Scientists have resolved a technical problem with the giant superconducting magnets of the ITER fusion reactor that threatened the project's scheduled completion in 2020.
After tests showed that the approved design for superconducting cables was showing signs of degradation too soon, samples from a U.S. manufacturer with a different design fared better. But Japan, the ITER partner responsible for manufacturing the cable, went its own way and has developed a cheaper alternative that seems to tick all the boxes. "It's totally stable. I'm convinced the problem is fully solved," says ITER technical director Rem Haange.
ITER, an international project under construction in France, aims to show it is possible to generate power by fusing hydrogen isotopes together, as happens in the sun and stars. The hydrogen fuel, in the form of a plasma, must be heated to about 150 million°C. Controlling it in that state requires huge and powerful superconducting electromagnets.
The conductor that has been causing problems is for the central solenoid, a 13.5-meter-high stack of six coils in the center of the reactor. The central solenoid acts like the primary of a giant transformer, creating a magnetic field of 13 tesla which induces a 15-million-amp current of plasma around the doughnut-shaped reactor, known as a tokamak. To produce such a field, the solenoid needs 43 kilometers of superconducting cable, made of a compound of niobium and tin (Nb3Sn).
Manufacturing the brittle compound is complex. The niobium and tin must be wound together in separate filaments, and once the coil is wound into its final shape it is heated to get the niobium and tin to react into the superconducting compound. Copper is also included as a safety measure in case the Nb3Sn suddenly loses its superconducting properties and the current needs somewhere to flow through.
Cable samples manufactured in Japan according to the ITER design were tested in late 2010 at the SULTAN facility at the Paul Scherrer Institute in Villigen, Switzerland. The ITER magnets must be able to endure 60,000 up-down current cycles during the reactor's lifetime, but the sample cables began to degrade after 6000 cycles. The problem seemed to be with the high mechanical loads put on the individual strands of the cables. Alternative cables from the U.S. supplier were tested in late 2011 for 10,000 cycles and showed much lower levels of degradation.
This sample was manufactured using a different method, known as "internal tin," whereas the Japanese had used the "bronze" process. Japanese suppliers had no experience with the internal tin process, so it was decided to stick with the bronze process and make other improvements.
Last November and December two Japanese suppliers put samples through their paces at SULTAN. The key innovation was a "short twist pitch"—the filaments in the cable formed a tighter spiral. Tests showed that these cables showed little signs of degradation. "The twist pitch made the difference," Haange says. "It's much more mechanically stable."
Haange doesn't think the new approach will delay the project because the United States—which is responsible for winding the cable into coils—had enough leeway built into its schedule. It will, however, cost Japan a few million additional euros, Haange says, because the tighter spiral uses more cable material.