Fusion Reactors: How Scientists Are Building a Star on Earth

Recreating Stellar Energy on Earth

The sun generates energy by fusing hydrogen nuclei at its core, where temperatures exceed 15 million degrees Celsius and gravitational pressure is enormous. On Earth, without the sun's gravity, the temperature required to initiate the same fusion reactions must be far higher — around 100 to 150 million degrees Celsius — to give nuclei enough kinetic energy to overcome their mutual electromagnetic repulsion and fuse. No material wall can contain something this hot. The engineering challenge of fusion is not physics but engineering: holding plasma at stellar temperatures long enough, densely enough, to extract more energy than was put in.

Fusion is attractive because its fuels are abundant, its reactions do not produce long-lived radioactive waste like fission, and individual accidents cannot cause runaway chain reactions — the plasma extinguishes within milliseconds if confinement fails. Deuterium, one fusion fuel, can be extracted from seawater; the global oceans contain enough deuterium to fuel humanity at current consumption levels for billions of years. The challenge is achieving net energy gain — consistently extracting more energy from fusion than is required to heat and confine the plasma.

The Deuterium-Tritium Reaction

The most accessible fusion reaction for near-term reactors combines deuterium (D, hydrogen-2) and tritium (T, hydrogen-3):

D + T → He-4 (3.5 MeV) + n (14.1 MeV)

The reaction releases 17.6 MeV total per fusion event. The neutron carries 80% of this energy (14.1 MeV), while the helium-4 nucleus (alpha particle) carries 20% (3.5 MeV). In a power reactor, the neutrons would be captured in a surrounding lithium blanket, transferring their energy as heat (converted to electricity by steam turbines) and generating tritium via the reaction Li-6 + n → He-4 + T. Tritium is radioactive with a 12.3-year half-life — which is why it cannot simply be stockpiled; it must be bred in the reactor itself from lithium.

Magnetic Confinement: Tokamaks

The dominant approach to fusion confinement uses magnetic fields to keep the hot plasma away from physical walls. The tokamak — a word coined from a Russian acronym for "toroidal chamber with magnetic coils" — is a donut-shaped chamber where two types of magnetic fields combine to confine plasma.

A powerful toroidal field, generated by superconducting coils wrapped around the torus, circulates along the donut's circumference. A poloidal field, generated partly by current driven through the plasma itself via a central solenoid (transformer action), circulates around the donut's cross-section. These two fields combine into a helical field that keeps plasma particles on stable orbits, preventing them from drifting to the walls.

The three conditions a fusion plasma must simultaneously satisfy for net energy gain are captured by the Lawson criterion:

• Temperature: ≥100 million °C for D-T fusion (ions must have sufficient kinetic energy to fuse)
• Density: High enough number of ions per unit volume to achieve significant fusion reaction rates
• Confinement time: Long enough that the energy produced by fusion exceeds energy losses from radiation and particle transport

The product n × τ × T (density × energy confinement time × temperature) must exceed a threshold — roughly 10²¹ keV·s/m³ for D-T. Modern tokamaks have approached this threshold from various directions. JET (Joint European Torus) set a record in February 2022, producing 59 megajoules of fusion energy sustained over 5 seconds — more than double its previous 1997 record — though it still required more external power than it produced.

ITER: The International Fusion Experiment

ITER (International Thermonuclear Experimental Reactor) under construction in Saint-Paul-lès-Durance, France, is the world's largest tokamak and a joint project of 35 nations representing more than half of the world's population. It is designed to demonstrate scientific breakeven and beyond.

• Plasma volume: 840 m³ (10 times JET's plasma volume)
• Fusion power: 500 MW from 50 MW of input heating power — a Q factor of 10 (Q = fusion power / heating power)
• Plasma current: 15 million amperes
• Central solenoid: The most powerful pulsed superconducting magnet ever built — 13 tesla, 1,000-tonne structure that generates a magnetic force equivalent to two aircraft carriers
• Superconducting coils: Niobium-tin (Nb₃Sn) cooled to 4 Kelvin by liquid helium
• First plasma target: 2025 (delayed from original 2020 schedule); deuterium-tritium experiments planned from the early 2030s

Inertial Confinement: The NIF Approach

The National Ignition Facility (NIF) at Lawrence Livermore National Laboratory takes a fundamentally different approach. Instead of magnetic confinement, NIF uses 192 laser beams delivering up to 2.15 megajoules of ultraviolet light, focused simultaneously on a pellet of frozen D-T fuel roughly 2 millimeters in diameter.

The laser energy ablates the outer shell of the pellet, creating a rocket-like implosion that compresses the D-T fuel to densities more than 100 times that of lead and temperatures exceeding 100 million °C for a few billionths of a second — long enough for thermonuclear ignition to occur. In December 2022, NIF achieved fusion ignition for the first time in a laboratory setting, producing 3.15 megajoules of fusion energy from 2.05 megajoules of laser energy delivered to the target — a Q > 1 on target. Total facility electrical input remains far larger, making net electricity generation still distant.

Private Fusion Ventures

Commonwealth Fusion Systems demonstrated in 2021 that its high-temperature superconducting magnet technology could achieve 20 tesla — more than twice the field strength of ITER's magnets — at far lower cost and complexity. Higher magnetic fields allow a smaller, cheaper tokamak to achieve the same plasma pressure. CFS's planned SPARC device, roughly 1/65 the volume of ITER, aims to demonstrate Q > 2 by the mid-2020s.

The Path to Commercial Fusion Power

The engineering gap between experimental fusion demonstrations and a commercial fusion power plant is enormous. A power plant must breed its own tritium, convert neutron energy to electricity at >30% efficiency, survive decades of neutron bombardment without material degradation, and operate with high availability. Materials science — developing structural materials that withstand 14 MeV neutrons without becoming severely embrittled or activated — is a critical unsolved engineering challenge. Fusion may provide energy before 2050; it will almost certainly be operating at commercial scale before 2100. The physics is solved. The engineering is the work of the next generation.