Kyoto Fusioneering wins US backing for the part of fusion nobody photographs

Fusion coverage tends to stop at the reactor — the chamber where atoms fuse and the energy comes from. A working plant needs a great deal more than that, and the reactor alone is expensive enough that many startups are content to buy the remainder from somebody else.

That is the market Kyoto Fusioneering occupies, and it has become one of the larger specialists in it. The Japanese company holds $121 million in committed capital according to FusionX, and has now picked up grants from the US Department of Energy and from Tennessee to build, at Oak Ridge National Laboratory, a prototype device that breeds fuel. It told TechCrunch it is moving its American headquarters to Oak Ridge at the same time.

The demand behind that is documented rather than assumed. A recent Fusion Industry Association survey found more than half of fusion startups intend to source fuel cycle technology from outside suppliers rather than develop it in house.

The device is called Unity-3, is being built with ORNL, and exists to test a breeding blanket — the component asked to do two jobs at once. It captures energy coming off the reaction, and manufactures fresh fuel while doing so.

One approach uses liquid lithium, which soaks up heat and neutrons together. Neutrons striking the metal split it into helium and tritium, a hydrogen isotope that many reactor designs run on. The tritium is drawn off and returned to the reactor; the heat is taken away to make electricity.

Described that way it sounds tidy. In practice it demands materials that survive the temperatures involved, pumps built for the purpose, and a quantity of equipment nobody sells off a shelf — and the blanket is one piece of a plant with several other unproven ones.

Kyoto Fusioneering is working on some of those as well: heating fuel until it becomes a plasma, recovering unburned fuel from the exhaust, and turning captured heat into power.

Unity-3 will not test lithium alone. The company says it will run a range of blanket materials through it, which matters because almost everything the industry believes about those materials so far comes from simulation rather than hardware. Realta Fusion, Thea Energy and Type One Energy are all expected to design against the results, as is Xcimer Energy.

That spread is the commercial point. Supplying several competing routes to fusion at once is how a company ends up being the one everybody calls, whichever route turns out to reach the grid first.