http://www.antipope.org/charlie/blog-static/2010/08/moonshin...
> Firstly, nobody's built a commercially successful fusion reactor yet. ITER plan to build a working test-bed; it's logical successor would be a working prototype first generation power reactor. There are huge obstacles to overcome, not least in developing neutron capture techniques and breeding D/T fuel. These are engineering problems (sorry, annoying paywall) and theoretically amenable to solution — but at a price of billions of euros and decades of work, and even then, it may turn out to be too costly to be a viable competitor for well-understood fourth generation fission technology and a mature waste disposal/fuel recycling chain. And that's before we look to a speculative second generation reactor, running on a different type of fuel, that — because of the higher Coulomb barrier between He nuclei — requires a far higher temperature (on the order of 500M to 1Bn degrees celsius, rather than the relatively chilly 100M degrees C required for D/T fusion).
> Given the average generation time for a new reactor technology of 20-30 years, and development costs on the order of $50Bn-100Bn per generation, we won't be even thinking about prototyping an He3 reactor until 2060 at the earliest.