The nuclear situation in Japan got me thinking once again about nuclear power... and its implications for engineering design.
I really dislike nuclear power. In fact, I think we shouldn't do it. But usually, when people criticize nuclear power they point to two issues: safety and waste disposal. My objection is related to those topics, but I'd like to describe it in more fundamental terms. I object to nuclear power because it's a critically incomplete idea.
Operating a nuclear power plant generates nuclear waste. However, as I wrote in 2008, we don't have a good plan for how to dispose of nuclear waste (and 3 years later, we still don't). Seriously, we're producing highly toxic waste that's going to be around for thousands of years... and our best plan is to put it in "long term storage." This is as foolish as taking off in an airplane with the expectation that we'll figure out how to land it someday. I suppose that's alright if you're a test pilot, but not so much if you're doing it with a fleet jetliners, each full of passengers, flying over major cities.
You might wonder, "How close are we to having a disposal plan?" Well, Newsweek recently did an article about a nuclear power advocate from France, named Anne Lauvergeon. In the article, she pointed out that "the technology exists to destroy it [nuclear waste] in a laboratory setting—a technology she predicts will jump to real life within 20 years."
Um, really? If the technology we're depending on won't be ready for real world use until 2030, shouldn't that give us pause? Isn't there a chance our prediction of what life will be like in 20 years could -just maybe - be a little bit off? Sure, go ahead and produce 20 years worth of nuclear waste. I'm sure we'll figure out what to do about it someday...
The same thing applies to all sorts of other designs. For example, the latest GAO report on selected weapon systems pointed out that the Joint Strike Fighter went into production even though three critical technologies are not mature, the manufacturing process isn't proven yet and the testing is incomplete. Yeah, you guys go on ahead. Don't worry - the technology, manufacturing and testing crew will just catch up later.
I promise you, that approach is not on anyone's list of Best Practices.
OK, back to nuclear power. What are we doing with all that spent fuel while we're waiting 20 years for the big breakthrough? Mostly, we're keeping it on site, in some cases storing five times more waste than the container was designed to handle. And of course we're continuing to produce even more waste... without a good plan on what to do with it all.
On the question of safety, Ms. Lauvergeon's aforementioned interview listed a series of catastrophic scenarios and explained that thanks to our robust designs, "Whatever happens, you will have no leak in the air or the ground.” This was about a week before the earthquake hit Japan. Oopsie!
Yes, most nuclear power plants operate with perfect safety records. But when one fails, it fails hard. We can't prevent it 100% of the time, and we don't really know how to deal with it when the inevitable failures occur. Just one more example of how this is not a complete idea.
Also, Ms. Lauvergeon should find a new job.
I'm all for exploration and experimentation. I'm a lab guy at heart. But when the objective is to develop an operational system, one that'll be built in large quantities and used in the real world, where failure has serious consequences on a major scale, we must make sure we have a complete idea, to include things like safety & disposal. Until then, it's best to use other solutions that are already proven and complete.