Tokamak research entered a new and important phase last month. Shortly before midnight on 9 December, the 11-year-old Tokamak Fusion Test Reactor at the Princeton Plasma Physics Laboratory was fired for the first time with a 50-50 mixture of deuterium and tritium. That's the mixture of hydrogen isotopes envisioned for the first generation of fusion reactors, because DT fusion can be harnessed at much lower temperatures than DD or D3He fusion, and the cross section for the proton-proton reaction that powers the Sun is hopelessly small. But until now most tokamak research has been done, for practical reasons, with pure deuterium plasmas.
Two years ago the Joint European Torus in England, a larger machine of TFTR's generation, did perform two shots with a 10% admixture of tritium. JET's tritium experiment was purposely limited so as to minimize radioactivation of the machine. Not only is tritium a radioactive beta emitter; it also generates 14-MeV neutrons when it fuses with deuterium. The accumulating neutron flux from an extended series of 50:50 DT shots gradually activates a tokamak's superstructure.
That's less of a problem for the TFTR, because it is scheduled for retirement this fall, after a series of about a thousand DT shots, to free up funds for the proposed construction of PPPL's next major undertaking: the Tokamak Physics Experiment. The $600 million TPX would be this country's first fully superconducting tokamak.(See Physics Today, November, page 79.) One of the principal purposes of this new national facility would be to study techniques for running tokamak reactors in a continuous, as distinguished from pulsed, mode. If the TPX is not eventually funded, the useful life of the TFTR could be extended by the installation of remote handling equipment.
The reaction that is to power the first generation of tokamak fusion reactors is 2H + 3H ^ n(14 MeV)+4He(3.5 MeV). Pure deuterium plasmas are very useful surrogates for the requisite DT mixture when it comes to the investigation of magnetohydrodynamic stability and energy transport. But they don't generate the energetic alpha particles (helium nuclei) that are essential for maintaining the temperature of an ignited plasma. (The even more energetic neutrons deposit almost none of their energy in the plasma. Their function is to carry useful energy out to the external world.)
So the crucial physics of alpha heating has remained a terra incognita. Are the alphas adequately confined in the plasma, and how efficiently do they heat it? Do they contribute to magnetohydrodynamic stability, or instability? Veteran theorist Marshall Rosenbluth told us that "we've all been speculating for decades about what MHD effects might be induced by the large gyroradii of the energetic alphas. Now we'll finally see." Then there's the question of isotope effects: What are the consequences, if any, of the greater atomic mass of the triton (irrespective of its fusion proclivity) on the macroscopic MHD stability and microscopic transport properties of a tokamak plasma? The first shot:
On the night of 9 December, with Rosenbluth and former PPPL Directors Harold Furth and Melvin Gottlieb in festive attendance, the very first 50:50 DT shot generated a peak power of 3 megawatts of fusion energy. (Each of these experimental shots lasts for only a few seconds.) That was, admittedly, only 12% of the 24-MW input heating power from the beams of energetic deuterons and tritons fired into the plasma. But it was, far and away, a world record.
By the following night the peak power had risen to 6.2 MW. The total fusion output energy integrated over the duration of that shot came to 3.5 megajoules. In 1976 Furth had set "1 to 10 megajoules" as a target for the culminating tritium phase of the proposed TFTR. The more ambitious goal of demonstrating "scientific break-even" would require that the output fusion power match the input heating power. The experimenters at JET plan a brave assault on that goal with tritium in 1996. But break-even, or even its simulation in a pure deuterium plasma, is probably beyond the capabilities of the TFTR, whose plasma volume is five times smaller. The smaller the surface-to-volume ratio, the easier it becomes to confine a plasma's heat.
The TFTR experimenters gradually raised the output power during this first series of shots by increasing the tritium fraction beyond 50%. Of course one wants equal numbers of deuterons and tritons in the plasma. But years of pure deuterium running have saturated the graphite inner walls of the TFTR with adsorbed deuterium. (See the cover of this issue.) The result is that one has to inject more tritium than deuterium to get an effective 50:50 fusion plasma.
The first week of DT shots at Princeton has already provided some tantalizing glimpses into the new territory of the tritons and their alpha progeny. To see if anything new was happening, the experimenters compared every DT shot with the results of a pure deuterium shot done under otherwise identical conditions. "We saw right away that the ion temperature rose from 27 keV [in the deuterium shots] to 35 keV with tritium, "says PPPL Deputy Director Dale Meade. That's much more of a temperature rise than one can simply attribute to alpha heating, Meade explained, because the lion's share of the alpha heating goes to the electrons in the plasma rather than to the ions. "So we seem to be seeing some sort of subtle enhancement of energy confinement by the tritium."
An important figure of merit is the energy confinement time t, which characterizes the rate at which heat leaks out of the plasmas. It was not known a priori, whether introducing tritium would have any effect, beneficial or deleterious, on t. "Our very first tritium shot showed a t of 190 milliseconds, where the corresponding t in pure deuterium had been 160," Meade told us. "And it's been 20 to 30% t enhancement ever since. Our next job is to sort out whether the improved confinement is due to the alphas or directly to the tritons." The detailed study of confinement and alpha heating in DT plasmas will take up much of the TFTR's remaining months. - by Bertram Schwarzchild