All that ever seems to separate fusion power from reality is 20 years and tens of billions of dollars.
But a new proposal from MIT fusion scientists for a working fusion power plant may change at least one part of that equation. While it isn’t likely to shorten the timeframe, it may lower the cost significantly. And even if it doesn’t, it has a design trick up it’s sleeve that could help scientists finally crack the fusion reactor riddle.
The so-called ARC reactor—a thinly veiled nod to the diminutive device built by comic book character and fictional MIT alumnus Tony Stark (aka Iron Man)—stands for “affordable, robust, and compact.” It would be significantly more modular than ITER, the prototype fusion reactor currently under construction in France. By allowing researchers to swap parts, the ARC reactor would make it easier to adjust the design for optimal performance.
The scientists behind ARC made a few other clever design choices. The tokamak—the donut-shaped vessel that contains the plasma—will be surrounded by molten fluorine-lithium-beryllium salt that addresses several requirements at once. By absorbing neutrons, it moderates the nuclear reaction, shields external components from damage, and breeds tritium, which can then be used to fuel the reactor. The molten salt can also be circulated through a heat exchanger, which is how the plant will convert all that heat to electricity. And, because it’s a liquid, the salt won’t suffer the same kind of damage that befalls solid shielding materials.
Another innovation is their choice of superconductors: REBCO, a rare-earth barium-copper-oxide tape that superconducts at temperatures as high as to 80 K while still producing extraordinarily strong magnetic fields. That should keep a lid on the power required to run the reactor, keeping the plant’s parasitic consumption of the projected 270 MW of electricity to a minimum.
But as John Timmer points out over at Ars Technica, the killer feature for the reactor is, again, its modularity:
The authors calculate that interdigited REBCO junctions can maintain the required superconductivity. That, in turn, means that the magnet assembly doesn’t have to be produced as a single unit. Instead, they propose making a top and bottom half that come apart a bit like the upper and lower halves of a clam shell. Each of the pieces inside the shell—the shielding, the molten salt, and the reactor walls—can then be accessed, modified, or even replaced.
So, if there were a complete failure of containment and the reactor walls were trashed, then they could simply (for some definitions of “simply”) be replaced. Or, if used in a research reactor, different designs could be tested. As the authors phrase it, “a starting design philosophy of ARC is that failure should and will occur as various fusion materials and power exhaust technologies are tried and tested.”
The ARC team projects that the plant would cost about $5.5 billion, some $4.6 billion of which would go toward the the structure that would support the novel magnets. That’s a fraction of the more than $20 billion that ITER is expected to cost, though I should point out that ITER, too, was expected to cost about $5 billion when it was first proposed.
Image credit: ARC team/MIT