Permit me to share a few thoughts about US participation in the German stellarator program Wendelstein 7-X (W7-X) and about the best overall strategy for the fusion program (PHYSICS TODAY, September 2011, page 30). While one hopes W7-X will work well, skepticism is also appropriate. A great deal is already known about stellarators, and in many respects that information is not encouraging. For example, the Japanese have built a large tokamak, JT-60, and a stellarator of comparable size, the Large Helical Device. One important figure of merit for any magnetic fusion device is the product of number density n, the temperature T, and the confinement time τ, nTτ. It is roughly proportional to the fusion power divided by the input power. The Large Helical Device has achieved an nTτ of 4 × 1019 m−3 keV s; JT-60 has achieved 1.6 × 1021, 40 times greater.1 

But more important than the debate over tokamaks versus stellarators for magnetic fusion, or lasers versus heavy-ion beams for inertial fusion, is the question of the best strategy for fusion. Up to now, the choice has been pure fusion—using the 14-MeV fusion neutron’s kinetic energy to boil water. But an alternate strategy is fission-suppressed hybrid fusion (PHYSICS TODAY, July 2009, page 24). Also called fusion breeding, it uses the energy of the excess neutrons to breed fissile fuel for use in conventional nuclear reactors—for instance, light water reactors (LWRs).

The concept of fusion breeding was proposed by Andrei Sakharov2 around 1950; Hans Bethe advocated it in 1979 (PHYSICS TODAY, May 1979, page 44). However, the fusion community has always rejected fusion breeding, most likely because it involved partnering with the nuclear industry, something fusion scientists saw as having many environmental, proliferation, fuel supply, and safety problems. Pure fusion seemed nearly perfect by comparison.

As a plasma physicist participating in and observing the fusion program, I have become convinced that Sakharov and Bethe were right and the conventional strategy is wrong. Over the past 15 years, I have documented this view and the science backing it up.3 

Fusion breeding is similar to fission breeding from a reactor such as the integral fast reactor4 (IFR), but with two enormous advantages. First, the fusion breeder is much more prolific. One fusion breeder can fuel about five LWRs of equal power; it takes two IFRs at maximum breeding rate to fuel a single LWR of equal power. Second, an IFR needs a great deal of fissile material to start up, but a fusion breeder needs none. One advantage of an IFR, however, is that once started, it can burn any actinide. Run at a low conversion ratio, one IFR can burn the actinide wastes from as many as five LWRs of equal power.3,4

Fusion breeding is the optimum choice because I think the world will need an additional 10–30 TW of carbon-free power by midcentury.5 The options for achieving that are few. The most optimistic proponents of pure fusion admit that it has no hope of making any major contribution in that time frame. Fusion breeding just might. The requirements on the fusion reactor are considerably relaxed, and that may be of great importance, particularly for the tokamak approach. Limits on density, pressure, and current, which have constrained tokamak operation for half a century, make it difficult to see how it can ever be viable as a pure fusion reactor.3 But it can operate within those limits as a fusion breeder. For any fusion device, breeding is much easier than pure fusion.

Pure fusion has already been delayed more than 30 years. It seems to recede further and further into the future, and sponsors may well lose patience. Fusion breeding is hardly a cakewalk; it will take decades to fully develop. But there are about 400 LWRs in the world today, and about 70 more are in various stages of construction or planning. Proponents claim to have decades worth of fuel. But then what? It is far better for fusion scientists to attempt something that is achievable in the relevant time and that will fit within the likely midcentury infrastructure. Let’s not lose the fusion program because perfect is the enemy of good enough.

Japan Atomic Energy Agency, Naka Fusion Institute, Fusion Plasma Research,;
Large Helical Device Information,
, R. Lourie, trans.,
Vintage Books
New York
), p.
Phys. Soc.
), and references therein,
W. H.
G. E.
G. S.
Phys. Soc.