A breeder reactor is a type of nuclear reactor that creates more fuel than it uses. These reactors can use common types of uranium and thorium, such as uranium-238 and thorium-232, instead of the less common uranium-235, which is used in traditional reactors. These materials are called fertile materials because breeder reactors can change them into usable fuel.

Breeder reactors work this way because they use neutrons very efficiently. Extra neutrons from the reactor are absorbed by the fertile materials placed inside the reactor with the fuel. These materials then change into new fuel that can be used in nuclear reactions.

At first, breeder reactors were seen as useful because they used uranium more completely than other types of reactors. However, interest in them decreased after the 1960s when more uranium supplies were discovered and new methods made fuel cheaper.

Breeder reactors have been built and tested in Russia, India, Japan, the United States, France, and China. As of April 2026, only Russia is currently running a commercial breeder reactor.

Types

Many types of breeder reactors are possible. A "breeder" is a nuclear reactor designed to use neutrons very efficiently, with a conversion rate higher than 1.0. This means it creates more fissile material (like plutonium or uranium) than it uses. In theory, almost any reactor design can be modified to become a breeder. For example, the light-water reactor, which uses water to slow down neutrons, evolved into the RMWR concept. This design uses light water in a special form called supercritical to improve neutron efficiency enough for breeding.

Besides water-cooled reactors, other types of breeder reactors are being considered. These include reactors cooled by molten salt, gas, or liquid metal. These designs can use different fuels, such as uranium, plutonium, minor actinides, or thorium. They can be built for different purposes, like producing more fissile fuel, operating continuously, or reducing nuclear waste.

Existing reactor designs are often divided into two categories based on their neutron spectrum. One category uses fast neutrons (not slowed down) and is called fast breeder reactors (FBRs). These reactors breed fissile plutonium from fertile uranium-238. They can also breed fissile uranium-233 from thorium if needed. The other category uses thermal neutrons (slowed down) and is called thermal breeder reactors. These reactors breed fissile uranium-233 from thorium. Thermal breeders are considered more commercially feasible with thorium fuel because they avoid creating heavier transuranics.

As of 2026, all large-scale FBR power stations use sodium-cooled liquid metal fast breeder reactors (LMFBR). These reactors come in two main designs:
– Loop type: The primary coolant is circulated through heat exchangers outside the reactor tank but inside the biological shield.
– Pool type: The heat exchangers and pumps are immersed directly in the reactor tank.

As of 2017, only two breeder reactors were commercially operating: the BN-600 (560 MWe) and the BN-800 (880 MWe), both Russian sodium-cooled reactors. These reactors use liquid metal as the primary coolant to transfer heat from the core to steam that powers electricity generators. Other FBRs have used coolants like mercury, lead, lead-bismuth eutectic, molten tin, or sodium-potassium alloy (NaK). Mercury and NaK are liquids at room temperature, which is useful for experiments but less practical for large-scale reactors due to toxicity and cost. Lead-cooled reactors were used in the Soviet Union, including in submarines and the BREST reactor.

Three of the proposed Generation IV reactor types are FBRs:
– Gas-cooled fast reactor cooled by helium.
– Sodium-cooled fast reactor based on LMFBR and integral fast reactor designs.
– Lead-cooled fast reactor inspired by Soviet naval propulsion units.

FBRs typically use a mixed oxide fuel core with up to 20% plutonium dioxide (PuO₂) and at least 80% uranium dioxide (UO₂). Another option is metal alloys of uranium, plutonium, and zirconium. Enriched uranium can also be used alone.

Many FBR designs surround the reactor core with a blanket of tubes containing non-fissile uranium-238. This uranium captures fast neutrons from the core, converting into fissile plutonium-239, which is later reprocessed as fuel. Other designs rely on fuel geometry to achieve neutron capture. In fast reactors, the fissile cross-section of plutonium-239 or uranium-235 is smaller than in thermal reactors, requiring higher concentrations of fissile material to sustain a chain reaction. Fast reactors do not need a moderator to slow neutrons, which allows more neutrons per fission. Liquid water, which acts as a moderator and neutron absorber, is not ideal for fast reactors because it reduces neutron yield and breeding efficiency. Some designs, like the supercritical water reactor (SCWR), use supercritical water to cool the core with less water, enabling a fast-spectrum water-cooled reactor.

The extreme conditions of fast reactors—high temperatures, radiation, and coolant interactions—challenge fuel cladding materials. Austenitic stainless or ferritic-martensitic steels are commonly used, but oxide dispersion-strengthened alloy steel is seen as a long-term solution for radiation resistance.

One specific fast reactor design, the integral fast reactor (IFR), was created to address waste disposal and plutonium issues. The IFR used on-site fuel reprocessing to recycle uranium and transuranics (not just plutonium) via electroplating, leaving only short-half-life fission products in waste. These fission products could be used for industrial or medical purposes, with the rest sent to a waste repository. The IFR’s pyroprocessing system used molten cadmium cathodes to reprocess metallic fuel directly at the reactor site. This system combined all minor actinides with uranium and plutonium, eliminating the need to transport plutonium-containing materials. Such reactors would operate with breeding ratios close to 1.00, requiring only small amounts of natural uranium for refueling. A 1 gigawatt reactor might need about the size of a milk crate of natural uranium delivered monthly. These self-contained reactors are seen as a future goal for nuclear design. The IFR project was canceled.

Fuel resources

Breeder reactors can, in theory, extract most of the energy from uranium or thorium. This would use about 100 times less fuel than light water reactors, which currently extract less than 1% of the energy from uranium or thorium mined from the Earth. The high fuel efficiency of breeder reactors could reduce worries about fuel supply, energy used in mining, and the storage of radioactive waste. If uranium is extracted from seawater (a process that is not yet cost-effective), there is enough fuel for breeder reactors to meet the world’s energy needs for 5 billion years at the energy consumption rate of 1983. This would make nuclear energy effectively renewable. In addition to seawater, the average granite rock in Earth’s crust contains significant amounts of uranium and thorium. With breeder reactors, these materials could supply energy for the entire lifespan of the sun during its main sequence phase.

Spent nuclear fuel has three main parts. The first is fission products, which are the leftover pieces of fuel atoms after they split to release energy. These fission products include many different elements and isotopes, all lighter than uranium. The second part is transuranics, which are atoms heavier than uranium. These are created when uranium or heavier atoms in the fuel absorb neutrons but do not split. All transuranics belong to the actinide series on the periodic table and are often called actinides. The largest part of spent fuel is remaining uranium, which is about 98.25% uranium-238, 1.1% uranium-235, and 0.65% uranium-236. The uranium-236 comes from a reaction where uranium-235 absorbs a neutron but releases only a high-energy gamma ray instead of splitting.

Fission products behave differently from actinides. Unlike actinides, fission products cannot split and cannot be used as nuclear fuel. In fact, fission products often act as neutron poisons, absorbing neutrons that could sustain a chain reaction. Because of this, fission products are sometimes called nuclear "ashes" left after using fissile materials. Only seven long-lived fission product isotopes have half-lives longer than 100 years, making their storage or disposal less complicated than for transuranic materials.

With growing concerns about nuclear waste, breeding fuel cycles have gained renewed interest because they can reduce actinide waste, especially plutonium and minor actinides. Breeder reactors are designed to split actinide waste as fuel, converting it into more fission products. After spent fuel is removed from a light water reactor, it decays over time as each atom decays at a different rate. There is a large difference in the decay rates of fission products and transuranic isotopes. If transuranics remain in the waste, after 1,000 to 100,000 years, their slow decay would create most of the radioactivity in the spent fuel. Removing transuranics from the waste reduces much of the long-term radioactivity.

Today’s commercial light-water reactors produce some new fissile material, mostly plutonium. However, these reactors were not designed as breeders and do not convert enough uranium-238 into plutonium to replace the uranium-235 used. Still, at least one-third of the power from commercial reactors comes from splitting plutonium created within the fuel. Even with this level of plutonium use, light water reactors only consume part of the plutonium and minor actinides they produce. Nonfissile plutonium isotopes and other minor actinides build up over time.

Breeding fuel cycles have drawn interest because they can reduce actinide waste, especially plutonium and minor actinides like neptunium, americium, and curium. Breeder reactors in a closed fuel cycle would use nearly all the actinides fed into them as fuel, reducing fuel needs by about 100 times. The amount of waste produced would also decrease by about 100 times. While breeder reactors create less waste, the radioactivity of the waste is similar to that from light-water reactors.

Waste from breeder reactors behaves differently because it is made of different materials. Breeder reactor waste is mostly fission products, while light-water reactor waste includes unused uranium isotopes and large amounts of transuranics. After 100,000 years, transuranics would be the main source of radioactivity in light-water reactor waste. Removing them would reduce long-term radioactivity.

In theory, breeder fuel cycles can recycle and use all actinides, leaving only fission products. Fission products have a unique "gap" in their half-lives, meaning none have a half-life between 91 and 200,000 years. Because of this, after several hundred years of storage, the radioactivity from breeder reactor waste would drop quickly to the level of long-lived fission products. However, this benefit requires highly efficient separation of transuranics from spent fuel. If reprocessing methods leave transuranics in the waste, this advantage is reduced.

Fast neutrons in breeder reactors can split actinide nuclei with even numbers of protons and neutrons. These nuclei do not have the low-speed "thermal neutron" resonances found in fuels used in light water reactors. The thorium fuel cycle naturally produces fewer heavy actinides. The fertile material in the thorium cycle has an atomic weight of 232, while the fertile material in the uranium cycle has an atomic weight of 238. This mass difference means thorium-232 requires six more neutron capture events per nucleus before transuranic elements can form. In addition, the reactor has two chances to split nuclei as the mass increases: first as uranium-233 and later as uranium-235.

A reactor designed to destroy actinides rather than produce more fissile fuel is sometimes called a burner reactor. Both breeding and burning depend on efficient neutron use, and many reactor designs can do either. Breeding reactors

Design

A reactor's performance can be measured by its "conversion ratio," which compares the number of new fissile atoms created to the number of fissile atoms used. Most nuclear reactors, except those specifically designed to burn actinides, produce some new fissile material. This happens whenever fertile material is exposed to the reactor's neutron flux. When the conversion ratio is greater than 1, it is called the "breeding ratio."

For example, light water reactors typically have a conversion ratio of about 0.6. Pressurized heavy-water reactors using natural uranium have a conversion ratio of 0.8. In a breeder reactor, the conversion ratio is higher than 1. When the conversion ratio reaches 1.0, the reactor produces as much fissile material as it uses, a point called "break-even."

The "doubling time" is the time needed for a breeder reactor to create enough new fissile material to replace its original fuel and produce an equal amount of fuel for another reactor. This was an important measure in the past, when uranium was thought to be limited. However, since uranium is now known to be more abundant and because spent fuel contains usable plutonium, doubling time is less important in modern breeder designs.

"Burnup" measures how much energy is extracted from a given amount of heavy metal fuel, often expressed in gigawatt-days per ton of heavy metal. Burnup affects the types and amounts of isotopes produced in a reactor. Breeder reactors have higher burnup than conventional reactors because they produce more fission products and aim to destroy most actinides through fission.

Historically, breeder reactors focused on low breeding ratios, such as 1.01 for the Shippingport Reactor using thorium fuel and 1.2 for the Soviet BN-350 reactor. Theoretical models suggest breeding ratios of at least 1.8 could be achieved with certain reactor designs. The Soviet BR-1 test reactor reached a breeding ratio of 2.5 under non-commercial conditions.

Fission in any reactor produces neutron-absorbing fission products. In breeder reactors, fertile material must be reprocessed to remove these neutron-absorbing materials. This step is necessary to fully use the reactor's ability to produce more fuel than it consumes. All reprocessing methods can raise concerns about nuclear proliferation because they may extract weapons-usable materials. The PUREX method, which separates plutonium, is particularly concerning. Early breeder reactor designs using PUREX posed even greater proliferation risks due to the isotopic form of plutonium produced.

Some countries are developing reprocessing methods that do not separate plutonium from other actinides. For example, the pyrometallurgical electrowinning process used in integral fast reactors leaves radioactive actinides in the fuel. Other systems, such as SANEX, UNEX, DIAMEX, COEX, and TRUEX, and combinations with PUREX, offer better proliferation resistance than PUREX, though they are not widely used.

In the thorium cycle, thorium-232 converts to protactinium-233, which then becomes uranium-233. If protactinium remains in the reactor, small amounts of uranium-232 are also produced, which emits strong gamma radiation through its decay chain. In commercial thorium reactors, high levels of uranium-232 would accumulate, creating extremely high gamma radiation doses from uranium derived from thorium. These gamma rays make it difficult to handle uranium safely and complicate the design of nuclear weapons, which is why uranium-233 has not been used for weapons beyond early tests.

While the thorium cycle may reduce proliferation risks related to extracting uranium-233 from fuel (due to the presence of uranium-232), it still poses a risk if protactinium-233 is chemically extracted and allowed to decay to pure uranium-233 outside the reactor. This process is not required for normal reactor operation but could occur outside oversight by groups like the International Atomic Energy Agency (IAEA) and must be prevented.

Production

Fast breeder reactors have caused a lot of discussion and disagreement for many years. In 2010, the International Panel on Fissile Materials stated, "After six decades and spending the equivalent of tens of billions of dollars, the promise of breeder reactors has not been fully achieved, and efforts to use them in real-world applications have decreased over time." In Germany, the United Kingdom, and the United States, programs to develop breeder reactors have been stopped. The reasons for pursuing breeder reactors—sometimes clearly stated and sometimes implied—were based on these key ideas:

• It was believed that uranium would run out quickly if nuclear power expanded, but uranium has been more available and cheaper than expected since the Cold War ended.
• It was believed breeder reactors would be cheaper than current reactors, but they cost at least 25% more to build.
• It was thought breeder reactors would be as safe as other reactors, but sodium coolant leaks can cause fires in fast reactors.
• It was believed risks from plutonium recycling could be controlled, but all breeding cycles could pose risks. However, U-232 in U-233 produced in breeder reactors makes weapons harder to handle and easier to detect.

Some people who once opposed nuclear power now support it because breeder reactors recycle waste, solving a major issue. In the documentary Pandora's Promise, breeder reactors are praised for providing a clean energy alternative. The movie says one pound of uranium equals 5,000 barrels of oil in energy.

The Soviet Union built several fast reactors. The first used mercury coolant and plutonium metal. Later ones used sodium coolant and plutonium oxide. BR-1 (1955) was 100W (thermal), followed by BR-2 at 100 kW, then the 5 MW BR-5. BOR-60 (first criticality 1969) was 60 MW, with construction starting in 1965.

Future plants

India is working on fast breeder reactors as part of its three-step nuclear energy plan. The 500 MWe Prototype Fast Breeder Reactor (PFBR) in Kalpakkam reached a key step on April 6, 2026. This reactor uses thorium-232 to create uranium-233. India focuses on thorium because it has a lot of thorium, even though thorium is also found in many other countries.

BHAVINI, an Indian nuclear company, was created in 2003 to build and operate fast breeder reactors in the second stage of India’s nuclear plan. To support these plans, the FBR-600 is a sodium-cooled reactor with a power output of 600 MWe.

China’s Experimental Fast Reactor is a 25 MWe prototype for a larger planned reactor. It began producing power in 2011. China started a research project in 2011 to study thorium molten-salt thermal breeder-reactor technology. The goal was to develop a thorium-based molten salt nuclear system over about 20 years.

South Korea is designing a standardized modular fast breeder reactor for export. This would join their existing pressurized water reactor and CANDU designs. However, South Korea has not yet built a prototype.

Russia plans to greatly increase its number of fast breeder reactors. The BN-800 reactor (800 MWe) at Beloyarsk was completed in 2012 and reached full power in 2016. A larger BN-1200 reactor (1,200 MWe) was planned for completion by 2018, with two more planned by 2030. However, construction was paused in 2015 to improve fuel design and address cost concerns.

An experimental lead-cooled fast reactor, BREST-300, will be built at the Siberian Chemical Combine in Seversk. This design is expected to replace the BN series. The 300 MWe unit could lead to a 1,200 MWe version for commercial use. The project is part of a Russian program to improve uranium efficiency and reduce nuclear waste. The reactor’s core would be 2.3 meters wide and 1.1 meters tall, holding 16 tonnes of fuel. It would be refueled yearly, with each fuel element staying in the core for five years. The lead coolant would operate at about 540°C, achieving 43% efficiency and producing 300 MWe of electricity. The reactor could last 60 years. Construction was expected to begin between 2016 and 2020, with the cooling tower completed by 2024 and operation starting by 2026.

In 2006, the United States, France, and Japan agreed to research sodium-cooled fast reactors as part of the Global Nuclear Energy Partnership. In 2007, Japan selected Mitsubishi Heavy Industries to lead its fast breeder reactor development. Shortly after, Mitsubishi FBR Systems was created to develop and sell fast breeder technology.

In 2010, France provided €651.6 million to finalize the design of ASTRID, a 600 MWe fourth-generation reactor. The design was to be completed by 2020. By 2013, the UK showed interest in the PRISM reactor and worked with France on ASTRID. In 2019, France announced the design would not be built before mid-century.

Kirk Sorensen, a former NASA scientist, has promoted thorium fuel cycles and liquid fluoride thorium reactors. In 2011, he founded Flibe Energy to develop 20–50 MWe LFTR designs for military bases.

In 2010, GE Hitachi Nuclear Energy signed an agreement with the US Department of Energy’s Savannah River Site to build a demonstration plant for its S-PRISM fast breeder reactor before full approval. In 2011, the UK Nuclear Decommissioning Authority requested details about PRISM to help reduce its plutonium stockpile.

A traveling wave reactor, proposed by Intellectual Ventures, is a fast breeder reactor that does not require fuel reprocessing. The reactor’s design uses a "standing wave" core, where fuel rods are moved instead of a burn region moving through the fuel. This avoids cooling challenges and allows remote fuel reconfiguration without stopping operations.