
As you obviously know, the Fukushima disaster was not because of quake or tsunami damage to the reactors. It was because of water damage to the diesel generators that circulated the coolant to remove decay heat. Then failures in 3 (4?) redundant backup measure to provide that power occured, too.
What would happen, hypothetically in the event of an human generated EMP (or general infrastructure because of war) or a multi-stdev CME from the sun? What if a sufficiently dedicated group was able to infiltrate a reactor and intentionally damage it? There's always a way to imagine a failure scenario that is possible, even if unlikely.
Thankfully Fukushima wasn't nearly as bad as Chernobyl. Had the fuel breached the containment unit and triggered a steam explosion with groundwater. it would have been worse.