This year marks the 75th anniversary of the end of World War II. While there can be no doubt that it was the cause of immeasurable and far-reaching destruction and loss, that it was a catalyst for pivotal military and civilian technological advancement is equally undeniable. The energy that is produced in fission, fusion, or even the combustion of fossil fuel comes from mass defects: a very small part of the invariant mass of the reactants is lost and is converted into kinetic energy. Among those three energy sources, the fraction of mass converted to kinetic energy during a reaction is highest in fusion reactions, then fission reactions, and then fossil fuel reactions which aren’t even comparable to the other two.
Fusion energy isn’t yet viable because it necessitates a large supply of energy for plasma heating and confinement, which is difficult to effectively contain and maintain with the available materials and confinement technology. Adequate heating and confinement conditions for self-sustained fusion have only been achieved in nuclear weapons explosions.
Nuclear fission is the highest energy density source practically available today to energy applications. Fission was discovered in Nazi Germany by Otto Hahn and Lise Meitner in 1938. It occurs when an unstable nucleus like that of uranium-235 disintegrates and splits into lighter elements when hit by a free neutron. In fission reactions, energy and free neutrons are released. The number of free neutrons released depends, in part, on the energy of the incoming neutron. This is an important fact that makes fast “breeder” reactors possible, as explained later.
Much of our current knowledge in nuclear energy traces back to the Manhattan Project and the two decades that followed. The Manhattan Project developed the first nuclear weapons as well as the fundamental knowledge used later in peaceful nuclear energy.
Many of the most important contributions to the Manhattan Project were made by scientists of Jewish background, some of whom had fled the Holocaust in Europe and some were born in the United States. Rudolf Peierls made key contributions to neutron transport theory; John von Neumann and Stanislaw Ulam developed the Monte Carlo method and made fundamental contributions to numerical computation on ENIAC and theoretical design of nuclear weapons; Emilio Segre discovered that spontaneous fission in plutonium was caused by the plutonium-240 isotope; Edward Teller made significant contributions to the implosion design and hydrogen bomb development later on, and J. Robert Oppenheimer was the director of the Los Alamos National Laboratory. Many other scientists of Jewish background made significant contributions to the Manhattan Project, including Leo Szilard, Eugene Wigner, David Bohm, and Richard Feynman, but exploring nuclear history in detail is beyond the scope of this discussion.
Neutrons are normally present inside atomic nuclei just like protons, but there are various ways they can exit the nucleus and then freely move and interact. Most importantly, neutrons may get captured and transmute the nucleus that captured them into another isotope (often unstable), they may cause certain nuclides to fission; they may collide, change direction and lose energy. The probability of any interaction strongly depends on the material and on the speed/energy and, in some materials, the direction of the incoming neutron.
The free neutrons from fission can trigger other reactions. At the level of one neutron or a few neutrons, the interactions are probabilistic. But when a very large number of neutrons are available, the results are deterministic and the neutron population can be characterized mathematically. For the chain reaction to continue, at least one neutron produced from each fission should induce another fission reaction. Otherwise, the neutron population would decrease and the chain reaction would come to an end in the absence of another source.
Most materials, or more properly nuclides, have stable nuclei and aren’t fissionable. Some materials are fissionable and may fission if struck by an energetic enough neutron (e.g., uranium-238 and thorium-232). Other materials are “fissile” and may fission when struck by a neutron of any energy such as uranium-233, uranium-235, and plutonium-239. It is fissile materials that can sustain the chain reaction and are practically the fuel. Materials can also be “fertile” like thorium-232 and uranium-238, which can transmute to other isotopes that eventually decay to a fissile material like uranium-233 or plutonium-239. It is fertile materials that can create fuel when they capture a neutron.
The vast majority of uranium present in nature is not fissile. Uranium-235 (fissile) represents only 0.72% of the uranium in nature, while uranium-238 (nonfissile) represents 99.27%. This is what usually necessitates the enrichment of uranium in the uranium-235 isotope for use in nuclear reactors and other nuclear applications. On the other hand, plutonium is found in nature in trace quantities and is produced in nuclear reactors by the conversion of uranium-238.
Different nuclear reactor designs use different levels of uranium enrichment. As an example, the Canadian CANDU heavy water reactors don’t necessitate the enrichment of uranium because they use the heavier isotope of hydrogen (deuterium) which doesn’t absorb neutrons as much as the lighter and more common isotope of hydrogen (protium). This reduced absorption allows the chain reaction to continue in the presence of more quantities of uranium-238 than in light water reactors.
The fuel of light water reactors is enriched approximately to 3% to 5% uranium-235. Enrichment, nowadays done by spinning uranium hexafluoride gas in centrifuges to separate the heavier and lighter isotopes, is an energy-expensive process that is subject to international regulations to prevent enrichment to weapons grade. Most of the uranium in the reactor is not used: Some is converted to plutonium, and most is sent to spent fuel storage. Currently, only about 5% of the uranium in the reactor is consumed and about 1.15% of the fuel in light water reactors becomes plutonium, which can be used as a fuel.
A breeder is a reactor that produces at least as much fissile fuel (e.g., plutonium-239 or uranium-233) as the fissile fuel it burns in the chain reaction. But before you rush to conclusions, this is not exactly infinite fuel—as a fertile U-238 or Th-232 blanket is consumed and as the reactivity decreases as neutron-absorbing parasitic poisons differentially build up in the system, which limits the achievable burn-up.
The most important physical parameter to the neutron supply is the number of the neutrons emitted per absorption in the fuel, which is controlled by the neutron energy spectrum. Baby neutrons are born out of fission energetic and fast. They slow down as they age due to scattering collisions. When neutrons collide with heavier nuclei like that of lead, they don’t slow down as much as when they collide with lighter nuclei like that of hydrogen. However, fast neutrons, for most part, are less likely to interact with a nuclei per unit distance traveled than slower neutrons.
So what exactly is the advantage of fast neutrons? It’s the fact that in the fast spectrum, the neutron yield per absorption in the fuel is higher which contributes additional neutrons that allow for breeding fuel. The absolute theoretical minimum for breeding is two neutrons per absorption in the fuel—one to cause fission in order to sustain the chain reaction and the other to convert fertile material to fissile. Nevertheless, not every fast spectrum reactor is a breeder, as the losses in structural materials and losses due to leakage would adversely affect the neutron economy. The way you spend your neutrons matters.
Materials and geometry are of extreme importance in nuclear reactor design. Different materials have different neutronic and thermophysical properties which are dependent on many parameters from neutron energy spectrum to temperature. By changing the materials in the reactor, their distribution, and their shape, engineers can control the energy spectrum of the neutrons, the energy deposition and heat distribution, reactor kinetics and many other parameters.
Reactor design is an inherently iterative process that necessitates large computational resources. The engineers’ job is to find sweet spots for their design targets and priorities whether it’s higher core outlet temperature and better thermodynamic efficiency, or increased safety, or increased proliferation resistance, or increased conversion and fissile material production.
Money also plays a key role in the design choices. Thorough experiments are often very expensive. Designs that necessitate the use of undercharacterized materials are often not economically viable in the short term, which necessitates government investment in research to create information and knowledge that enables the safe use of those materials, allowing for technology to move forward.
Fast spectrum reactors use heavier coolants that don’t slow down neutrons such as sodium, 23 times as heavy as hydrogen, or even lead, 207 times as heavy as hydrogen. They are practically not moderated, unlike light water reactors. As a result, more neutrons are released per absorption in the fuel which can be used to increase plutonium production or transmute long-lived minor actinides to other actinides with shorter half-life.
There are only a handful of fast spectrum reactors active today and most of them are experimental reactors. Russia has the BN-600 and BN-800 sodium-cooled fast breeder reactors which generate 560 MWe and 880 MWe, respectively. They are the only two nonexperimental fast breeder reactors active today that deliver significant energy to a grid. There are a few other active experimental reactors in operation in Russia, China, and India that produce low power and are mainly used for material compatibility testing. There are other breeder reactors under construction such as the Prototype Fast Breeder Reactor in India and the CFR-600 in China which should generate 500 MWe and 600 MWe, respectively.
In the past, France had possessed the Superphénix reactor, which produced an impressive 1,242 MW of electric power. The facility was decommissioned in 1997 following lengthy political opposition by “green party” groups which led the then-prime minister, Lionel Jospin, to announce the closure of the plant citing “excessive costs,” which can’t be true given that the major costs are incurred on construction of the reactor, not operation. The facility had faced an RPG rocket attack by a member of the Swiss Green Party in 1982 before it was commissioned in 1986.
The problem of nuclear waste was solved many decades ago at the engineering level. At the social and political level, it remains the biggest challenge to nuclear energy, although there are legitimate concerns on maintaining long-lived waste that could remain active for thousands of years. Such waste, however, can be transmuted to shorter lived forms in breeder reactors. That’s one of the main advantages of having some neutrons to spare from fission reactions in the higher energy spectrum.
Although the United States led the conceptualization and the early experimentation in fast spectrum reactors, the United States doesn’t currently possess active fast spectrum reactors, unless we count NASA’s Kilopower space reactor which is designed for low energy application. In the past, the United States possessed experimental facilities such as the Fast Flux Test Facility at Hanford, Washington, and SEFOR in Arkansas that used sodium coolant. Perhaps the lack of interest stems from the fact that the United States uses a once-through nuclear fuel cycle. The spent fuel is sent to storage and isn’t processed. Some other countries like Japan and India reprocess the fuel to extract the plutonium for use in mixed oxide (MOX) fuel assemblies that use both plutonium and uranium; such a cycle is known as the closed fuel cycle—which, while more efficient, often raises nuclear weapons proliferation concerns.
Russia had also long used lead-bismuth eutectic cooled fast reactors for submarine propulsion. The use of lead-bismuth eutectic rather than lead is because it has a much lower melting point than lead (123°C vs. 327°C) which makes it easier to use as a coolant. Additionally, compared to light water reactors, lead cooled reactors can achieve higher power densities (~100 kW/L vs. ~200 kW/L) although less than sodium cooled reactors which can operate at core power densities greater than 300 kW/L, but aren’t self-shielded against gammas which negates this advantage in a space-limited environment.
Submarine reactors in general tend to use highly enriched uranium to increase burn-up and allow for longer operation without refueling as well as reduce the size of the core. The reduction of the core size increases the power density for the same amount of power produced and inherently necessitates additional safety features compared to reactors that operate with low enriched uranium.
The negative void reactivity feedback in lead cooled reactors, the high boiling point of lead and lead-bismuth (~1700°C), and the poor chemical reactivity of lead with air and water are all safety aspects that make lead even more attractive over sodium. The main problem with lead and lead-bismuth, however, is corrosion of commonly used structural steels which reduces the reactor life and makes it economically unattractive to the commercial energy sector. Lead selectively leeches elements such as nickel out of stainless steel. This problem is even worse in lead-bismuth and intensifies at higher temperatures, practically restricting the power production as increasing the coolant velocity increases erosion.
In the United States, there appears to be renewed interest in developing fast spectrum reactor technology. In 2018 and 2019, Congress allocated a total of $100 million to the Department of Energy to develop conceptual design and conduct research to support the development of an advanced test reactor known as the Versatile Test Reactor (VTR) based on requests from companies working on advanced reactor designs. If built, the VTR would likely be sodium-cooled and would provide a platform that allows for multiple cartridges for testing material compatibility for different reactor types such as molten salt reactors, gas cooled reactors, lead cooled reactors, and sodium cooled reactors, which is an important step toward restoring U.S. leadership in advanced reactor technology.
Compared to lead, sodium has the advantage of a lower melting point, good material compatibility that allows for high power densities greater than 300 kW/L, and being easier to work with and pump as it’s much lighter than lead. Sodium cooled reactors, however, face challenges with positive void reactivity feedbacks. A positive void feedback means that in the very unlikely case of a severe accident where sodium boils, the voids that form in sodium will increase the multiplication of the neutrons in the reactor, which would increase the power. Reactors are generally designed to avoid single point failure and don’t rely on just one layer of safety. The positive void reactivity in sodium, although undesired, may be compensated by strongly negative fuel temperature reactivity feedback.
Lead has some major advantages over sodium including potential for use as a spallation target in accelerator driven systems. However, lead-induced corrosion makes it economically unattractive to the commercial energy sector. Commercial reactors are generally designed to operate for about 40 years. There is an ongoing search for structural materials that are resistant to corrosion in flowing lead environment. The University of New Mexico’s Lobo Loop, run by former Los Alamos National Laboratory scientist, professor Osman Anderoglu, is the only such facility in operation in the United States.
There are legitimate concerns regarding fast breeder reactors in general and the proliferation of nuclear weapons. Fast breeder reactors produce a lot of plutonium. In the absence of proper and enforceable safeguards, fissile material may be diverted to weapons use rather than reactor use leading to proliferation of nuclear weapons.
When discussing proliferation one should note there are different grades of plutonium. Safeguards can ensure that plutonium isn’t withdrawn from the reactor in the weapons grade form. Although any plutonium grade is considered a proliferation threat, not all plutonium is desired in nuclear weapons. Fuel grade plutonium has 7% to 19% Pu-240, while reactor grade plutonium has even more. An atom of plutonium-240 is ~47,000 times more likely to spontaneously fission at any given second than an atom of plutonium-239. Although reactors produce more plutonium-239 than plutonium-240, the ratio of the mass of plutonium-240 to plutonium-239 increases with time in reactors, in much of the energy spectrum, due to the conversion of some of the plutonium-239 to plutonium-240. By keeping plutonium in the reactor for longer, it accumulates more plutonium-240 and becomes less desirable—though not useless—for weapons applications. A country possessing other means to obtain plutonium wouldn’t be likely to seek fuel grade or reactor grade plutonium from a civilian reactor for weapons use, especially in the absence of online refueling in the reactor design.
On the other hand, fast breeder reactors would raise serious concerns on proliferation of nuclear weapons in countries that have no other means of obtaining plutonium—particularly in countries that don’t already have nuclear weapons capability. There are, however, cheaper and easier options for countries that seek nuclear weapons materials than fast breeder reactors. Iran for instance had pursued heavy water reactors such as the IR-40 in Arak that can operate on natural uranium, although it requires the enrichment of water in deuterium. North Korea had used a graphite-moderated reactor at Yongbyon to produce plutonium for its nuclear weapons and is also fueled by natural uranium.
Finally, in a discussion about breeder reactors, one should honorably mention thorium cycle based reactors although they are often designed as slow (thermal) neutron spectrum breeders enabled by the fact that uranium-233 emits more neutrons per absorption than uranium-235 and plutonium-239 in the thermal spectrum. A common misconception is that they are “fueled by thorium”; sorry to disappoint, but thorium, no matter how cool it sounds, isn’t technically a fuel. It’s a fertile material that can be converted to uranium-233 which is fissile and can be used as a reactor fuel.
The primary advantage of the thorium cycle is the relative abundance of thorium compared to uranium (~3x). In some countries like India with large supplies of thorium and low supplies of uranium, thorium-cycle offers the prospect of energy security and independence. Thorium cycle reactors produce less nuclear waste and very little plutonium-239 given how far thorium-232 is from plutonium-239. Nevertheless, it produces fissile uranium-233, which could be used in weapons although less attractive than plutonium which releases more neutrons per fission in the fast spectrum.
Nuclear reactors in general are built with inherent safety features based on intrinsic physics and not only engineering. The materials in the reactor are chosen such that the reactivity decreases when temperature increases. As a result, the chain reaction inherently would come to an end in the event of an accident. In all reactor accidents that ever happened in history, the number of documented direct deaths is about 280 in total – which is far less than the deaths in hydroelectric power dam failures. Cancer-related deaths attributed to reactor accidents, like Chernobyl, which was the product of a flawed engineering design that would never have been approved in the West, are not fully understood, as they occurred long after the accident. It has been demonstrated from the Chernobyl data that exposure to significant doses of radiation can increase risks for certain cancers over the long term up to 20 years.
Despite all their advantages, the main challenge for new reactor technology whether fast breeder reactors or thermal breeder reactors is that they have to compete with mature technology that has been demonstrated and experimented with for decades. This necessitates expensive investments and very thorough research that can take decades until the technology matures. New government leaders may not share the same vision of their predecessors, as in the case of France’s Superphénix reactor. And yet, it is still possible to imagine that cheap and plentiful energy may be the most lasting legacy of the Manhattan Project of World War II.
Khaled Talaat is a postdoctoral scholar in nuclear engineering at the University of New Mexico. He has conducted research on multiple subjects including aerosols, radiological protection, and Generation IV lead-cooled fast reactors.