Slashdot Mirror
Reactor at Earth's Core?
An anonymous submitter sent in this story suggesting that uranium in the Earth's core may be acting as a giant breeder reactor, generating a large amount of heat and perhaps being responsible for the Earth's magnetic field.
Man, is this article bad.
The article proposes a 5-mile sphere of uranium at the Earth's core as the source of Earth's geothermal energy.
Now, the idea of radioactive decay powering the Earth's geothermal heat generation isn't a new one - in fact, it's one of the more plausible models, as it works quite well over a very long time frame and explains the presence of helium in the mantle. However, the uranium, thorium, and other elements involved would be diffused through the core material (i.e. in solution in the liquid outer core, or as dopants in the iron crystal of the inner core). The absolute concentration of uranium in the Earth is very, very low. Without a really strong mechanism operating to separate it out and concentrate it, it's going to remain a trace impurity in other ores, and not a ball at the center of the Earth.
Secondly, I'm pretty sure that a 5-mile ball would be over the critical mass/volume envelope for an uncontrolled chain reaction of the U-235 and U-233 present in uranium ore. The fact that the Earth hasn't exploded suggests that uranium is not concentrated into a ball. Anyone with the fast-neutron cross section data care to work this out?
Now, on to other questionables.
The Earth's magnetic field is created by movement of the conducting material in the Earth's outer core. The polarity changes in the field are adequately explained by the idea that turbulence destabilizes the dynamo fluid currents every so often. A fascinating article was published about this a while back, but the citation escapes me.
Turbulence happens; it's a known and expected phenomenon. However, the article authours propose no mechanism for their magical solid-state fission reactor to turn on and off every so often to reset the dynamo currents. Thus, I consider the turbulence conjecture the more plausible.
Now, on to Jupiter. While Jupiter undoubtedly also has heavy element fission contributing to its heat, the majority of its heat is expected to come from it essentially continuing to slowly compact itself. The idea is that as hydrogen progresses from gas to liquid to metallic liquid to metallic solid state, it undergoes several exothermic phase transitions (analagous to the heat of condensation for more common substances). This provides a feedback loop that limits the rate of conversion - for example, if a lot of liquid hydrogen starts converting to liquid metallic hydrogen, the boundary layer heats up, which makes it less favourable for the conversion to continue. The rate of conversion for some of these phases is expected to be slow enough for Jupiter not to have reached its final equilibrium composition. If the conversion is still going on, then as it's exothermic, it could indeed explain heat generation in Jupiter.
A similar mechanism involving crystallization of iron was proposed as a source of Earth's geothermal energy, though this is less convincing as the amounts of matter involved are small enough that Earth should have reached equilibrium long ago.
In summary, the "mysteries" that the article attempts to invoke compact reactors to solve are already adequately explained without the need for such an implausible mechanism.