Fusion and what “net energy gain” really meant

Fusion is the opposite of the reaction in today’s nuclear plants. Those split heavy atoms apart (fission); fusion forces light ones — forms of hydrogen — together, the reaction that powers the sun. It releases enormous energy and leaves no long-lived radioactive waste and no risk of meltdown. The catch has always been that making it happen on Earth costs more energy than it gives back.

So when a US lab announced in 2022 that a fusion experiment had produced more energy than was put in — “ignition,” net energy gain — it was reported as the moment fusion arrived. It was a genuine scientific milestone. It was also narrower than the headlines, and the narrowing is the whole story.

The experiment fired 192 lasers at a fuel pellet the size of a peppercorn. The fuel gave back a bit more energy than the laser light that landed on it. That is a real threshold — for the first time the reaction was self-sustaining enough to clear the energy that reached the fuel.

But firing those lasers drew far more electricity from the wall than ever reached the pellet — the lasers are wildly inefficient. Count that, and the machine consumed something like a hundred times more energy than the fusion produced. “Net gain” was true for the energy delivered to the fuel, and badly false for the energy delivered to the building.

This is the distinction worth keeping: scientific gain (did the fuel return more than the energy that hit it?) versus engineering gain (did the whole plant return more than it drew from the grid?). 2022 crossed the first. A power station needs the second, with margin to spare, running not once but continuously.

There are two main routes still being raced — crushing fuel with lasers (inertial, the approach here) and holding a superheated plasma in magnetic fields (magnetic, the doughnut-shaped tokamaks). Both have made real progress. Neither is close to putting steady power on the grid; that remains a matter of decades, not years.

The milestone deserved the celebration and not the conclusion. Fusion released more than it absorbed — at the fuel, once, briefly. Getting from there to a plant that beats its own wall socket, around the clock, is a different and much larger problem. When the next fusion headline lands, the one question that cuts through it is: net gain measured against what — the fuel, or the whole machine?