Purportedly, not a science project anymore – an actual plan with a deadline.
I’m putting this in the same category as a colony on Mars. Do-able, I guess, but you’ll have to show me.
That said, the video above is very well done and informative.
Commonwealth Fusion Systems is looking to join a power grid that is operated by PJM Interconnection and provides 182,000 megawatts of power to more than 67 million people living in 13 states and Washington, D.C. But technical hurdles to bringing fusion online remain—one major obstacle is actually producing a stable fusion reaction that generates more energy than it consumes.
The application process requires a potential energy provider to provide extensive technical information to the grid operator, including descriptions of the planned fuel type. In Commonwealth’s case, the company is developing a tokamak reactor design that uses powerful magnetic fields to create and insulate a highly energetic cloud of particles called a plasma until it’s hot enough for those particles to fuse. It’s a process that mimics the nuclear reactions in the sun, including the particles involved: isotopes of hydrogen called deuterium and tritium. The promise of the device is that a fusion reaction could feasibly generate limitless clean energy. That energy, in the form of heat, is used to boil water into steam, which then pushes a turbine to produce electricity.*
Much of that process remains theoretical, however, because physicists have yet to prove that fusion can work as a large-scale power source. Recent results from Germany’s Wendelstein 7-X demonstrated it could contain superheated plasma for 43 seconds. And its rival, the Joint European Torus, was apparently able to accomplish that feat for a full minute before its reactor was retired in 2023. While such capabilities are impressive, there is still a long way to go before a fusion device could be connected to a grid. Commonwealth plans to open its first power plant, called ARC (for “affordable, robust, compact”), in Virginia in the early 2030s. And the company aims to demonstrate an initial model, called SPARC (for “smallest possible ARC”), in 2027.
Commonwealth has demonstrated some success: The company’s toroidal field magnet technology was validated by the Department of Energy in September 2025. The superconducting magnets generate the magnetic field that is used to contain the high-temperature plasma generated by a fusion reaction. But Commonwealth has yet to test the full system.
Let us just work on the assumption that it will be an excellent addition to grids that do not ‘need’ it.
Would be vapourware?
This is like building the airport, and figuring out flight schedules, before the Wright Flyer has managed to get off the ground. Meanwhile China is currently building a dozen Chinese-flavoured AP1000 light water reactors, essentially the same as the two designed by Westinghouse and built in Georgia. China is building them in half the time, for one third the price. They have also developed a 1500 MW version, 35% more power on a smaller land footprint, that’s been running for a year, with a second plant nearly finished.
Nearly all the energy from fusion is in the form of very high energy neutrons, about 7x more per watt generated than from fission, and also about 7x more energetic. The energy from fission is mostly in the two fission fragments of uranium, which stay in the fuel rods, with the heat carried away by water (a good shield for neutrons and gamma rays.) A tokamak works in a vacuum – anything blasted off the chamber walls will contaminate the plasma, and lower its temperature from the 150 million degrees centigrade (ten times that of the sun’s core!) needed for commercial fusion. The neutron storm is also liable to be unhealthy for the supercritical magnets, which only work if they’re kept below twenty degrees above absolute zero.