Tuesday, November 12, 2013

In Where I Hammer Thor...ium, Again

So, once again, I am being sent links from advocates about thorium.  This time, though, it's not about the LFTR or breeder reactors or anything related to that.  This time, it's a little weirder...  it's about a Thorium-powered car.  Now, I am not strictly ruling out the theoretical possibility of such a car just based on that description alone.  One of the nice things about thorium is that you can breed U-233 out of it with relatively low-energy neutrons.  Also, the part about making a thorium reactor small enough that it can fit inside a car is conceivable in principle, although what you'd have is many times more massive than a traditional ICE.  However, there are practical barriers to using it in a car, not the least of which include the public fear of nuclear anything (and no amount of reason or factual knowledge can possibly weigh in).  More than that, the mechanism by which nuclear actually does its thing is tremendous heat output, and it's a fairly huge challenge not just to carry that heat away, but in a car, you also have to prevent the transmission of all that heat to the passengers and also to just about anywhere else within a few inches of your core.  That's not easy.

Using steam to generate power in a car is not an entirely new concept, mind you.  BMW toyed with the idea of using a water-alcohol (vodka?) coolant to draw excess heat away from the ICE's exhaust gases to drive a small steam turbine.  Although it was providing power assist and not generating electricity, it worked in practice, but current hybrid technology probably yields better results overall.  With a nuclear source, though, you pretty much have to generate electricity, and you can generate lots of it on relatively little fuel.  The actual press release as well suggests that start-up times are around 30 seconds, which sounds reasonable to start superheating steam from a dead cold state.  Nonetheless, as you read into it, it's pretty obvious that it's a big fat hoax, and it turns out that all these claims go back a few years as it is, and they've unsurprisingly gotten nowhere.

So what was the obvious problem with it?  Well, that comes when you get a little more specific about the actual claims made.
The first one that raised a few alarm bells was the claim that 1 gram of Th-232 provides more energy than 7400 gallons of gasoline.  While there's no doubt that thorium is more energy-dense than gasoline is, that's the sort of claim that bears double-checking.  U-235 reactions yield about 20 TJ of energy per kilogram, vs. about 48 MJ per kg of gasoline.  This is a pretty significant advantage -- a little over 416,000 to one.  Let's just assume for now that U-233 fission from Th breeding yields about the same making 1 gram of Th equivalent to 416 kg of gasoline.  The density of gasoline is 2.7 kg per gallon.  That means 1 gm of Th-232 has a yield more like that of 150-155 gallons of gasoline...  quite a difference from 7400.  If that figure is meant to be real, that means whatever process they're using is 45-50x more efficient than common light-water nuclear reactors.  A bit of a stretch, to say the least.

Then of course, there's the process itself.  The claim is not that they're using a small thorium breeder reactor, but that they're using an accelerator (presumably to bring neutrons up to appropriate temperature) to drive a thorium-based laser that generates steam to drive a turbine that generates electricity.  While that process of heating steam and so on by itself doesn't sound impossible, it's the thorium laser that starts creating that tailspin of lies.  First of all, if you're already generating all that heat from bombarding and fissioning U-233, why aren't you using that heat to generate steam?  Where does the laser come in, and why?  The idea of a radioactive material to lase is not new, but I'm not aware of any practical application of a Uranium laser (although plenty of research examples exist).  I can understand wanting to focus thermal energy exactly where you want it, but when you're dealing with chunks as small as a few grams of thorium, thermal density is pretty high to begin with.  So the interesting thing is that the article I was linked to had the following quote about the thorium laser --
The key twist to Stevens’ thorium-laser power concept is that it would use a radioactive element-based laser to produce heat, not a beam of coherent light.
Ummm...  if it doesn't produce a beam of coherent light, how is it a laser?  How is it anything more than a warm reactor core?  Sure a beam of coherent light can generate heat, but you've still got a coherent beam. While a hot reactor core can definitely power steam turbines -- heck, that's what nuclear power is in the first place -- that's not exactly the same thing as saying you've got a laser generating steam.  And certainly not one that operates at 50x the output of a conventional reactor.  So maybe you can get more thermodynamic efficiency with a laser than with a normal heat exchanger, but I'm not aware of any physical process that can get you 50x the efficiency of a nuclear reactor, something which is already pretty darn thermodynamically efficient to begin with.

That's aside from the fact that Th-232 breeder reactors are still in the experimental stages in the MW and GW scale.  And yet somehow this guy has a small-scale one generating a few hundred KW that can fit under the hood of a car and operate in a closed-loop steam cycle and drive?  And I'm supposed to take this seriously when there are still so many difficulties when trying to do the same thing at any other scale?

It makes for a series of red flags that I have to say just make it out to be fake in the utmost.  In practice, I think the only sense in which cars will be nuclear in the long run will be that they are electric in some sense (whether that be fuel cells or batteries) and they are backed by a predominantly nuclear infrastructure.  That, at least, is something I can get behind.  This thorium laser car, however, is bollocks.