What is Nuclear Fusion? And How Does it Work?

Currently, nuclear power plants produce energy by exploiting the fission process. Nuclear fission involves taking a heavy atom such as uranium isotope U-235, and splitting it into two lighter atoms. This process releases a tremendous amount of useable energy. 

Fission is initiated by bombarding U-235 with neutrons. As these neutrons collide with U-235 atoms, they force additional neutrons to be released which then collide with other U-235 atoms creating a chain reaction. This reaction can also produce electromagnetic radiation. To maximize energy production, the number of neutrons used in fission must be controlled to prevent the chain reaction from dying or running away. 

Fission reactors exploit produced energy the same way as conventional power plants. 
Fission occurs in a reactor vessel, moderated by control rods. The heat from the reaction is used to heat water and produce steam. This steam then turns a turbine to produce electricity. Additional steps are required to safely dispose of spent fuel as they are radioactive and can have a half-life of over 10,000 years. In many cases, decommissioning costs are up to 15% of the initial capital cost.

To create fuel for nuclear fission, enrichment is first required to increase uranium-235 concentration from the 0.7% abundance found naturally in uranium ore to the 4-5% needed for the fission process. This is done using 1,000-2,000 centrifuge stages. To do this, uranium ore is converted to UF6 gas, processed, and then converted into fuel pellets after the enrichment process. The centrifugal process is carefully controlled under specific vacuum parameters.

As of February 2026, 416 reactors with a net capacity of ~400 GWe were operational, operating in 31 countries. These reactors produce roughly 9% of the world’s electricity. Additionally, 70 nuclear power reactors are under construction. Uranium ore is sourced in 10 countries and has an energy density of 70,000 times that of fossil fuels. The cost of construction of a nuclear power plant can be between $2 and $20 billion, though the development of lower-cost, smaller units is feasible.

Fusion reactions of light elements power the stars and produce many elements in a process called nucleosynthesis.

Nuclear fusion is a developing technology that sits at the forefront of current scientific research. With the potential to revolutionize energy production, fusion is the focus of a significant number of national and international projects. 

In the nuclear fusion process, two light atoms are ‘fused’ together to create a larger atom. In this way, nuclear fusion is the opposite of fission: instead of splitting atoms, atoms combine to form a larger one. This results in the release of energy. 

Fusion combines hydrogen isotopes such as deuterium (D) and tritium (T). Deuterium, also called heavy hydrogen, is similar to a ‘regular’ hydrogen isotope like protium, but deuterium contains an additional neutron. For comparison, tritium is a hydrogen isotope containing two extra neutrons. 

Ideal fusion reactions use a 50:50 mix of deuterium and tritium with only a few grams needed at any time. However, this ideal has been difficult to achieve.

Light atoms can only fuse if they are ‘squeezed‘ together for a long enough time. The attractive nuclear force is stronger than the repulsive electrostatic force but only at very short distances. Based on these constraints, the main technical difficulty for fusion is getting the nuclei close enough to allow ‘fusion’. This is why a form of either magnetic or inertial confinement (or a combination of the two) is needed.  

Enormous temperatures are needed for both types of confinement: over 4 x 10⁷ K for deuterium–tritium (DT) fusion. That’s 10 x hotter than at the Sun’s core! 

Many different approaches that have been proposed to achieve fusion. However, there are two main kinds of confinement being explored.

Magnetic confinement fusion (MCF) is arguably the most encouraging route to commercial energy generation. It uses a series of very powerful superconducting magnets, with liquid helium used to cool the magnets and optimise device performance.

The most common magnetic confinement devices are the Tokamak and Stellarator, though some facilities have experimented with devices that use magnetic mirrors. All three of these device types consist of a central vacuum vessel, where the plasma is contained and fusion reaction occurs. The vacuum vessel is surrounded by a cryostat, where the superconducting magnets are housed and cooled. The Tokamak has a toroidal shape, the Stellarator resembles a twisted ribbon,  whereas the mirror machine has a cylindrical shape. The magnetic field continually forces the atoms together and the high temperatures enables ‘plasma ignition’.  

The second common confinement type is inertial confinement fusion (ICF). This technique extracts fusion energy by bombarding small pellets (1 to 3 mm) of DT fuel with lasers or charged particle beams to compress and heat the mixture to ignition. Here, ignition refers to the condition that all alpha particles are stopped, which causes the plasma to heat.

The compression and heating is done very rapidly with short pulse (1 to 10 ns) high power lasers, X-rays or ion beams which intersect the fuel from many directions. Direct-drive devices deliver compressive energy pulses to the fuel pellet directly. Indirect-drive ICF techniques have the fuel pellet placed inside a small metal vessel (hohlraum/burning chamber). Lasers are then focused on the hohlraum, heating it into a plasma which radiates X-rays that, in turn, are absorbed by the pellet. The pellet then compresses in the same way as ‘direct-drive’. Compared with direct-drive devices, the indirect-drive approach is significantly less efficient. The indirect-drive technique is used in nuclear weapon simulations. 

To create electricity from ICF, fuel pellets are dropped into a vacuum reaction chamber where they are ignited. The radiation from the plasma and fusion neutrons is absorbed by liquid lithium. Heat produced is used in conventional steam generation and tritium is produced to be recycled as new fuel.

Fusion has the potential to produce massive amounts of energy from only a very small amount of fuel — only around 250kg a year in a commercial reactor — once sustained ignition is achieved. 

The main advantages of fusion are the abundance of fuel compared with hydrocarbon sources, the lack of harmful by-products such as greenhouse gases, and its intrinsic safety compared with nuclear fission. 

Compared with fission, nuclear fusion has additional advantages: 

• While the fusion reaction releases neutrons that are dangerous to humans, these neutrons are contained and the production of neutrons ceases almost as soon as the plant is turned off.  
  
• Fusion is inherently safer as there are no chain reactions, run-aways, or melt-downs.
  
• Fusion results in extreme temperatures but if any plasma were to escape, then the heat would quickly dissipate without causing harm.
  
• Neutrons can present a hazard but are easily stopped.
  
• While nuclear fusion does produce radioactive waste, this radioactivity is short-lived — especially when compared with fission-produced waste. This includes the neutron bombardment vessel.  
  
• Waste produced through fusion does not need to be transported for disposal, storage or reprocessing.

The vacuum requirement in fusion reactors depends on the type of confinement and the size of the reactor.

The largest need for vacuum in magnetic confinement is in the central vacuum chamber and surrounding cryostat chamber. This can be over 100 m³ which must be under UHV (around 10^-9 mbar) to prevent contamination of the plasma. 

Vacuum is also needed for the production of tritium, pumping helium and fuel recirculation. In the case of inertial confinement fusion, all the laser or particle beam lines must be evacuated along with the reaction chamber.

As with magnetic confinement, this can be a very large volume requiring UHV to prevent losses and aberration in the beams. Turbomolecular and cryogenic pumps are the most commonly used and they must be able to cope with the localized peak in radiation and high magnetic fields. Similar concerns apply to primary pumps sets, gauges and valves.

One of the key requirements for vacuum pumps is the ability to handle tritium. This means that metal seals are preferred and pumps should be fluorine/halogen and hydrocarbon free. 

In the case of magnetic confinement, the pumps must also be able to operate under high magnetic fields and shielding is often required.  Space around the reaction chambers is very limited and shutdowns are extremely expensive, so pumps must have a small footprint and be highly reliable with minimal service requirements. 

There is an ongoing debate over when net energy gain from fusion will be achieved, however optimistic estimates now project <5 years. 

Supporting this optimism is the fact that fusion projects around the world have been breaking records including at the JET, NIF and EAST facilities. 

The private fusion industry is growing at a rapid pace with $13B of declared investments to 2026. 

ITER, a tokamak-style reactor under construction is the largest (multi) government funded fusion project in the world. It aims to produce 10 x more energy than it uses. One of ITER’s goals is to lead the way for DEMO, a prototype for commercial fusion reactors.  

Across 18 countries, there are over ~100 fusion devices either running or being built including privately funded devices. 

Currently, no fusion facility produces a net energy gain. But as the energy density of fusion fuel is even higher than fission, and around 10,000,000 times that of fossil fuels, the technology promises to reshape the future of energy production.