Doubts Cast on Grand Fusion Experiment

New calculations suggest that a megaproject on nuclear fusion may be doomed to fail. Scientists have struggled for decades to demonstrate that fusion could be a practical source of power. The $10 billion International Thermonuclear Experimental Reactor (ITER), a joint project of the United States, Russia, Europe, and Japan that is due to be up and running before 2010 if construction funds materialize, is supposed to prove the case. At its heart, in a giant donut-shaped tokamak, spiraling magnetic fields would cage deuterium and tritium ions. Heated to millions of degrees Celsius, this plasma would fuse and generate abundant power--enough, designers hope, to ignite the world's first controlled, self-sustaining fusion burn.

But that grand vision may be colliding with physical reality, in the form of results that are now roiling the fusion community. Two researchers at the Institute for Fusion Studies (IFS) of the University of Texas, Austin--William Dorland and Michael Kotschenreuther--have come up with what Marshall Rosenbluth, a physicist at the University of California, San Diego, calls "a remarkable intellectual achievement": a new theory of how turbulence rattles plasma in a tokamak.

For decades, physicists designing new tokamaks have been forced to extrapolate from experiments to estimate how fast this complicated turbulence will cause heat to leak across such fields. The IFS work derives the rates directly from basic physics principles. "This differs from all previous attempts to understand [plasma] turbulence," says IFS director Richard Hazeltine, who was not involved in the work. According to computer models based on the theory, turbulent heat conduction in ITER will likely be strong enough to seriously undermine its performance. Indeed, Science has learned that since March of last year, Dorland and Kotschenreuther have been telling ITER scientists and officials that turbulence could shorten the energy confinement time in ITER to the point where, far from generating the 1.5 billion watts in fusion power that ITER's official documents project, it may give back no more than a few times the energy used to heat the plasma in the first place--much too little to ignite a fusion burn.

On the face of it, the calculations are "pretty worrying for ITER," says Rosenbluth, a member of the ITER Joint Central Team, although he thinks they are still far from conclusive enough to seal ITER's fate. But some prominent fusion scientists suggest that the calculations could force a redesign that would delay ITER's construction.