A large piece of hardware
The plasma fusion research community has released conceptual designs for fusion power plants, and in every one that I’ve seen, “fusion power” means “heat used to produce hot gas”, usually by boiling water to produce steam. The gas drives turbines that turn generators, producing electric power with a typical efficiency of about 33%. In other words, 2 GW (fusion) = 2 GW (thermal) ≈ 660 MW (electric).
Here’s a larger image of the International Thermonuclear Experimental Reactor (ITER), intended to begin operational experiments sometime around 2018:
ITER will confine a 100,000,000 K plasma in a toroidal, ultra-high vacuum chamber surrounded by liquid-helium cooled superconducting magnets that produce a field several times higher than that of an MRI scanner. Most of the energy from the fusion reactions will be carried by neutrons; these both heat the wall and create radioactive material, some of it (tritium) to be extracted and used as fuel. In the proposed follow-on power-producing reactor, DEMO (~2030?), and in commercial machines (~2050?), the inner walls would be cooled by molten lithium or hot helium gas to extract the heat that ultimately drives the turbines. As is often remarked, the fuel would cost almost nothing.
DEMO would be somewhat larger than ITER, and is designed to produce about as much electric power as a medium-sized conventional nuclear plant.
In the ITER image above, note the man in the blue coat.
{ 2 trackbacks }
{ 10 comments… read them below or add one }
Josh Maxwell 05.02.09 at 10:00 am UTC
I must say this is a great article i enjoyed reading it keep the good work :)
Pat Galea 05.02.09 at 6:06 pm UTC
It’s one of the things that’s often amused/bemused me about the fusion reactor designs, that you have this really high tech stuff going on, and it’s basically the heater element in a huge kettle. It just seems like it ought to be possible to directly harness the energy coming off the reactor and convert it directly to electricity.
No, I don’t have any plausible ideas for actually doing that!
M. Simon 05.02.09 at 7:30 pm UTC
Here is a fusion plant that can do direct conversion of particle energy to electricity eliminating the steam plant:
Bussard’s IEC Fusion Technology (Polywell Fusion) Explained
Why hasn’t Polywell Fusion been fully funded by the Obama administration?
Eric Drexler 05.03.09 at 1:46 am UTC
@ M. Simon — I track a range of technologies, and if I had to bet on this one as a potential energy source, I’d bet against. However, if I had billions of dollars to allocate, and if I were acting like a proper rational decision maker (maximizing expected utility under uncertainty, etc.), I’d want to learn more before making any firm decision to ignore it.
A direct exchange of technical arguments between critics and advocates of the technology, followed by with an evaluation of the result by a few knowledgeable physicists who have no stake in the matter, would probably give satisfactory information. This a formal process of this sort could be cheap and fast, and would be a good policy for analogous cases — false negative evaluations can have enormous opportunity costs, and in such instances, policy should be based on something more substantial than informal community opinion.
BTW, I was surprised to learn that the Polywell device is related to the low cost, desktop-scale machine (the Farnsworth–Hirsch Fusor) that can produce enough fusion neutrons to be worth worrying about (though the x-rays are apparently worse). It’s hopeless as an energy source, of course.
Al 05.03.09 at 8:46 pm UTC
I do recall there being some MCF designs where you could draw power directly from the reactions themselves—something about the plasma generating its own current or something like that. But obviously I don’t recall exactly how it would work, and if nobody is pursuing it, it can’t be that easy.
The problem with fusion is that it has always been the “next best thing” since the early 1950s, but it has utterly failed to live up to expectations. (It is, for a variety of very important technical reasons, very, very hard. So far we have managed to make it work well only with an atomic bomb as its starter, and the other place it seems to work well is inside the Sun. Both are pretty extreme environments!) Until it is demonstrated that one can actually get more energy out of it than it takes to start the reaction, it isn’t going to feature into any our *current* energy needs. We don’t currently have 20-40 years to figure out how to do it; we need energy solutions that can be implemented in the next 5-10 years. If I were the Obama administration, I’d fund some research into fusion, but it simply could not be part of a short-term (next few decades) energy policy.
Eric Drexler 05.03.09 at 9:38 pm UTC
@ Al — Speaking of bombs, there was a 1970s proposal for controlled fusion (PACER) that had well-understood physics and guaranteed a large energy gain: it would have worked by scaling up the means for controlling the energy output, rather than by scaling down the energy source. The idea was to generate heat (and boil water) by detonating a series of more-or-less conventional thermonuclear explosive devices in a very large,very strong underground chamber. It didn’t become very popular.
The Chad 05.04.09 at 5:20 pm UTC
As a more direct way to generate energy, couldn’t one design a ‘nuclear combustion engine’? Similar to a car engine, maybe one atom gets fed into the combustion chamber at a time? All sorts of technical limitations, of course, but in theory wouldn’t that work pretty well?
Eric Drexler 05.05.09 at 2:14 am UTC
See the next post, “Nanotechnology and Nuclear Reactions”.
Mark Bahner 09.23.09 at 8:02 pm UTC
Hi,
I agree that fusion is the “ultimate” energy source. But have you considered liquid fluoride thorium reactors (LFTRs)?
Mark
RWJ 01.28.10 at 9:43 pm UTC
Electric power with an efficiency of more than 90% is possible if using aneutronic reactor, without radioactive material(tritium).