A thorium reactor is an experimental nuclear reactor, intended to be safer and cheaper than the usual uranium-fueled reactors used in existing powerplants.

All by itself, 90Th232(thorium) is a very poor fuel. However, it can be transformed into 92U233, which is a very good fuel. The decay of a 92U233 atom releases 30 to 60 times as much energy as was needed to initiate the transformation.

The thorium reaction cycle is as follows:

  1. A 90Th232 atom absorbs a neutron and is converted into 90Th233
  2. 90Th233 transforms 1 neutron into a proton and ejects an electron from the nucleus (beta decay), becoming 91Pa233(Protactinium)
  3. 91Pa233 undergoes beta decay and becomes 92U233
  4. 92U233 absorbs another neutron and fissions, ejecting two neutrons

A basic thorium reactor consists of a reactor block with a heat exchanger and a controllable neutron source. No control rods are needed to absorb excess neutrons, because they are only provided as needed to keep the reaction going. The specific version designed by Dr. Carlo Rubbia places a toroidal reactor block at the bottom of a shaft filled with molten lead. A tube runs down the center of the shaft, opening into the middle of the reactor block. Protons are accelerated down the tube to impact and fission atoms of lead at the bottom, which release neutrons into the core, a process known as spallation. Pb208, which makes up slightly more than half of all naturally accuring lead, is transparent to neutrons, so nearly all of the neutrons produced this way make it into the core. Heated lead around the core rises by convection to the top of the shaft, where a heat exchanger cools the lead and carries the heat away to make steam to run a turbine.

If the proton beam is shut off, the reaction stops almost immediately. Some heat is still produced by residual fission byproducts. In a conventional uranium reactor, this can be enough to cause meltdown, but in a thorium reactor the lead will just keep circulating and carrying heat away.

Thorium is cheeper than uranium because it is nearly pure and usable as mined, whereas uranium must be refined to bring up the concentration of 92U235 up from 0.7%. The safety of the thorium reactor is due to two main factors:

  1. The decay of 92U233 only produces 2 neutrons as opposed to three. As it requires one neutron to create another 92U233 atom and a second to cause its decay, a chain reaction can only occur with 100% utilization, which is impossible in practice. Ergo, if we turn off the neutron supply to the reactor, it will quickly shut itself off with no need for control rods. Since a thorium reactor cannot sustain a chain reaction, a meltdown is impossible.
  2. Since the fissionable materials are produced as needed, thorium produces far less long-lived radioactive waste than uranium or plutonium reactors.

As far as I know, there are currently no plans to build working reactors as described above. For now, they're confined to small-scale tests at CERN and a spritzing of labs in Japan, Russia, and the US. A quick googling reveals, however, that India is planning on building thorium fast breeder reactors to produce 92U233 for conventional nuclear reactors.

rootbeer277 tells me that the reason the US won't build one of these is that it refuses to build reactors that create or use weapons-grade fuels. Seems like a pretty stupid reason to me, given that in the Rubbia reactor design it is far more difficult to extract the uranium than it is to extract accumulated plutonium from a conventional reactor. Now, not building a Radkowski style hybrid (which mixes thorium with the regular fuel and puts it in the control rods) makes rather more sense.