Fusion nuclear reactors are best described as machines that generate energy by forcing light atomic nuclei together, the opposite of conventional nuclear power plants, which split heavy atoms apart. Instead of uranium, fusion reactors use forms of hydrogen as fuel. When two hydrogen atoms fuse, they create helium and release enormous energy, the same process that powers the sun. No fusion reactor currently produces electricity for the grid, but multiple experimental machines have demonstrated the core physics, and the first commercial-scale projects are targeting operations within the next two decades.
How Fusion Differs From Fission
Every nuclear power plant operating today runs on fission: a neutron strikes a large atom like uranium, splitting it into smaller atoms and releasing energy. Fusion works in reverse. Two small atoms, typically isotopes of hydrogen called deuterium and tritium, collide at extreme speeds and merge into a single heavier atom (helium), releasing an energetic neutron in the process. That neutron carries most of the energy, and capturing it as heat is what would eventually generate electricity.
The distinction matters for safety. A fission reactor contains a large inventory of fuel that must be actively controlled to prevent overheating. A fusion reactor holds less than four grams of fuel at any moment, and the reaction depends on a continuous feed of fresh material. If anything disrupts the process, the reaction stops on its own. A runaway chain reaction or meltdown of the kind that occurred at Fukushima is physically impossible in a fusion system.
Why Fusion Needs Extreme Temperatures
Atomic nuclei are positively charged, so they naturally repel each other. To force them close enough to fuse, the fuel must be heated to roughly 100 million degrees Celsius, about six times hotter than the core of the sun. At these temperatures, the fuel becomes plasma, a state of matter where electrons separate from their nuclei entirely. No solid material can contain plasma this hot, so engineers have developed two fundamentally different strategies to hold it in place.
Magnetic Confinement: Tokamaks and Stellarators
The most widely pursued approach uses powerful magnets to suspend the plasma inside a vacuum chamber, never letting it touch the walls. The leading design is the tokamak, a doughnut-shaped machine (physicists call the shape a torus) that combines multiple magnetic fields to keep the plasma stable. One set of coils wraps around the doughnut the long way, creating a strong field in the toroidal direction. A central magnet generates a second field the short way around. Together, these produce a twisted magnetic cage that traps charged particles along spiraling paths. A third set of coils on the outside fine-tunes the plasma’s shape and position.
A close relative of the tokamak is the stellarator. Where a tokamak relies on an electrical current running through the plasma itself to help create its magnetic cage, a stellarator generates the entire confining field with external coils alone. The coils are elaborately twisted, giving the machine an organic, almost sculptural appearance. This design trades simplicity for stability: because there is no internal plasma current, stellarators avoid a whole category of instabilities that tokamaks must actively suppress. In a tokamak, pressure-driven disturbances called tearing modes can grow and disrupt the plasma. In a stellarator, the same physics actually causes those disturbances to shrink, making the plasma self-healing in a sense. The tradeoff is that stellarators are harder to engineer and historically have not confined energy quite as well, though modern designs are closing that gap.
Inertial Confinement: Lasers Instead of Magnets
The second approach skips magnets entirely. At the National Ignition Facility (NIF) in California, 192 of the world’s most powerful lasers converge on a target the size of a peppercorn filled with deuterium and tritium. The lasers heat a small metal cylinder surrounding the fuel capsule to over 3 million degrees Celsius, generating intense X-rays that blast the capsule’s outer shell outward. By Newton’s third law, the fuel implodes inward at extraordinary speed, compressing and heating it until fusion ignites in a tiny hot spot at the center.
If the implosion is symmetric enough, the helium nuclei produced in that hot spot spread outward and heat the surrounding fuel, triggering a self-sustaining burn. This is what physicists call ignition, and NIF achieved it for the first time in December 2022, producing 3.15 megajoules of fusion energy from 2.05 megajoules of laser energy delivered to the target. The facility has repeated the result several times since, with its best shot in July 2023 yielding 3.88 megajoules. These experiments are landmark proof that controlled fusion energy gain is possible, though inertial confinement is generally considered farther from a commercial power plant design than magnetic confinement.
Fuel Supply and Availability
Deuterium is abundant. It occurs naturally in seawater at a ratio of about 1 in every 6,500 hydrogen atoms, meaning the oceans contain enough to power civilization for millions of years. Tritium is rarer because it is radioactive with a short half-life, so it barely exists in nature. Current plans call for fusion reactors to breed their own tritium by surrounding the plasma chamber with lithium blankets. When the high-energy neutrons from the fusion reaction strike lithium, they produce fresh tritium that can be fed back into the reactor. Lithium itself is relatively common in Earth’s crust.
Radioactive Waste Compared to Fission
Fusion does produce some radioactive waste, but it is a fundamentally different problem than fission waste. The high-energy neutrons that escape the plasma strike the reactor’s inner walls, gradually making the structural materials radioactive through a process called neutron activation. The total radioactivity can be comparable in magnitude to fission waste, but the isotopes involved are generally less hazardous and have much shorter half-lives. Most of this material would decay to safe levels within decades to a century, rather than the thousands of years required for spent fission fuel. The choice of structural materials matters: some alloys decay quickly, while others, like certain niobium-zirconium combinations, can require storage times closer to those of fission waste. Designing reactors with low-activation materials is an active area of engineering specifically to keep waste manageable.
Fusion produces no carbon dioxide during operation and requires no mining of uranium or plutonium. The primary fuel comes from water and lithium, and the main reaction product is ordinary helium.
Where Commercial Fusion Stands Today
The largest fusion project in the world is ITER, an international tokamak being built in southern France by a consortium of 35 nations. ITER is designed to produce 500 megawatts of fusion power from 50 megawatts of heating input, a tenfold energy gain that would demonstrate commercial viability in principle. The project has faced repeated delays and cost overruns. Current estimates place the start of full operations at 2039.
Meanwhile, dozens of private companies are pursuing smaller, faster paths to fusion power using a variety of approaches: compact tokamaks with high-temperature superconducting magnets, pulsed magnetic compression, and hybrid designs. Several have announced target dates for grid-connected electricity in the early 2030s, though none has yet demonstrated a net energy gain in a magnetic confinement device. The gap between producing fusion in a laboratory and running a power plant remains substantial. A commercial reactor needs to sustain the reaction continuously (or in rapid pulses), convert neutron energy into heat efficiently, breed its own tritium fuel, and do all of this affordably with components that can withstand years of neutron bombardment.
What is no longer in doubt is the underlying physics. Fusion works. The engineering challenge is making it work reliably, economically, and at scale.