Iterative processes

HERE is one way to squeeze energy from nuclear fusion: create and contain a roiling soup of ionised hydrogen atoms known as a plasma, and heat it to ten times the temperature of the sun’s core. Some of the fast-moving atomic nuclei will bash together with enough oomph to fuse. Gather the energy from fast-moving particles created in these collisions and you have a limitless (for hydrogen is abundant), comparatively clean energy source. It is an idea conceived in the 1950s, but yet to be born in a laboratory.

Here is one way that might make it happen: gather an international consortium of the fusion-minded, including the European Union, America, China, India, Japan, Russia and South Korea. Conspire to build a 23,000-tonne doughnut-shaped vessel called a tokamak, that is wrapped with 80,000km of superconducting wire, all to contain the plasma magnetically and, for the first time, produce fusion energy continuously. Call it the International Thermonuclear Experimental Reactor; shorten the name to ITER for better PR. And farm out the design to the seven “domestic agency” partners, each completely in charge of the procurement and production of their bit (they will all have to agree to any changes, though, as the design of this technological beast inevitably evolves).

It sounds wonderfully egalitarian, in a technocratic sort of way. Timelines are slippery things, though, so delays will occur and costs will go up. That makes for more delays. Partners may drop out and come back as the political will to pay for the project comes and goes.

The story of ITER has become a tale of these shortcomings. The first of its components arrived at the reactor’s site in Cadarache, in the south of France, earlier this month, just as the foundations were finished. In the next week or so construction should start on the walls that will house its core: the doughnut-shaped vacuum vessel. Perhaps tellingly, no one can say exactly when even that will happen. There has already been a 30-month delay in the manufacture of the vacuum vessel. The most recently published schedule says the first plasma will be created in the vacuum vessel in 2020. That will now have to slip to 2023 or 2024, but the revised official schedule will not be published until mid-2015. The overall cost? Also unknown, but it is sure to surpass by a considerable sum the current official estimate of $20 billion.

The factors that have hobbled ITER from the start have not been concerns of nuclear physics or large-scale engineering. They have been problems of leadership and project management. Few would now argue that the initial design was adequate, or that the seven domestic agencies should have been allowed to have the absolute control they got over the bits of it they worked on. As delays and dissent cropped up within the domestic agencies, ITER’s management kept the other agencies in the dark, and stuck too long to timelines that were never feasible.

Most embarrassingly, a management assessment from last October that was leaked earlier this year derided the organisation along all these lines, and added that the project lacked a “nuclear safety culture”. America withheld 12% of its ITER funding this year, pending the implementation of every one of the assessment’s recommendations. A dissenting subcommittee of the Senate has proposed a budget that would see America pull out altogether next year, as it did in 1999 (it came back into the fold in 2003).

ITER’s core philosophy is to share between countries the risks, efforts and rewards of trying to crack the fusion-power problem—costs and delays be damned. The idea is that if the project proves successful, any of the domestic agencies involved should then be able to build its own version with the knowledge collaboratively gained.

Consensus among the ITER faithful is that it will accomplish its stated goal of extracting 500 megawatts of power from a continuously fusing plasma—about ten times as much power as is put in. Scientifically, nothing seems to stand in the way of this. But consider the National Ignition Facility, a star-crossed American effort to use lasers rather than magnetic fields to create fusion. It looked certain, on paper, that if engineers built a laser of a given size (the world’s largest, by some margin) and fired it at a target of a prescribed shape and composition, the result would be a net gain of energy. The laser and the targets were made; fusion “break-even”, however, was not. The facility has now switched its focus to “nuclear-stockpile stewardship” (modelling the behaviour of atom bombs).

What looks good according to the equations is thus not always—or even often—what works best in reality. But if ITER wants even the chance to test the science, it will first have to solve, in a comprehensive fashion, its human problems.