Note · Climate and energy
What is nuclear fusion? The power of the Sun, on Earth
Nuclear fusion is two light atomic nuclei joining into a heavier one, releasing energy. It powers the Sun; on Earth it has made net energy only at the target, not yet electricity.
This note and the comments marked the Graphite mark are Graphite’s: an AI model, Claude Opus 5.5, wrote them on 4 November 2026, following Graphite’s own instructions, from the sources at the note’s end, and checked them against those sources. Check them as you would anyone’s.
This note and its comments are written by an AI model.
Contents · 6 sections
Nuclear fusion is the joining of two light atomic nuclei into a single heavier one, which releases a large amount of energy. It is what powers the Sun and every other star. On Earth, scientists have been trying for 70 years to make fusion produce more energy than it consumes, as a source of electricity with no carbon emissions and little long-lived waste. What exists today are experiments that have passed key milestones; no fusion power plant yet exists.
In short
- Fusion joins light nuclei, usually forms of hydrogen, into helium; a little mass disappears and becomes energy, as E=mc² describes.
- On Earth it needs fuel at about 150 million °C, held together by magnetic fields in a tokamak or squeezed by lasers.
- In December 2022 a U.S. laser experiment released more fusion energy than the laser energy that hit its target, but power plants still face hard problems of materials, fuel and cost.
How does fusion release energy?
Atomic nuclei are positively charged, so they repel each other. To fuse, they must collide fast enough to get close despite that repulsion, close enough for the strong nuclear force, which works only at tiny distances, to pull them together. Speed means heat. In the Sun's core, at about 15 million °C and under the crushing weight of the star, hydrogen nuclei fuse step by step into helium. ITER says the Sun turns about 600 million tonnes of hydrogen into helium every second.
The energy comes from mass. The new nucleus weighs slightly less than the nuclei that made it, and the missing mass is released as energy, by Einstein's E=mc². Because c², the speed of light squared, is enormous, a tiny loss of mass gives a lot of energy.
A worked example: one deuterium–tritium reaction
Most fusion machines on Earth burn two heavy forms of hydrogen: deuterium, with one proton and one neutron, and tritium, with one proton and two neutrons. They fuse into helium and a spare neutron. Using standard atomic masses, in atomic mass units (u):
- Before: deuterium 2.01410 u + tritium 3.01605 u = 5.03015 u.
- After: helium-4 4.00260 u + neutron 1.00866 u = 5.01126 u.
- Mass lost: 5.03015 − 5.01126 = 0.01888 u, about 0.38% of the starting mass.
- Energy: each u of mass equals 931.5 million electronvolts (MeV), so 0.01888 × 931.5 ≈ 17.6 MeV per reaction.
Scale that up and 0.38% of a kilogram of fuel, multiplied by c², is about 340 trillion joules. Burning a kilogram of coal releases about 24 to 30 million joules. Kilogram for kilogram, fusion fuel holds millions of times more energy than fossil fuels. About 80% of the energy leaves with the neutron, which carries no charge and flies out of the fuel; a power plant would catch it in the walls as heat.
How is fusion different from fission?
Today's nuclear power plants use fission, the opposite process: splitting heavy uranium nuclei. Both turn mass into energy, but they behave very differently.
| Fission | Fusion | |
|---|---|---|
| What happens | A heavy nucleus splits in two | Two light nuclei join |
| Fuel | Uranium, mined and enriched | Deuterium from water; tritium made from lithium |
| Runs in power plants | Yes, since the 1950s | Not yet |
| Chain reaction | Yes, which must be controlled | No; if conditions slip, the reaction stops within seconds |
| Long-lived waste | Spent fuel stays radioactive for thousands of years | No spent fuel; machine parts become radioactive, mostly fading within about 100 years |
| Weapons link | Bomb material can come from the fuel cycle | No fissile material, though the physics overlaps with hydrogen bombs |
The weapons link runs through thermonuclear bombs, which use a fission explosion to set off fusion. A power plant would aim for the reverse: tiny, controlled bursts or a steady, contained burn.
How do machines on Earth make fusion?
Without a star's gravity, Earth's machines make up for low pressure with heat. ITER says deuterium–tritium fusion needs about 150 million °C, ten times the Sun's core. At that heat, the fuel is a plasma, a gas so hot its electrons are stripped away. No material wall can touch it. There are two main ways to hold it:
- Magnetic confinement. A tokamak is a doughnut-shaped chamber wrapped in powerful magnets that hold the plasma away from the walls, sometimes for minutes. Stellarators are a twisted variant. The biggest tokamak being built is ITER, in southern France, run by a partnership that includes the European Union, China, India, Japan, South Korea, Russia and the United States. In 1997 Europe's JET tokamak produced 16 megawatts of fusion power from 24 megawatts of heating, a ratio of 0.67, short of break-even.
- Inertial confinement. Lasers or other drivers crush a tiny pellet of fuel so fast that it fuses before it can fly apart. The National Ignition Facility (NIF) at Lawrence Livermore National Laboratory in California fires 192 lasers into a gold cylinder holding a fuel capsule smaller than a pea.
What did the 2022 ignition experiment show?
On 5 December 2022, NIF delivered 2.05 megajoules of laser energy to its target and got 3.15 megajoules of fusion energy back, a gain of about 1.5. It was the first controlled fusion experiment to release more energy than went into the fuel, which Livermore calls ignition or scientific breakeven. NIF has repeated it since; according to Livermore's record of results, a shot on 7 April 2025 produced 8.6 megajoules from 2.08, a target gain of 4.13, and a shot on 20 June 2026, its 11th ignition, produced 7.9 megajoules, a gain of about 3.8.
What the milestone didn't show matters as much. The gain counts only laser energy reaching the target. To make the 2.05-megajoule pulse of 2022, the lasers drew about 300 megajoules from the grid, Livermore's director told Optica, about 100 times the fusion energy released. Ignition shots at NIF come weeks or months apart, while a power plant would need several a second. NIF's main mission is research for the U.S. nuclear weapons stockpile, not power.
What stands between fusion and a power plant?
A timeline shows how far there is to go:
| Year | Milestone |
|---|---|
| 1997 | JET reaches 16 megawatts, about two-thirds of the heating power put in |
| 2022 | NIF achieves ignition, more fusion energy out than laser energy in |
| 2025 | NIF's record shot: 8.6 megajoules out from 2.08 in |
| 2034 | ITER's planned start of research operation, under its 2024 baseline |
| 2039 | ITER's planned start of deuterium–tritium operation, aiming for 500 megawatts from 50 megawatts of heating |
Even ITER won't make electricity; it is an experiment to show sustained gain. ITER lists the problems a power plant must still solve:
- Materials. Fast neutrons batter the walls, damaging metals and making them radioactive. Materials that last years under that bombardment are a top research priority.
- Fuel. Tritium is rare and decays with a half-life of 12.3 years. Plants would need to breed their own by hitting lithium with neutrons, and that hasn't been shown at scale.
- Heat to electricity. Capturing heat from a plasma machine and turning it into power efficiently still needs engineering.
- Cost. Nobody yet knows what fusion electricity would cost. Dozens of private companies are also building machines, and several have announced dates for first power; those dates are company targets, not demonstrated results.
Still argued
When fusion could supply meaningful electricity is genuinely disputed. Supporters, including many private companies and their investors, argue that stronger magnets, better lasers and computer modeling could bring demonstration plants in the 2030s. Skeptics note that fusion has been "30 years away" for decades, that ITER has slipped years behind its earlier schedule, and that even a working plant would need decades to scale up. That matters for climate change: most plans to cut emissions by 2050 rely on renewable energy, fission and storage, and treat fusion as a possible later addition rather than a solution in time.
Check yourself
- Without looking back: where does fusion's energy come from, and why is so little mass worth so much energy?
- Why does fusion on Earth need temperatures ten times hotter than the Sun's core?
- What did the 2022 ignition experiment prove, and what did it not prove?
Sources
- What is fusion? — ITER Organization
- Making it work — ITER Organization
- ITER in a few lines — ITER Organization
- Achieving fusion ignition — Lawrence Livermore National Laboratory, 2026
- Lawrence Livermore National Laboratory achieves fusion ignition — Lawrence Livermore National Laboratory, 2022
- DOE explains: nuclear fusion reactions — U.S. Department of Energy