Sodium-cooled fast reactor

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search
Pool type sodium-cooled fast reactor (SFR)

A sodium-cooled fast reactor is a fast neutron reactor cooled by liquid sodium.

The initials SFR in particular refer to two Generation IV reactor proposals, one based on existing liquid metal cooled reactor (LMFR) technology using mixed oxide fuel (MOX), and one based on the metal-fueled integral fast reactor.

Several sodium-cooled fast reactors have been built and some are in operation, particularly in Russia.[1] Others are in planning or under construction. For example, in the US, TerraPower is planning to build its own reactors in partnership with GEHitachi, under the Natrium appellation.[2]

Aside from the Russian experience, Japan, India, China, France and the USA are investing in the technology.

Fuel cycle[edit]

The nuclear fuel cycle employs a full actinide recycle with two major options: One is an intermediate-size (150–600 MWe) sodium-cooled reactor with uranium-plutonium-minor-actinide-zirconium metal alloy fuel, supported by a fuel cycle based on pyrometallurgical reprocessing in facilities integrated with the reactor. The second is a medium to large (500–1,500 MWe) sodium-cooled reactor with mixed uranium-plutonium oxide fuel, supported by a fuel cycle based upon advanced aqueous processing at a central location serving multiple reactors. The outlet temperature is approximately 510–550 degrees C for both.

Sodium coolant[edit]

Liquid metallic sodium may be used to carry heat from the core. Sodium has only one stable isotope, sodium-23. Sodium-23 is a weak neutron absorber. When it does absorb a neutron it produces sodium-24, which has a half-life of 15 hours and decays to stable isotope magnesium-24.

Pool or loop type[edit]

Schematic diagram showing the difference between the Pool and Loop designs of a liquid metal fast breeder reactor

The two main design approaches to sodium-cooled reactors are pool type and loop type.

In the pool type, the primary coolant is contained in the main reactor vessel, which therefore includes the reactor core and a heat exchanger. The US EBR-2, French Phénix and others used this approach, and it is used by India's Prototype Fast Breeder Reactor and China's CFR-600.

In the loop type, the heat exchangers are outside the reactor tank. The French Rapsodie, British Prototype Fast Reactor and others used this approach.


All fast reactors have several advantages over the current fleet of water based reactors in that the waste streams are significantly reduced. Crucially, when a reactor runs on fast neutrons, the plutonium isotopes are far more likely to fission upon absorbing a neutron. Thus, fast neutrons have a smaller chance of being captured by the uranium and plutonium, but when they are captured, have a much bigger chance of causing a fission. This means that the inventory of transuranic waste is non existent from fast reactors.

The primary advantage of liquid metal coolants, such as liquid sodium, is that metal atoms are weak neutron moderators. Water is a much stronger neutron moderator because the hydrogen atoms found in water are much lighter than metal atoms, and therefore neutrons lose more energy in collisions with hydrogen atoms. This makes it difficult to use water as a coolant for a fast reactor because the water tends to slow (moderate) the fast neutrons into thermal neutrons (although concepts for reduced moderation water reactors exist).

Another advantage of liquid sodium coolant is that sodium melts at 371K and boils / vaporizes at 1156K, a difference of 785K between solid / frozen and gas / vapor states. By comparison, the liquid temperature range of water (between ice and gas) is just 100K at normal, sea-level atmospheric pressure conditions. Despite sodium's low specific heat (as compared to water), this enables the absorption of significant heat in the liquid phase, while maintaining large safety margins. Moreover, the high thermal conductivity of sodium effectively creates a reservoir of heat capacity that provides thermal inertia against overheating.[3] Sodium need not be pressurized since its boiling point is much higher than the reactor's operating temperature, and sodium does not corrode steel reactor part, and in fact, protects metals from corrosion.[3] The high temperatures reached by the coolant (the Phénix reactor outlet temperature was 560 C) permit a higher thermodynamic efficiency than in water cooled reactors.[4] The electrically conductive molten sodium can be moved by electromagnetic pumps.[4] The fact that the sodium is not pressurized implies that a mich thinner reactor vessel can be used (e.g. 2 cm thick). Combined with the much higher temperatures achieved in the reactor, this means that the reactor in shutdown mode can be passively cooled. For example, air ducts can be engineered so that all the decay heat after shutdown is removed by natural convection, and no pumping action is required. Reactors of this type are self-controlling. If the temperature of the core increases, the core will expand slightly, which means that more neutrons will escape the core, slowing down the reaction.


A disadvantage of sodium is its chemical reactivity, which requires special precautions to prevent and suppress fires. If sodium comes into contact with water it reacts to produce sodium hydroxide and hydrogen, and the hydrogen burns in contact with air. This was the case at the Monju Nuclear Power Plant in a 1995 accident. In addition, neutron capture causes it to become radioactive; albeit with a half-life of only 15 hours.[3]

Another problem is leaks. Leaking sodium at high temperatures that escapes into a compartment filled with air will start to burn, albeit very slowly. Thus, sodium metal in contact with water will explode, however, it burns gently in air.[citation needed] A Russian breeder reactor, the BN-600, reported 27 sodium leaks in a 17-year period, 14 of which led to sodium fires.[5]

Design goals[edit]

Actinides[6] by decay chain Half-life
range (a)
Fission products of 235U by yield[7]
4n 4n+1 4n+2 4n+3
4.5–7% 0.04–1.25% <0.001%
228Ra 4–6 a 155Euþ
244Cmƒ 241Puƒ 250Cf 227Ac 10–29 a 90Sr 85Kr 113mCdþ
232Uƒ 238Puƒ 243Cmƒ 29–97 a 137Cs 151Smþ 121mSn
248Bk[8] 249Cfƒ 242mAmƒ 141–351 a

No fission products
have a half-life
in the range of
100 a–210 ka ...

241Amƒ 251Cfƒ[9] 430–900 a
226Ra 247Bk 1.3–1.6 ka
240Pu 229Th 246Cmƒ 243Amƒ 4.7–7.4 ka
245Cmƒ 250Cm 8.3–8.5 ka
239Puƒ 24.1 ka
230Th 231Pa 32–76 ka
236Npƒ 233Uƒ 234U 150–250 ka 99Tc 126Sn
248Cm 242Pu 327–375 ka 79Se
1.53 Ma 93Zr
237Npƒ 2.1–6.5 Ma 135Cs 107Pd
236U 247Cmƒ 15–24 Ma 129I
244Pu 80 Ma

... nor beyond 15.7 Ma[10]

232Th 238U 235Uƒ№ 0.7–14.1 Ga

Legend for superscript symbols
₡  has thermal neutron capture cross section in the range of 8–50 barns
ƒ  fissile
metastable isomer
№  primarily a naturally occurring radioactive material (NORM)
þ  neutron poison (thermal neutron capture cross section greater than 3k barns)
†  range 4–97 a: Medium-lived fission product
‡  over 200 ka: Long-lived fission product

The operating temperature must not exceed the fuel's melting temperature. Fuel-to-cladding chemical interaction (FCCI) has to be accommodated. FCCI is eutectic melting between the fuel and the cladding; uranium, plutonium, and lanthanum (a fission product) inter-diffuse with the iron of the cladding. The alloy that forms has a low eutectic melting temperature. FCCI causes the cladding to reduce in strength and even rupture. The amount of transuranic transmutation is limited by the production of plutonium from uranium. One work-around is to have an inert matrix, using, e.g., magnesium oxide. Magnesium oxide has an order of magnitude lower probability of interacting with neutrons (thermal and fast) than elements such as iron.[11]

High-level wastes and, in particular, management of plutonium and other actinides must be handled. Safety features include a long thermal response time, a large margin to coolant boiling, a primary cooling system that operates near atmospheric pressure, and an intermediate sodium system between the radioactive sodium in the primary system and the water and steam in the power plant. Innovations can reduce capital cost, such as modular designs, removing a primary loop, integrating the pump and intermediate heat exchanger, and better materials.[12]

The SFR's fast spectrum makes it possible to use available fissile and fertile materials (including depleted uranium) considerably more efficiently than thermal spectrum reactors with once-through fuel cycles.


In 2020 Natrium received an $80M grant from the US Department of Energy for development of its SFR. The program plans to use High-Assay, Low Enriched Uranium fuel containing 5-20% uranium. The reactor was expected be sited underground and have gravity-inserted control rods. Because it operates at atmospheric pressure, a large containment shield is not necessary. Because of its large heat storage capacity, it was expected to be able to produce surge power of 550 MWe for 5+ hours, beyond its continuous power of 345 MWe.[13]


Sodium-cooled reactors have included:

Model Country Thermal power (MW) Electric power (MW) Year of commission Year of decommission Notes
BN-350  Soviet Union 135 1973 1999 Was used to power a water de-salination plant.
BN-600  Soviet Union 600 1980 Operational Together with the BN-800, one of only two commercial fast reactors in the world.
BN-800  Soviet Union/ Russia 2100 880 2015 Operational Together with the BN-600, one of only two commercial fast reactors in the world.
BN-1200  Russia 2900 1220 2036 Not yet constructed In development. Will be followed by BN-1200M as a model for export.
CEFR  China 65 20 2012 Operational
CRBRP  United States 1000 350 Never built Never built
EBR-1  United States 1.4 0.2 1950 1964
EBR-2  United States 62.5 20 1965 1994
Fermi 1  United States 200 69 1963 1975
Sodium Reactor Experiment  United States 20 65 1957 1964
S1G  United States United States naval reactors
S2G  United States United States naval reactors
Fast Flux Test Facility
 United States
PFR  United Kingdom 500 250 1974 1994
FBTR  India 40 13.2 1985 Operational
PFBR  India 500 2020 Under construction Under construction
Monju  Japan 714 280 1995/2010 2010 Suspended for 15 years. Reactivated in 2010, then permanently closed
Jōyō  Japan 150 1971 Operational
SNR-300  Germany 327 1985 1991
Rapsodie  France 40 24 1967 1983
Phénix  France 590 250 1973 2010
Superphénix  France 3000 1242 1986 1997 Largest SFR ever built.

Most of these were experimental plants that are no longer operational. On November 30, 2019, CTV reported that the Canadian provinces of New Brunswick, Ontario and Saskatchewan planned an announcement about a joint plan to cooperate on small sodium fast modular nuclear reactors from New Brunswick-based ARC Nuclear Canada.[14]

See also[edit]


  1. ^
  2. ^ Gardner, Timothy (August 28, 2020). "Bill Gates' nuclear venture plans reactor to complement solar, wind power boom" – via
  3. ^ a b c Fanning, Thomas H. (May 3, 2007). "Sodium as a Fast Reactor Coolant" (PDF). Topical Seminar Series on Sodium Fast Reactors. Nuclear Engineering Division, U.S. Nuclear Regulatory Commission, U.S. Department of Energy. Archived from the original (PDF) on January 13, 2013.
  4. ^ a b Bonin, Bernhard; Klein, Etienne (2012). Le nucléaire expliqué par des physiciens.
  5. ^
  6. ^ Plus radium (element 88). While actually a sub-actinide, it immediately precedes actinium (89) and follows a three-element gap of instability after polonium (84) where no nuclides have half-lives of at least four years (the longest-lived nuclide in the gap is radon-222 with a half life of less than four days). Radium's longest lived isotope, at 1,600 years, thus merits the element's inclusion here.
  7. ^ Specifically from thermal neutron fission of uranium-235, e.g. in a typical nuclear reactor.
  8. ^ Milsted, J.; Friedman, A. M.; Stevens, C. M. (1965). "The alpha half-life of berkelium-247; a new long-lived isomer of berkelium-248". Nuclear Physics. 71 (2): 299. Bibcode:1965NucPh..71..299M. doi:10.1016/0029-5582(65)90719-4.
    "The isotopic analyses disclosed a species of mass 248 in constant abundance in three samples analysed over a period of about 10 months. This was ascribed to an isomer of Bk248 with a half-life greater than 9 [years]. No growth of Cf248 was detected, and a lower limit for the β half-life can be set at about 104 [years]. No alpha activity attributable to the new isomer has been detected; the alpha half-life is probably greater than 300 [years]."
  9. ^ This is the heaviest nuclide with a half-life of at least four years before the "Sea of Instability".
  10. ^ Excluding those "classically stable" nuclides with half-lives significantly in excess of 232Th; e.g., while 113mCd has a half-life of only fourteen years, that of 113Cd is nearly eight quadrillion years.
  11. ^ Bays SE, Ferrer RM, Pope MA, Forget B (February 2008). "Neutronic Assessment of Transmutation Target Compositions in Heterogeneous Sodium Fast Reactor Geometries" (PDF). Idaho National Laboratory, U.S. Department of Energy. INL/EXT-07-13643 Rev. 1. Archived from the original (PDF) on 2012-02-12.
  12. ^ Lineberry MJ, Allen TR (October 2002). "The Sodium-Cooled Fast Reactor (SFR)" (PDF). Argonne National Laboratory, US Department of Energy. ANL/NT/CP-108933.
  13. ^ "Bill Gates's next-gen nuclear plant packs in grid-scale energy storage". New Atlas. 2021-03-09. Retrieved 2021-06-03.
  14. ^ "Three premiers plan to fight climate change by investing in small nuclear reactors". CTVNews. November 30, 2019.

External links[edit]