Breeder Reactors
Abstract
Breeder reactors are designed to generate nuclear fuel at the same time as producing energy for electricity production. This is possible because a small number of isotopes will capture neutrons produced in a reactor, starting a reaction that leads to a new, heavy fissile isotope. This permits both neutron-induced fission for power generation and neutron capture, for fuel generation, to occur simultaneously. The two main, fertile, isotopes that are useful for this purpose are uranium-238 that will generate fissile plutonium and thorium-232 that will produce fissile uranium. Breeder reactors have a long history but have so far not proved successful commercially. However, a number of advanced designs use the concept. Most use a liquid metal as the coolant because it does not slow neutrons.
Keywords
Breeder reactor; fast neutron reactor; nuclear core blanket; uranium; thorium; plutonium; liquid metal coolant; pool-type reactor; loop-type reactor
Breeder reactors are a special type of nuclear reactor that can produce additional nuclear fuel at the same time as producing energy for electricity generation. All energy producing nuclear reactors rely on one of a small number of fissile elements to generate energy. These elements, discussed in Chapter 3, have nuclei that will react with neutrons and split into two smaller nuclei with the release of large quantities of energy. The only fissile nucleus that occurs naturally in significant quantities is uranium-235 which makes up roughly 0.7% of the uranium found in uranium ores.
There are two other fissile isotopes of value for energy production, plutonium-239 and uranium-233, but neither of these can be found naturally. However, both of them can be produced in nuclear reactions of a different type to those that result in fission. The most important of these alternative processes is a reaction between uranium-238 and a fast neutron that produces plutonium-239. Since uranium-238 is much, much more abundant on the Earth than uranium-235, this reaction offers a way of providing a much larger source of fissile material that the naturally occurring fissile uranium-235 isotope can provide alone.
The second breeder reaction involves thorium-232 that will react with a neutron to produce uranium-233. Thorium-232 is even more abundant than uranium-238 and so could in principle provide an even large source of fissile material. However, this reaction has yet to be exploited on a large scale for nuclear energy production. In contrast the production of plutonium has been widely explored because it is also a key element in nuclear weapons. The two isotopes, uranium-238 and thorium-232, which will react with neutrons to produce new fissile isotopes are called fertile isotopes.
The development of plutonium breeder reactors can be traced to the earliest days of the nuclear age and like other reactors these were a product of weapons development. Furthermore the potential to build reactors that could produce more fuel than they consumed was an attractive one during the early years of nuclear power when the availability of uranium was uncertain. During the two decades following World War II, the United States, the United Kingdom, France, Germany, Japan, the Soviet Union, and India all set up breeder reactor research programs. By the middle of the 1980s, with the development of nuclear power slowing and the abundance of uranium much higher than had previously been thought, the need for breeder reactors became less urgent and for most countries this research was no longer a priority. However, both India and Russia have continued development as a result of domestic fears of a uranium shortage, while China is building prototype units to a Russian design.
In spite of the wealth of research the problems encountered during development of breeder reactors have meant that no commercial unit has ever entered service, although research reactors have generated electricity. However in the last two decades a renewed interest in fast neutron breeder reactors has been stimulated by the need to find a way of disposing of nuclear waste from existing power plants. Breeder reactors can burn the plutonium formed in conventional reactors, thereby eliminating at least part of the waste fuel disposal problem. They can also destroy the actinides1 in radioactive waste from water-cooled reactors that cannot be reused for fuel. For this to be effective as a means of dealing with nuclear waste, there would need to be many more operating breeder reactors than exist today.
Plutonium Breeder Reactors
The majority of breeder reactors, whether experimental, prototypes or demonstration plants, that have been built have been plutonium breeder reactors. These are also sometimes known as nuclear fast reactors or fast breeder reactors. Plutonium-239 is a fissile material and its nucleus will split when struck by a neutron, generally producing two nuclei of smaller elements and a number of fast neutrons. Plutonium-239 is the most common nuclear fuel used in fast breeder reactors and it provides both the source of energy for electricity production and a source of fast neutrons. These fast neutron are then exploited both to generate further fission reactions and to react with uranium-238 which is also present in the reactor. Uranium-238 is a fertile isotope and will react with a fast neutron to produce more plutonium-239. In this way the breeder reactor can produce both energy and more fuel.
One of the key differences between a conventional nuclear reactor and a plutonium breeder reactor is that the latter does not have a moderator. In the conventional reactor the fast neutrons produced from uranium-235 fission reactions are slowed because slow neutrons are much more likely to react with further uranium-235 nuclei that are fast neutrons. Fast neutrons will react, but the probability of reaction is much lower. Plutonium-239 also reacts with both slow and fast neutrons but, critically, it has a higher probability of reaction with a fast neutron than uranium-235.
The fast breeder reactor requires a high density of fast neutrons because it is these that will react with uranium-238 and produce more plutonium. In order for a fast neutron reactor to achieve criticality, the core will contain a much higher percentage of fissile material—typically around 20% or more of plutonium-239—than would be found in a slow neutron reactor. This higher concentration allows a controlled chain reaction to be achieved with fast neutrons. Plutonium has a second advantage too, it produces around 25% more fast neutrons from each fission reaction than uranium-235 and this means there are more neutrons to share between fission and production of more plutonium.
The structure of a fast neutron reactor typically involves a core containing the enriched plutonium fuel, usually mixed with depleted uranium to achieve the required level of enrichment. The latter is the uranium-238 left from the enrichment of uranium and it is referred to as depleted because it has a much lower concentration of fissile uranium-235 than would be found in natural uranium. This uranium-238 within the core will produce some additional plutonium. However, in order to be able to make more plutonium that it burns, the reactor core is surrounded by a further blanket of depleted uranium. Stray fast neutrons from the core pass into this blanket and generate more plutonium. Moreover, the neutron density here is too low to lead to many fission reactions so most of the plutonium remains in the blanket, once produced.
The other key element of the fast neutron reactor is the coolant. Since the reactor uses fast neutrons the coolant cannot be either a moderator or a neutron absorber. The material that has proved the most popular coolant for fast neutron reactors is liquid sodium. A cross-section of a reactor of this type is shown in Fig. 6.1. Sodium has good heat-carrying properties and, importantly, does not absorb or slow neutrons. However the material is very reactive if exposed to air or water and so the cooling circuits have to be extremely strictly engineered. The core of a fast neutron reactor is usually smaller than that of a conventional slow neutron reactor and it has a higher power density within the core. This leads to a higher core temperature of 500–550°C. The core usually operates at atmospheric pressure, again unlike slow neutron reactors which usually operate at high pressure. But like the latter they have control rods to manage the nuclear reaction and these are made of boron carbide.
The liquid metal coolant in a fast neutron reactor is passed through a heat exchanger through which water is passed and steam generated. The steam is then used to drive a steam turbine for power production, in much the same way as a slow neutron reactor. As with the latter, the heat exchanger/steam generator may be located either inside the containment vessel that encloses the reactor core, or outside.2 Fast neutron reactors require these same protective enclosures and the same safety features as conventional reactors.
While liquid sodium is the most popular coolant, others have been tested too. Candidates include liquid lead or a lead-bismuth mixture. Gas-cooled reactors, often using helium, are also possible. However the liquid sodium offers the best breeding potential. The latter is defined by the breeding ratio, a figure that shows how much new fissile material is produced for each unit of fissile material burnt. The figure must be greater than one if the reactor is to produce more fuel than it consumes. With a sodium-cooled reactor a breeding ratio of 1.3 can be achieved. With other coolants such as lead–bismuth, the breeding ratio is usually less than one.
The fuel that is loaded into a nuclear fast reactor is normally in the form of plutonium oxide and uranium oxide. These oxides do not react with sodium or lead but they have relatively low thermal conductivities. Alternatives with high thermal conductivity such a mixed metal fuel or fuels made from uranium and plutonium carbides of nitrides have also been tested but these present other problems that make them less easy to manage than the conventional oxide fuels.
In order to close the fuel cycle for a fast neutron reactor, the fuel and the blanket material from the reactor must be processed to isolate the plutonium so that it can be used to manufacture more fuel. While developing fast neutron reactors, several countries such as the United Kingdom and France also developed nuclear waste reprocessing facilities that are capable of carrying out the large-scale separation of plutonium for fuel manufacture. Today these processing plants are more likely to be used for waste fuel reprocessing from slow neutron reactors. The plutonium produced from the fuel is then used to make a mixed oxide fuel containing both fissile uranium and fissile plutonium. This can be used as fuel in some conventional reactors. However, it could be used in breeder reactors in the future.
Fast Neutron Reactor Projects, Past and Present
According to the World Nuclear Association, around 20 fast neutron reactors have operated since the 1950s. Most of these have been experimental or pilot schemes but a small number have been demonstration projects that have produced electric power for delivery to the grid.
In the United States, the first experimental breeder reactor called EBR-1 began operating in Idaho in 1951. The plant used a sodium–potassium alloy as coolant and metallic uranium fuel. It was able to produce around 200 kW of power, enough to power the facility and ran until 1963. It was succeeded by EBR-II, a liquid sodium research reactor with a power output of 20 MW, which ran until 1994. This reactor also formed the basis for a project called the integral fast reactor which was intended to integrate a reactor, reprocessing and fuel manufacture in a single facility, again using enriched metallic uranium fuel. The only commercial breeder reactor in the United States was the Fermi 1 demonstration project in Michigan. The project started in 1963 but the plant only operated for 3 years before coolant problems forced it to close. It had a nameplate generating capacity of 60 MW but never realized it.
The United Kingdom had a fast reactor program based at Dounreay in the north of Scotland. Its first reactor was a 15-MW experimental fast reactor using sodium-potassium coolant which operated between 1959 and 1977. The reactor initially used uranium metal fuel but later tested uranium oxide fuels. This was followed by a prototype fast reactor with a generating capacity of 250 MW that ran from 1974 until 1994. The latter was a liquid sodium–cooled reactor and was intended to form the basis for a commercial reactor fleet but no commercial versions were built. It used mixed oxide fuel.
France built an experimental fast neutron reactor called Rapsodie in 1967 with no power production capability. The liquid sodium–cooled plant was shut down in 1983. However, soon after the project was started, France also began work on the 250-MW Phénix nuclear fast reactor. The reactor contained just under 1 tonne of plutonium in a fuel that was enriched to 77% plutonium-239. It was connected the grid in 1973 and was finally shut down in 2009. Even as this plant was being built, France, Germany and Italy signed a series of agreements for the construction of two commercial fast neutron reactors, one in Germany and one in France. After a tortuous gestation period work began in 1976 on the first plant, in France and christened Superphénix. The plant finally went critical in 1985 with a rated generating capacity of 1200 MW, the largest such plant ever built. However, the liquid sodium–cooled plant suffered a number of operational problems including a major sodium leak. It was finally shut down in 1998. Superphénix would have formed the basis for a European fast breeder reactor with a generating capacity of 1450 MW but work on that project has virtually been abandoned.
Japan has operated two fast breeder reactors as part of a program to develop fast neutron reactor technology. The first, called Joyo began operating in 1977 and is still functioning as a test bed for fast neutron developments. Alongside this experimental project, Japan also began construction of a prototype fast reactor called Monju. This plant, with a nameplate capacity of 280 MW, achieved criticality in 1994 but experienced a serious sodium leak in 1995 which closed the unit for the next 15 years. It restarted briefly in 2010 but has since been offline. The Japan Atomic Power Company planned a 660-MW demonstration commercial plant, but the project was canceled in the late 1990s following problems with Monju.
The other country with a long history of fast neutron reactors is Russia. Research into the technology has been underway since the 1950s with a small experimental reactor designated BR-5 operating between 1959 and 2004. This was followed by BOR-60, a 12-MW demonstration plant that used uranium oxide fuel enriched to between 45% and 75%. The plant began operating in 1969 and is still active. In 1972, a 350-MW plant, BN-350, was built in Kazakhstan. The liquid-sodium-cooled unit had two sets of heat exchangers outside the reactor vessel. Part of its output was used for desalination. The plant finally closed in 1999.
In 1980, another fast neutron reactor, designated BN-600 with a gross generating capacity of 600 MW, began supplying electricity to the grid. This plant is still operating. The unit is described as a pool-type reactor in which the reactor vessel is immersed in a pool of the liquid sodium coolant as shown in Fig. 6.2. Heat exchangers also filled with liquid sodium carry heat from the pool outside the reactor containment to a heat exchanger/steam generator where it is used to raise steam for steam turbines.
The Russian program has steered away from plutonium-239 and instead uses uranium-235 as the primary fissile material. In the case of the BN-600 the fuel is in the form of uranium oxide which has been enriched to up to 26% with uranium-235. However, this unit is expected to be adapted to burn plutonium from the Russian nuclear weapons stockpile.
The BN-600 has been followed by a larger unit of the same design, the BN800 with a gross generating capacity of 864 MW. It uses oxide fuels with a fissile isotope content of 20–30%, but can also burn uranium and plutonium nitride, or metal fuels too. The plant has primary and secondary cooling loops using sodium to carry heat to the steam generators. A loop-type metal-cooled reactor of this type is shown in Fig. 6.3. Steam is produced at 470°C and the thermal efficiency is claimed to be 39%. Meanwhile the BN-800 is the prototype for a commercial BN-1200 fast neutron reactor with a gross output of 1220 MW.
In addition to its series of liquid-sodium-cooled reactors, Russia has also been exploring reactors that use lead or a lead-bismuth mixture as the coolant. The initial work on these reactors was for submarine propulsion, but more recent designs are intended for civil power generation.
Slow Neutron Breeder Reactors
In addition to plutonium breeder reactors, there is a second type of breeder reactor in which thorium is converted into uranium-233. This isotope of uranium is also fissile, like uranium-235, and so can be used as fuel in a slow neutron reactor. The main interest in this cycle is in India where there are abundant reserves of thorium but much less uranium.
Thorium can be converted into fissile uranium with both slow and fast neutrons. This means that slow neutron reactors such as pressurized water reactors are capable of producing uranium-233 from thorium. In fact pressurized heavy water reactors (PHWRs) are some of the best reactors for this purpose and the Canadian CANDU reactor has been used as a test-bed for thorium conversion into uranium. However the creation of a closed fuel cycle in which more fissile material is generated and then the material reprocessed to make more fuel requires specially designed breeder reactors.
Breeder technology has been under investigation in India since the 1950s and the government has proposed a three-phase strategy leading to a fleet of thorium/uranium breeder reactors. The first stage involves building PHWRs to provide plutonium for a breeder reactor program. In the second stage, this plutonium would form the core of a breeder reactor that is used initially to generate more plutonium but eventually, when a large enough plutonium has been stockpiled, to generate uranium-233 from thorium in the breeder reactor blanket surrounding the core. The third stage of this program then involves building breeder reactors that use uranium-233 as their fissile core materials while the core is surrounded with a blanket containing thorium to be converted to uranium-233 using neutrons from the core.
The progress of this program has been slow. An experimental plant, the fast breeder test reactor (FBTR), was approved in 1971 but only completed in 1985 and did not begin to generate steam until 1993. The FBTR has experienced several accidents including a sodium coolant leak in 2002. Alongside the FBTR, design work also began on a protoype fast breeder reactor with a generating capacity of 500 MW. Work on this began in 2004. It will have a uranium-plutonium oxide core and this will be surrounded by a blanket of thorium, producing fissile uranium-233. Plutonium content in the core will be 21–27%. The unit was supposed to achieve criticality in 2015 but was still not connected to the grid, 1 year later.