The pebble bed reactor (PBR) is a graphite-moderated, gas-cooled, nuclear reactor. It is a type of Very high temperature reactor (VHTR) [formally known as the high temperature gas reactor (HTGR)], one of the six classes of nuclear reactors in the Generation IV initiative. Like other VHTR designs, the PBR uses TRISO fuel particles, which allows for high outlet temperatures and passive safety.
The base of the PBR's unique design is the spherical fuel elements called "pebbles". These tennis ball-sized pebbles are made of pyrolytic graphite (which acts as the moderator), and they contain thousands of micro fuel particles called TRISO particles. These TRISO fuel particles consist of a fissile material (such as 235U) surrounded by a coated ceramic layer of SiC for structural integrity. In the PBR, 360,000 pebbles are placed together to create a reactor, and is cooled by an inert or semi-inert gas such as helium, nitrogen or carbon dioxide.
This type of reactor is also unique because its passive safety removes the need for redundant, active safety systems. Because the reactor is designed to handle high temperatures, it can cool by natural circulation and still remain intact in accident scenarios, which may raise the temperature of the reactor to 1600 oC. Also because of its design, its high temperatures allow higher thermal efficiencies than possible in traditional nuclear power plants (up to 50%).
Also, the gases do not dissolve contaminants or absorb neutrons as water does, so the core has less in the way of radioactive fluids.
The technology was first developed in Germany but political and economic decisions were made to abandon the technology. In various forms, it is currently under development by MIT, the South African company PBMR, General Atomics (U.S.), the Dutch company Romawa B.V., Adams Atomic Engines , Idaho National Laboratory, and the Chinese company Huaneng .
In June 2004, it was announced that a new PBMR would be built at Koeberg, South Africa by Eskom, the government-owned electrical utility. There is considerable opposition to the PBMR from groups such as Koeberg Alert and Earthlife Africa, the latter of which has sued Eskom to stop development of the project.
The uranium, thorium or plutonium nuclear fuels are in the form of a ceramic (usually oxides or carbides) contained within spherical pebbles made of pyrolytic graphite, which acts as the primary neutron moderator. Each sphere is effectively a complete "mini-reactor", containing all of the parts that would normally be separate components of a conventional reactor. Simply piling enough of the fuel spheres together will eventually reach criticality.
The reactor design is such that it is power-limited or inherently self controlling due to Doppler broadening.
The pebbles are held in a bin or can. An inert gas, helium, nitrogen or carbon dioxide, circulates through the spaces between the fuel pebbles to carry heat away from the reactor. Ideally, the heated gas is run directly through a turbine. However, if the gas from the primary coolant can be made radioactive by the neutrons in the reactor, it may be instead brought to a heat exchanger, where it heats another gas, or produces steam. The exhaust of the turbine is quite warm and may be used to warm buildings or chemical plants, or even run another heat engine.
Much of the cost of a conventional, water-cooled nuclear power plant is due to cooling system complexity. These are part of the safety of the overall design, and thus require extensive safety systems and redundant backups. A water-cooled reactor is generally dwarfed by the cooling systems attached to it. Additionally, the core irradiates the water with neutrons. The water, and impurities dissolved in it become radioactive. Furthermore, the high pressure piping in the primary side becomes embrittled and requires continual inspection and eventual replacement.
In contrast, a pebble bed reactor is gas cooled, sometimes at low pressures. The spaces between the pebbles form the "piping" in the core. Since there is no piping in the core and the coolant contains no hydrogen, embrittlement of the pipes from neutrons and hydrogen cannot occur. The preferred gas, Helium, does not easily absorb neutrons or impurities. Therefore, compared to water, it is both more efficient and less likely to become radioactive.
A large advantage of the pebble bed reactor over a conventional light-water reactor is that it operates at higher temperatures. The reactor can directly heat fluids for low pressure gas turbines. The high temperatures allow a turbine to extract more mechanical energy from the same amount of thermal energy; therefore, the power system uses less fuel per kilowatt-hour.
A significant technical advantage is that some designs are throttled by temperature, not by control rods. The reactor can be simpler because it does not need to operate well at the varying neutron profiles caused by partially-withdrawn control rods. For maintenance, many designs include control rods, called "absorbers" that are inserted through tubes in a neutron reflector around the reactor core.
If throttled by temperature, the reactor can change power quickly just by changing the coolant flow rate. A coolant-throttled design can also change power more efficiently (say, for utility power) by changing the coolant density or heat capacity.
Another advantage is that fuel pebbles for different fuels might be used in the same basic design of reactor (though perhaps not at the same time). Proponents claim that some kinds of pebble-bed reactors should be able to use thorium, plutonium and natural unenriched uranium, as well as the customary enriched uranium. There is a project in progress to develop pebbles and reactors that use MOX fuel, that mixes uranium with plutonium from either reprocessed fuel rods or decommissioned nuclear weapons.
The concept was invented by Professor Dr. Rudolf Schulten in the 1950s. The basic concept was to make a very simple, very safe reactor, with a commoditized nuclear fuel. The crucial breakthrough was the idea of combining fuel, structure, containment, and neutron moderator in a small, strong sphere. The concept was enabled by the realization that engineered forms of silicon carbide and pyrolytic carbon were quite strong, even at temperatures as high as 2000 °C (3600 °F). The natural geometry of close-packed spheres then provides the ducting (the spaces between the spheres) and spacing for the reactor core. To make the safety simple, the core has a low power density, about 1/30 the power density of a light water reactor.
The core generates less power as its temperature rises, and therefore cannot have a criticality excursion when the machinery fails. At such low power densities, the reactor can be designed to lose more heat through its walls than it would generate. In order to generate much power it has to be cooled, and then the power is extracted from the coolant.
The "modular" concept of the pebble bed reactor uses several small reactors in a large power plant. This is convenient because new investment can be gradual, and tuned to the actual demand for electric power. Sites that require larger generation capacity can simply install more reactors. Depending on the design, there also can be economies of scale and better reliability when several reactors share equipment, and can switch sets of equipment when some part fails.
The modular design also allows a small reactor to be mass-produced, reducing the life-cycle costs of safety-certification and design qualification.
In modular systems, the equipment to cool the turbine's exhaust must be adapted to the site. The cooling equipment adaptable to the most sites is a cooling tower. However, near water, water cooling is far less expensive because the larger heat capacity of water permits the equipment to be much smaller.
The AVR was originally designed to breed 233Uranium from 232Thorium. 232Thorium is about 400 times as abundant in the Earth's crust as 235Uranium, and an effective thorium breeder reactor is therefore considered valuable technology. However, the fuel design of the AVR contained the fuel so well that the transmuted fuels were uneconomic to extract—it was cheaper to simply use natural uranium isotopes.
The AVR used helium coolant. Helium has a low neutron cross-section. Since few neutrons are absorbed, the coolant remains less radioactive. In fact, it is practical to route the primary coolant directly to power generation turbines. Even though the power generation used primary coolant, it is reported that the AVR exposed its personnel to less than 1/5 as much radiation as a typical light water reactor.
Tsinghua's program for Nuclear and New Energy technology also plans in 2006 to begin developing a system to use the high temperature gas of a pebble bed reactor to crack steam to produce hydrogen. The hydrogen could serve as fuel for hydrogen vehicles, reducing China's dependence on imported oil. Hydrogen can also be stored, unlike electricity, and distribution by pipelines may be more efficient than conventional power lines. See hydrogen economy.
PBMR's primary coolant is helium. The helium directly turns low-pressure turbo-machinery, without intervening losses from heat-exchangers. Helium is well-favored because it is chemically inert, and neutrons do not transmute it to a radioactive element. This means that the turbo-machinery should not become radioactive, even though it operates on primary coolant. While the fuel design is robust, a fuel defect could still contaminate the power production equipment. One disadvantage is that the turbine must be somewhat larger, and therefore more expensive.
The turbine's compressors are decoupled from the turbine, which permits the turbine's pressurization to be decoupled from the generator speed. Utility generators must be synchronized to the power grid. The prototype test of the closed-cycle helium system including compressors, turbine and recuperator has been developed in the engineering lab at the Potchefstroom Campus of the North-West University.
Helium is lighter than air, so air can displace the helium if the reactor wall is breached. Pebble bed reactors need fire-prevention features to keep the graphite of the pebbles from burning in the presence of air.
The pebble bed reactor's design can be throttled in real time to meet peak electric power loads just like conventional reactors, where power follows steam demand in seconds. The modular design also supports the speculation that it will be useful in building peak load plants. South Africa lacks natural gas for the gas turbines that normally power peak loads, but it exports uranium and thorium.
The S. African module's capacity is 165 MWe. The reactor could be a significant export item for South Africa.
An inherently self controlling modular reactor that can provide peaking-power and fresh water would be a genuinely useful addition to the market, and a valuable export item. If the trial is successful, PBMR says it will build up to ten local plants on South Africa's coast. PBMR also wants to export up to 20 plants per year. The estimated export revenue is 8 billion rand (roughly US$ 1.1 billion) per year, and could employ about 57,000 people. The program's total cost is about US$ 1 billion, and the developers estimate that about 30 plants will need to be produced to break even.
In 2005, environmental group Earthlife Africa won a court challenge requiring further hearings on the Koeberg reactors (which were originally approved in September 2003) The Cape Town city government and other civic and environmental groups also say they oppose the plant. In July 2003, following the approval of the environmental impact assessment, there were public demonstrations against the project in both Johannesburg and Cape Town. Earthlife Africa also opposed the Pelindaba fuel plant.
In December 2005, South Africa's PBMR company awarded a contract for engineering, procurement and construction management to SLMR - a Canadian-South African joint venture made up of Montreal-based engineering firm SNC-Lavalin and South-African construction and engineering firm Murray & Roberts - for its demonstration Pebble Bed Modular Reactor at Koeberg. Construction is envisaged starting 2007, and a second round of environmental hearings is under way at present. Meanwhile the BNFL share in PBMR has been passed to Westinghouse Electric Company and negotiations are under way with other possible investors to enable Eskom (the South African Power Utility) to reduce its stake from 30% to 5%.
On 30 January 2007 it was reported that the South African government had approved the manufacture of PBMR fuel at Nuclear Energy Corporation of South Africa's Pelindaba Beva complex in the North West Province, and transporting of the raw materials to this site and manufactured fuel from it to Koeberg.
This followed the dismissal by Environment Minister Marthinus van Schalkwyk of appeals brought by Earthlife Africa including opposition to the de-linking of the fuel plant and the PBMR. The appeal claimed that "neither process should be viewed in isolation". The appellants also registered concern about the long-term storage of high-level radioactive waste and contaminated materials, and alleged inadequate consideration of alternatives to the fuel plant.
In dismissing the appeals, van Schalkwyk noted that the two projects would be established in different places, were of different natures and came with "vastly different" environmental risks. He added that "negative environmental impacts ... can be sufficiently mitigated, provided the conditions contained in this record of decision are implemented and adhered to."
Romawa's reactor heats helium, which in turn heats air that drives a conventional gas turbine. The Romawa design reduces the size and expense of heat exchangers. The main heat exchangers, the reactor and air-heater, operate at very high temperatures, and should therefore be small, inexpensive and efficient. The design exhausts the air from the turbine, avoiding the large, inefficient, expensive low-temperature heat exchanger that would otherwise be necessary to cool the turbine's exhaust. On a conventional light water reactor the analogous item is the steam condenser, the largest part of a light water reactor—the big cooling tower. The exhausted air is not radioactive because it never passed through the reactor's core.
Additionally, gas turbines designed for air are well-developed for the aircraft and stationary power industries.
Romawa proposes two types of throttling. For vehicular power, they advocate a reliable, quick-acting, inexpensive valve between the turbine and reactor. For efficient utility-style throttling, they advocate a system that reduces the pressure of helium in the coolant loop that connects the reactor to the turbine.
The basic design is at least as safe as a light water reactor, because only the helium passes through the reactor. The design attempts to reproduce the very safe operational experience of the AVR by using helium as the primary coolant.
The air passing through the turbine never passes through the reactor, and is therefore never exposed to neutron flux, and therefore particles and gasses cannot become radioactive. The turbine is likewise not part of the primary loop, and uses air as its working fluid. The technology is therefore very standard. Most moving parts do not touch the primary loop, and therefore service should be relatively easy and safe.
Romawa also proposes a refueling and maintenance plan, based on "pool service." Users of large gas turbines customarily pool their repair resources to minimize expensive equipment, spares and training. By shipping entire reactors, Romawa plans to eliminate on-site service, and provide all service in one or a few centralized, specialized workshops.
Romawa has neither produced nor is licensed to produce a nuclear reactor at this time.
Romawa has a business agreement with Adams Atomic Engines in the U.S., which promotes a similar reactor system.
AAE holds the U.S. patent on direct throttling of a turbine heated by a pebble-bed reactor. Adams Atomic Engines has neither produced nor is licensed to produce a nuclear reactor at this time.
When the fuel heats, the 238U reacts with a broader spectrum of neutron speeds, thereby lowering the number of available neutrons for fission with 235U. A slower fission rate generally lowers the temperature of the fuel. 238U tends to absorb instead of fission, thus contributing a negligible amount of energy. This places a natural limit on the power produced by the reactor. The reactor vessel is designed so that without mechanical aids it loses more heat than the reactor can generate in this idle state. The design adapts well to safety features (see below). In particular, most of the fuel containment resides in the pebbles, and the pebbles are designed so that a containment failure releases at most a 0.5 mm sphere of radioactive material.
The reactor is cooled by an inert, fireproof gas, so it cannot have a steam explosion as a light-water reactor can. The coolant has no phase transitions—it starts as a gas and remains a gas.
The moderator is solid carbon. It does not act as a coolant, move, or have phase transitions (i.e. between liquid and gas) as the light water in conventional reactors does.
A pebble-bed reactor thus can have all of its supporting machinery fail, and the reactor will not crack, melt, explode or spew hazardous wastes. It simply goes up to a designed "idle" temperature, and stays there. In that state, the reactor vessel radiates heat, but the vessel and fuel spheres remain intact and undamaged. The machinery can be repaired or the fuel can be removed.
These safety features were tested (and filmed) with the German AVR reactor.. All the control rods were removed, and the coolant flow was halted. Afterward, the fuel balls were sampled and examined for damage. There was none.
PBRs are intentionally operated above the 250 °C annealing temperature of graphite, so that Wigner energy is not accumulated. This solves a problem discovered in a famous accident, the Windscale fire. One of the reactors at the Windscale site in England (not a PBR) caught fire because of the release of energy stored as crystalline dislocations (Wigner energy) in the graphite. The dislocations are caused by neutron passage through the graphite. At Windscale, a program of regular annealing was put in place to release accumulated Wigner energy, but since the effect was not anticipated during the construction of the reactor, the process could not be reliably controlled and led to a fire.
The continuous refueling means that there is no excess reactivity in the core. Continuous refueling also permits continuous inspection of the fuel elements.
The hollow contains fifteen thousand small "seeds" with further containment layers. Each seed surrounds a sand-grain-sized (0.5 mm) kernel of fissionables. Breaking the fissionables into pebbles, and pebbles into seeds assures that the maximum release by a cascade of containment failures will be small—at most the fissionables in one seed.
Each seed, from the inside out, consists of:
Pyrolytic graphite is the main structural material in these pebbles. It sublimes at 4000 °C, more than twice the design temperature of most reactors. It slows neutrons very effectively, is strong, inexpensive, and has a long history of use in reactors. Its strength and hardness come from anisotropic crystals of carbon. Pyrolytic graphite is also used, unreinforced, to construct missile reentry nose-cones and large solid rocket nozzles. It is nothing like the powdered mixture of flakes and waxes in pencil leads or lubricants.
Pyrolytic carbon can burn in air when the reaction is catalyzed by a hydroxyl radical (e.g. from water). Some famous examples include the accidents at Windscale and Chernobyl—both graphite-moderated reactors. Some engineers insist that pyrolytic carbon cannot burn in air, and cite engineering studies of high-density pyrolytic carbon in which water is conspicuously excluded from the test. Of course, water is present in real environments. However, all pebble-bed reactors are cooled by inert gases to prevent fire. All pebble designs also have at least one layer of silicon carbide that serves as a fire break, as well as a seal.
The fissionables are also stable oxides or carbides of uranium, plutonium or thorium which have higher melting points than the metals. The oxides cannot burn in oxygen, but have some potential to react via diffusion with graphite at sufficiently high temperatures; the carbides might burn in oxygen but cannot react with graphite. The fission materials are about the size of a sand grain, so they are too heavy to be dispersed in the smoke of a fire.
The layer of porous pyrolytic graphite right next to the fissionable ceramic absorbs the radioactive gases (mostly xenon) emitted when the heavy elements split. Most reaction products remain metals, and reoxidize. A secondary benefit is that the gaseous fission products remain in the reactor to contribute their energy. The low density layer of graphite is surrounded by a higher-density nonporous layer of pyrolytic graphite. This is another mechanical containment. The outer layer of each seed is surrounded by silicon carbide. The silicon carbide is nonporous, mechanically strong, very hard, and also cannot burn.
Many authorities consider that pebbled radioactive waste is stable enough that it can be safely disposed of in geological storage. Thus used fuel pebbles could just be transported to disposal.
All kernels are precipitated from a sol-gel, then washed, dried and calcined. U.S. kernels use uranium carbide, while German (AVR) kernels use uranium dioxide.
The precipitation of the pyrolytic graphite is by a mixture of argon, propylene and acetylene in a fluidized-bed coater at about 1275 °C. The fluidized bed moves gas up through the bed of particles, "floating" them against gravity. The high-density pyrolytic carbon uses less propylene than the porous gas-absorbing carbon. German particles are produced in a continuous process, from ultra-pure ingredients at higher temperatures and concentrations. U.S. coatings are produced in a batch process. Although the German carbon coatings are more porous, they are also more isotropic (same properties in all directions), and resist cracking better than the denser U.S. coatings.
The silicon carbide coating is precipitated from a mixture of hydrogen and methyltrichlorosilane. Again, the German process is continuous, while the U.S. process is batch-oriented. The more porous German pyrolytic carbon actually causes stronger bonding with the silicon carbide coat. The faster German coating process causes smaller, equiaxial grains in the silicon carbide. Therefore, it may be both less porous and less brittle.
Some experimental fuels plan to replace the silicon carbide with zirconium carbide to run at higher temperatures.
Some designs for pebble bed reactors lack a containment building, potentially making such reactors more vulnerable to outside attack and allowing radioactive material to spread in the case of an explosion. However, the current emphasis on reactor safety means that any new design will likely have a strong reinforced concrete containment structure . Also, any explosion would most likely be caused by an external factor, as the design does not suffer from the steam explosion-vulnerability of water-cooled reactors.
There is also significantly less experience with production scale Pebble Bed Reactors than Light Water Reactors. As such, claims made by both proponents and detractors are more theory-based than based on practical experience.
Since the fuel is contained in graphite pebbles, the volume of radioactive waste is much greater, but contains about the same radioactivity when measured in becquerels per kilowatt-hour. The waste tends to be less hazardous and simpler to handle. Current US legislation requires all waste to be safely contained, therefore pebble bed reactors would increase existing storage problems. Defects in the production of pebbles may also cause problems. The radioactive waste must either be safely stored for many human generations, typically in a deep geological repository, reprocessed, transmuted in a different type of reactor, or disposed of by some other alternative method yet to be devised. The graphite pebbles are more difficult to reprocess due to their construction, which is not true of the fuel from other types of reactors. Proponents point out that this is a plus, as it is difficult to re-use pebble bed reactor waste for nuclear weapons.
Critics also often point out an accident in Germany in 1986, which involved a jammed pebble damaged by the reactor operators when they were attempting to dislodge it from a feeder tube. This accident released radiation into the surrounding area, and led to a shutdown of the research program by the West German government.