On Wednesday, exactly when the LHC saw the first beams, ITER announced good news, too.
JET, the Joint European Torus, was so far the most successful realization of the Russian "Tokamak" (co-authored by Tamm and Sakharov) toroidal technology (the acronym stands for "Toroidal Kamera/Chamber with Magnetic Katushkas/Coil") to sustain thermonuclear fusion: it was able to produce 16 MW for one second.
ITER, the International Thermonuclear Experimental Reactor (or "way" in Latin), officially funded since 2006 and under construction in Cadarache, France (Google Maps, Google Earth overlay), is designed to break the record and produce 500 MW for at least 1,000 seconds. However, it is a USD 10 billion project scheduled for 30 years and the energy production might occur after 2018 (the energy will be lost because this is an experimental gadget).
Huge currents (52 kA) and magnetic fields (6.4 T) are needed here. Russia managed to construct the superconducting Nb-Ti wires, Europe succeeded in incorporating them into a broader structure, and Japan successfully tested the superconductor during the last 4 months (see above). It will work.
Great. But shouldn't they try to speed up the project a bit?
ITER plans to use the reaction
Deuterium + Tritium -> Helium-4 + neutron + 0.0176 GeVYou see that from the mass of 5 GeV that enters, you produce about 0.4% of energy. The particular reaction above is preferred because it has the lowest activation energy/temperature among the competitors. The optimal temperature is 100 megakelvins (=9 keV: this is no high-energy accelerator!) and one only needs 20% of the produced energy to heat up the plasma at the beginning.
At any rate, if this technology succeeded in a decade or two, I suppose that many industries would be switching to electricity (e.g. heating and a large fraction of transportation; there would surely be a lot of power outlets on the street where you could recharge with your "credit card" of the day). The fuel reserves for thermonuclear fusion are "almost" unlimited.
Deuterium is a part of heavy water which is about 0.016% of sea water. Tritium, being unstable (half-life is 12 years), has to be artificially produced in nuclear reactions, most conveniently from lithium ("breeding").