The lost nuclear fusion reactor design?

I had the following communication from Rider, in 2007, via email:
Thanks for tracking me down. Here is the info I have been sending out in response to similar recent inquiries. Please let me know if you still have any questions after reading it. You are quite welcome to post this complete response and the attached files anyplace on the web where they might be helpful to other folks with the same questions. I am nearly computer illiterate when it comes to the internet… 🙂
Once upon a time, I was wowed by the idea of IEC fusion. When I was an undergrad working in other fields, I picked up T.A. Heppenheimer’s The Manmade Sun, a history of the fusion program. It contained 2-3 pages on the Farnsworth-Hirsch poissors/fusors and their eye-popping 10 billion neutrons per second; a separate section covered Bussard’s Riggatron tokamak project and its seemingly premature death. I was fascinated by the idea of scientific underdogs with great ideas, and I devoured all of their papers and patents. That made me decide to pursue fusion (among other things) in grad school. The two strands of fusion ideas merged when Bussard picked up the old Farnsworth approach about the time I was starting my thesis research at MIT. I hoped to take their cumulative work and find ways to push it a few steps closer to realization. To bring myself up to speed, I first went back through all of Bussard’s and everyone else’s basic calculations about the approach. And I found massive errors everywhere–incorrect assumptions and/or incorrect calculations of a variety of effects. The rest is history… I assume you know about my M.S. ’94 and Ph.D. ’95 theses from MIT, as well as my various journal articles and conference presentations based on them. The M.S. thesis points out all the errors and fatal flaws in IEC, and the Ph.D. thesis generalizes the results to limitations on all IEC and all non-IEC fusion approaches.
In a nutshell, there are a large number of fatal flaws with IEC and related approaches, each flaw independent of the others and each far more than sufficient to keep IEC from ever producing net power. This isn’t simply a “he said, they said” disagreement. These are all very well documented and established plasma physics effects, found in textbooks all the way back to the 1950s and able to be accurately predicted for lab experiments. Moreover, the physical basis of each problem can be explained even without the accompanying mathematical analyses. These problems include, but are not limited to:
(1) Bremsstrahlung radiation. Ions must be given very high energies (equivalent to very high temperatures) to stand a decent chance of fusing when they run into each other. Unfortunately, the ions are very generous and promptly give much of this energy to the electrons. Electrons have a very low mass and are easily pushed around whenever they run into an ion or even another electron. Every time an electron runs into something, it slams on its brakes and emits a loud squeal–a bremsstrahlung X-ray. Thus a great deal of the input energy gets turned into useless X-rays, not fusion reactions. This is true for all fusion approaches, not just IEC. Bremsstrahlung is a relatively insignificant problem for deuterium + tritium fuel, which fuses at comparatively low energies. It is a large but theoretically tolerable problem for deuterium + deuterium and deuterium + helium-3 reactions, which require higher energies. Even with the most optimistic assumptions, the bremsstrahlung power loss is much larger than the fusion output power for all other fuels, including helium-3 + helium-3, proton + boron-11, and proton + lithium-6. My theses and papers used the most accurate standard methods of calculating the ion-electron energy transfer and bremsstrahlung losses, including relativistic and other effects. Nonetheless, I was able to show that the ion-electron energy transfer or bremsstrahlung would have to be miraculously lowered by a huge factor, at least 10 to 100 times, in order for the X-ray losses to be tolerable. In all of the papers I have seen from IEC and related researchers, bremsstrahlung is ignored entirely, the ion-electron energy transfer is ignored, or one or the other is merely assumed to be artificially low without providing any physical justification.
(2) Ion-ion collisions. Ions are all positively charged and thus strongly repel each other at close range. Only every once in a while do they quantum-mechanically “tunnel” through that repulsive barrier and fuse together. Even at high energies, two colliding ions are 100-1000 times more likely to randomly scatter off each other than to fuse. IEC, colliding-beam fusion, and related concepts assume that the ions all keep the right velocity pointed in the right direction. Yet collisions will turn these organized beams into a bell curve distribution of velocities going in all directions 100-1000 times faster than the time that would be required for a significant number of the ions to fuse. IEC and related approaches generally provide no mechanism to “herd” the ions back to the desired velocities, but even if they did, my Ph.D thesis and papers showed that even the most efficient such mechanism would consume too much power. You are fighting entropy, and entropy will win. Just like death and taxes.
(3) Electron-electron collisions. A very similar process occurs for the electrons, converting any preferred electron distribution into a random bell curve distribution. Again, any attempt to fight this entropy generation would consume far too much energy, as shown in the theses and papers.
(4) Counter-streaming electrostatic instabilities. Particles in orderly colliding beams are determined to reach a three-dimensional bell curve distribution. In addition to individual collision events as described above, the whole population of particles can act collectively via electrostatic or electromagnetic fields to rebel. Such mass rebellions are called instabilities. Colliding beams and particles outside the central core of an IEC are subject to the well-documented counter-streaming instabilities, in which particles bunch and unbunch like a Slinky, rapidly destroying any initial order in the system.
(5) Weibel electromagnetic instabilities. These approaches are also subject to Weibel instabilities, in which ions and electrons wiggle side-to-side, again destroying any initial order in the system.
(6) Ion upscattering losses from IEC well. The basic idea of IEC is that all the ions have too little energy to escape from the electrostatic potential well in the center of the plasma. However, once collisions and/or instabilities have randomized the ion distribution, ions in the upper tail of the bell curve will have enough energy to escape. Once they do, other ions are randomly scattered to reform the tail of the bell curve, and then they escape too. These escaping ions represent a power loss greater than any fusion output power.
(7) Losses of ions hitting grids in gridded IEC (Farnsworth-Hirsch) devices. Most IEC devices have high-voltage grids inside the plasma to create the potential well. Even with very optimistic assumptions, the power lost by ions hitting these grids is far larger than the fusion output power.
(8) Losses of electrons hitting grids in gridded IEC (Farnsworth-Hirsch) devices. Likewise, even with very optimistic assumptions, the power lost by electrons hitting the grids is far larger than the fusion output power.
(9) Losses of electrons escaping from magnetic cusps in Polywell IEC devices. Instead of grids, Bussard’s Polywell IEC approach uses a polyhedral cusp magnetic field to confine the electrons. Even with very optimistic assumptions, the power lost by electrons escaping at all the cusp points is much larger than the fusion power. As I showed in my theses, any foreseeable method of extracting part of the energy of the escaping electrons would either be vastly inadequate or would actually increase the ion losses instead.
(10) Arcing in direct electric converter and other high voltage grids. Most IEC devices have high-voltage grids, and additional high-voltage grids are frequently proposed to extract energy from escaping particles or fusion products in IEC, colliding-beam reactors, and related approaches. Marshall Rosenbluth, the widely acknowledged “dean” of plasma physics, has shown that at plasma densities typical of proposed full-fledged fusion reactors, arcing between these grids due to all the charged ions and electrons would be intolerable (Plasma Physics and Controlled Fusion, vol. 36, pp. 1255-1268, 1994).
These effects may appear relatively small in the sort of low-density, short-pulse IEC or colliding beam experiments that have typically been used to date. Yet the standard formulas of plasma physics, strongly supported by over 60 years of theoretical and experimental results in this field, indicate that every one of these effects will be intolerably large in a full-fledged reactor-scale device. In order for IEC or related approaches to produce net fusion power, they must defy not just one of these standard predictions but *all* of them. And they must be not just a little better than predicted, but for several of these effects they must be a factor of ~100 better than predicted by the basic physical laws. To date, I have not seen any papers from Bussard, Rostoker, Kulcinski, or others that seriously address any of these problems, let alone propose feasible ways to avoid the problems.
In my opinion, there are several conclusions to be drawn from all of this:
(A) Due to problem (1) above, the *only* fusion reactions which can theoretically produce net power in *any* foreseeable fusion reactor, whether conventional or unconventional, are deuterium + tritium (easiest but quite radioactive), deuterium + deuterium (also radioactive), and deuterium + helium-3 (cleaner but still radioactive).
(B) Due to problems (2) through (10) above, the chances of IEC and related approaches producing net power even with those reactions are minute. (Deuterium + tritium may at best be marginal in some non-IEC beam-like systems, but its radioactivity makes it less appealing than the other reactions, and in any event more conventional fusion approaches like tokamaks stand a much better chance of burning it.) No matter how large or how small an investment the IEC researchers may be requesting, potential sponsors should be aware that their gamble is virtually guaranteed not to pay off, due to all of these fundamental physics problems that can be easily seen in advance.
(C) Despite these problems, IEC or similar approaches could be useful for applications *other* than power production. An IEC device is basically a very compact particle accelerator and could be useful for the same applications as larger accelerators, including radioisotope production and production of radiation, especially neutrons.
(D) Several non-IEC, non-colliding beam approaches are not as subject to problems (2) through (10) above and thus are theoretically able to produce net power with deuterium + helium-3. In fact, Bussard, Rostoker, and Kulcinski have worked on many of these approaches, including field-reversed configurations and tokamaks. The engineering challenges for these reactors are admittedly very difficult and may even ultimately be insurmountable, but that cannot be proven from the outset, so such approaches are certainly very worthy of further work.
(E) I believe fusion research is more limited by the lack of good ideas than by the amount of funding available to build larger and larger machines. Perhaps someone could think of a confinement system that would be better than tokamaks and other current approaches that are theoretically feasible but problematic from an engineering standpoint. Or better yet, perhaps someone could find an efficient method to catalyze fusion reactions, so that a given input energy would yield significantly more fusion output. (Spin-polarized and muon-catalyzed fusion were two clever but ultimately inadequate attempts to do just that.)
To summarize, I would love to see a working, power-generating fusion reactor, especially one that would be less radioactive than the deuterium-tritium reactors that are normally proposed. There are many technological and physical challenges that must be overcome in order to do that, and to overcome them one must openly acknowledge those challenges and then find clever ways around them. Unfortunately, the inertial-electrostatic confinement and colliding-beam fusion researchers have not even attempted to do that. They may choose to ignore the basic laws of physics in designing and building their prototype fusion devices, but when the switch is turned on, I doubt that the laws of physics will return the favor by ignoring them.
If you would like independent confirmation of the physics principles I have outlined, please feel free to contact Peter Catto at the MIT Plasma Science and Fusion Center (catto@psfc.mit.edu, 617-253-5825) or Bill Nevins at Lawrence Livermore National Lab (nevins1@llnl.gov, 925-422-7032). They are both familiar with the IEC approach and my analyses.
I am attaching five .pdf files of several of my journal articles and conference presentations on this topic. Please let me know if you have any problems opening these files or if you have any other questions.
=============================================================
N.B. I either don’t have the pdf’s, anymore, or else they are stashed away on some hard drive in storage. Rider’s email (at the time) was thor@ll.mit.edu .