I've come upon Dr. J. Marvin Herndon's theory that the earth's magnetic field is generated by a hot nuclear reactor operating in the center of the earth. This is backed by various papers, some of them peer reviewed:


His theory purports to explain various anomalies such as the high level of He3 in basalt in Hawaii and their dependency with time. A recent Herndon arXiv paper is:

Uniqueness of Herndon's Georeactor: Energy Source and Production Mechanism for Earth's Magnetic Field http://arxiv.com/abs/0901.4509

His papers suggest that the natural reactors sometimes seen in uranium deposits must also occur at the center of the earth. He says that the conventional explanation for the earth's magnetic field fails because the Rayleigh number for the core is inappropriate as far as determining whether convection exists.

Since this is not accepted geology, what are the problems with the theory?

  • 1
    $\begingroup$ "Magnetic field generated by heat"? $\endgroup$ – endolith Jun 7 '11 at 1:24
  • $\begingroup$ Surprisingly enough, heat does create magnetic fields. In fact, the magnetic fields in the motor in my washing machine back in Albuquerque was generated by heat. Now my washing machine motor magnetic fields are generated by moving water. If you want to know how heat in the core could generate magnetic fields you can read the links. $\endgroup$ – Carl Brannen Jun 7 '11 at 1:39
  • $\begingroup$ Magnetic fields are generated by moving electric charge $\endgroup$ – endolith Jun 7 '11 at 1:59
  • $\begingroup$ @endolith I think we should quickly agree that this discussion is about fission heat indirectly powering the Earth's magnetic field by powering the currents in the core, which then have the ability to move bulk amounts of electric charge. $\endgroup$ – Alan Rominger Jun 8 '11 at 14:34

KamLAND Borexino has set moderately strict limits of the total power of a central geo-reactor. See for instance Geo-neutrino: Experiments (pdf link) a talk by one of my colleagues. (Jalena notes that Borexino's limit is the strongest one going, but KamLAND was the leader for a while.)

The upper extreme of these limits is less than half the total geological power, but quite non-trivial. The bottom goes all the way to zero.

There is also a recent paper (that I have yet to read) on a variation of this idea: The KamLAND-experiment and Soliton-like Nuclear Georeactor. Part 1. Comparison of Theory with Experiment.

I've no idea, how the rest of these ideas stand.

  • $\begingroup$ This is getting more and more interesting. +1 when I get my votes back tomorrow. $\endgroup$ – Carl Brannen Mar 16 '11 at 5:48
  • $\begingroup$ I am in doubt, that anybody can calculate the energy supply to generate earths magnetic field to such precision as then to say its (or must be) fed by a georeactor. What about the radioactive decay of transuranes, the traditional theory? Is it proven to be wrong? $\endgroup$ – Georg Mar 16 '11 at 10:49
  • $\begingroup$ @Georg: As far as I know, radiothermal is still understood to be the leading contributor. $\endgroup$ – dmckee Mar 16 '11 at 15:37

This theory has been around a while. The one problem I see is that in order to get a uranium reactor running you need a fair concentration of U235. The core of the Earth most likely does not have that. There might be a low level or very subcritical amount of fission which goes on in the core of the Earth. However, I suspect this is not larger than the amount of energy produced by strong and weak nuclear decay processes.

  • $\begingroup$ It would be nice to add some calculations or references to this. $\endgroup$ – Carl Brannen Mar 17 '11 at 0:28
  • 1
    $\begingroup$ The needed quantities of U235 are not generally present today as a result of the difference in decay constants between 235 and 238, meaning that a concentrated U deposit is neutronically poisoned in modern times. Man has built reactors with Carbon and natural Uranium, but it's not easy and unlikely for nature to do it. Regardless of one's position on this question, we should all agree that a fission reaction was much more likely in core of the early Earth. Heat from a critical fission reaction >1Gy ago could be powering the magnetic fields now, in fact, that is more likely. $\endgroup$ – Alan Rominger Jun 8 '11 at 14:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.