A nuclear reactor is a device in which a controlled nuclear fission chain reaction takes place. The fission reaction is initiated by the absorption of a neutron in a heavy nucleus such as uranium-235 (U-235). The process produces additional neutrons that can be used to induce further fissions, thereby propagating the chain reaction. When the reactor materials are appropriately adjusted, it is possible for the chain reaction to be self-sustaining. Such a reactor is called "critical." If there are insufficient neutrons being produced to sustain the process, then the reactor is "subcritical." Conversely, if too many neutrons are being produced, the reaction rate increases with time and the reactor is called "supercritical" (see \Tnuclear energy\t). Nuclear reactors are most commonly used to produce electric energy, although they are occasionally used as sources of thermal energy for heating. They are also designed as sources of neutrons used in research, or for the transmutation of elements. Reactors designed to produce materials for nuclear weapons by transmutation are called production reactors. Numerous devices use nuclear processes other than fission as their energy source, although these devices are not called nuclear reactors. For instance, power supplies on spacecraft use the energy from radioactive decay, and hence are called "radioisotope power supplies." Similarly, devices based on the fusion process are called "thermonuclear" or "fusion" reactors. FISSION CREATION AND CONTROL The energy released in the fission process takes several forms. Almost 85% of the energy produced is kinetic energy of the fission fragments. About 3% appears as the kinetic energy of the neutrons released, and another 3% as gamma-ray energy. All of the energy from these sources is released immediately and can be recovered from the reactor. A small amount of energy, about 5%, is carried away by neutrinos, which do not interact readily with matter. This fraction of the energy is lost from the reactor. Finally, about 6% of the total energy is obtained from the decay of radioactive fission fragments. This delayed source of energy plays a significant role in the safety of nuclear reactors. Long after the fission process has been shut off, the inventory of accumulated fission products will continue to produce energy. It is essential to cool the reactor to prevent overheating. LIGHT WATER REACTORS The typical U.S. power reactor is termed a light-water reactor (LWR) because it uses water in the form of H(2)0 as a moderator and coolant. Another type of power reactor uses a type of water in which the hydrogen has been replaced by deuterium --D(2)0--as a moderator; it is called a heavy-water reactor. The design characteristics stem from a set of physical, engineering, and economic constraints. The physical aspects of the design seek to provide the nuclear fuel and other constituents so that a safe and controllable reactor can operate at the desired power level for an extended period of time. The engineering aspects of the design seek to convert the fission energy into a useful form of heat, usually high-pressure steam, to drive a turbine which is connected to an electric generator. The economic aspects of the design seek to optimize the physical and engineering design so as to minimize the cost of energy from the plant. Typically, it is possible to provide a range of options in design that are satisfactory from the technical view. The final choices rest with the relative economics of the various options. Reactor physicists are primarily concerned with finding means to promote the fission reaction so as to keep the reactor critical. This implies a careful balance between the neutron production rate (the fission rate) and the neutron loss rate. Neutrons are lost via two mechanisms: they may be captured by nuclei that do not fission, or they may simply migrate out of the region containing nuclear fuel. The core of a reactor is the region that contains the nuclear fuel. Neutrons from the fission process are born with relatively high energy. However, the probability of a neutron causing a fission in the fuel nuclei is much larger for low energy neutrons than for high energy neutrons. In order to slow neutrons down, it is common to surround the fuel with a moderator. Neutrons can interact with nuclei much like collisions between hard spheres. The neutron will lose energy most efficiently, i.e., in fewest collisions, if the moderator nuclei are close to the mass of the neutron. Thus, moderators are made from light materials such as hydrogen in water, deuterium in heavy water, or carbon in graphite. The physical arrangement of the fuel and moderator is a major element of reactor physics. The LWR uses H(2)0 as the moderator and uranium dioxide, U0(2), as the fuel. The fissionable isotope of uranium is U-235, which makes up only 0.7% of natural uranium. It is not possible to design a critical reactor using natural uranium. In order to increase neutron production, the U-235 concentration in the fuel is increased. Such fuel is called "enriched." Fuel for an LWR has a relatively simple structure. Uranium is pressed into small cylindrical pellets that are stacked in zirconium alloy tubes--the "cladding"--about 3.05 m (10 ft) in length. The tubes are arranged in a "fuel assembly," a square array containing about 17 tubes on a side. A modern pressurized water reactor has about 200 fuel assemblies in its core. Reactor control is achieved by carefully balancing the neutron production rate by fission with the neutron loss rate. The common process for obtaining control is to adjust the amount of neutron absorber in the core. Control materials are placed in rods with the same dimensions as fuel rods and the set of control rods are inserted in the middle of a fuel assembly. The control rods are attached to a drive mechanism that moves the control rods into or out of the core region. A typical set of control rods contains materials that are highly absorbent to neutrons such as silver, indium, and cadmium. The control rods are inserted into the core when reactor shutdown is desired. The rods are also inserted automatically in the event that unexpected conditions are detected. The core, including fuel assemblies, control rods, and moderator, is a very large system on the order of 3.65 m (12 ft) diameter and 3.65 m high. The entire assemblage fits into a 12-m-high (40 ft), thick-walled steel pressure vessel, designed to withstand very high pressures, up to 2,500 psi. For LWRs, water is both the moderator and the coolant, that is, the agent used to remove fission energy from the core and transfer it to the electric generating segment of the system. In the pressurized water reactor (PWR)--a type of LWR--water is heated to a high temperature without boiling, by keeping the system under very high pressure. Water is piped into the pressure vessel and flows down the vessel sides to a region below the core. It then flows up through the core, gaining heat while keeping the core cooled. The heated water flows through pipes to a steam generator. The sets of pipes and associate pumps are called "loops." Typically, a large PWR has 4 loops. In the primary loop, hot water from the core flows through a steam generator, where it exchanges heat with water in a second loop. Second-loop water, under lower pressure, will boil and produce steam, which in turn is fed to turbines to drive a generator. The primary water returning from the steam generator is pumped through the reactor core again. In passing through the core region, the primary water is subject to neutron irradiation, and some of its constituents will become radioactive. Further, small amounts of activity are scrubbed off the cladding material and other internal parts of the reactor. As a result, primary water is contaminated with radioactivity. It is very important that this water be kept apart from the plant operating staff. The shielding of the entire vessel and its internals, the coolant loops, and steam generator is accomplished by keeping all of these components inside a very large concrete shell called the containment building. The large spheres of containment buildings are characteristic features of nuclear power plants. The steam from the secondary side of the steam generator is not radioactive; it flows through the containment walls to an auxiliary building that houses the turbine and generator. The turbine, generator, condenser, pumping equipment, and transformers make up the balance of the plant, and are like those of a conventional energy plant. OTHER REACTOR TYPES The pressurized water reactor is the most common type of nuclear reactor used for the generation of electric energy. Over one half of all nuclear power reactors in the world are PWRs. The second most common reactor is also moderated and cooled with light water. This is the boiling water reactor (BWR), in which the coolant is permitted to boil within the reactor core. The steam emerging from the core is sent directly to a turbine rather than through a steam generator. The BWR has somewhat less equipment that must be held within the containment building and hence they are slightly smaller than PWR containments. The steam that enters into a BWR turbine is radioactive, however, and slightly contaminates the turbine. There are other small differences in detail between the two types of LWRs, but the major components such as the vessel, fuel, control rods, coolant loops, etc., are quite similar in function. The characteristics of the two types of plants have so much in common that several utilities own plants of both types. Reactors that use various gas coolants, rather than liquids, are collectively called gas-cooled reactors. The first plants of this type were built in Great Britain and France. The commercial versions used C0(2) as a coolant and graphite for moderation. The designs are significantly different from those of LWRs. The fuel is usually uranium, but it is implaced in steel tubes that are in turn imbedded in graphite blocks. There are channels through the graphite for the coolant gases to pass. Typically the gaseous coolant can be heated to a higher temperature in a graphite reactor than one containing structural metals such as steel or zirconium. The heat is again transferred through a steam generator to produce steam to drive a turbine. Because the coolant is at a high temperature the quality of steam produced in the steam generator is very high. Overall, gas cooled reactors tend to have a higher thermodynamic efficiency than LWRs. In spite of the thermodynamic advantages of the gas reactors there were other problems, including economic ones, that led to the economic demise of these plants. One major problem was the corrosion the hot gases caused in various surfaces of the reactor and steam generator. A second factor was the cost of constructing such systems. They tended to be very large in volume, requiring a great deal of material and high labor costs. In the early 1970s the French shifted their emphasis to the LWR technology. Great Britain made a similar decision in the mid 1980s. A different type of gas-cooled reactor has been under development in Germany and the United States and is called the high temperature gas-cooled reactor, or HTGR. The significant change from earlier gas reactors is the adoption of helium as the coolant gas. Helium is chemically inert, and as a result causes little corrosion. The fuel has the uranium imbedded in graphite rather than steel, thus the system can operate at very high temperatures, which promotes thermodynamic efficiency. A few HTGRs have been built to demonstrate the technology. The economics of large plants has not been demonstrated. There is a move toward building small HTGRs, called modular HTGR or MHTGR. It has been suggested that a small plant can be manufactured in a fabrication plant rather than in the field. Simplified installation of such systems should reduce costs. Further, if the module is small enough there will be important safety advantages to the system. Graphite has a very high heat capacity and melts at extremely high temperatures. If the decay heat stored in a core is sufficiently small, it is possible for an HTGR to withstand any accident that prevents coolant from reaching the core. Such a reactor may be genuinely "walkaway safe," i.e., no human intervention would be required to prevent an accident that could release radioactivity. A final argument made for the modular reactor is that the total plant size can be increased in increments as the power demand increases. Thus, the financial commitments can be kept small if the electric demand stops growing. The few HTGRs built to date are not sufficient to demonstrate the economic or safety advantages anticipated for the MHTGR. All the reactors mentioned above are known as thermal reactors because the moderation slows the neutrons down to reach thermal equilibrium with the moderator. An entirely separate class of reactors are the "fast" reactors, in which there is no effort to moderate neutron energy. The chance of a fast neutron causing a fission is much lower than that of a thermal neutron. However, if fast neutrons do cause a fission, more neutrons emerge, and in a carefully designed system, the excess number of neutrons can be greater than one for each fission. The excess neutrons can be used to transmute nonfissionable species into fissionable species. The most common example is the transmutation of uranium 238 into plutonium 239. Under appropriate conditions it is possible to produce more fissionable material than is consumed in operating the reactor. Thus, it is possible to breed fuel at a faster rate than it is being consumed. Such reactors are called "breeders", and they have been built and demonstrated in the United States, Great Britain, France, Germany, the Soviet Union, and Japan. The largest such plant is a 1200 MW commercial plant in France called "Superphoenix" (see \Tbreeder reactor\t). The breeders demonstrated to date have used liquid sodium as a coolant. The plants are called liquid-metal cooled fast breeder reactors, or LMFBRs. The core is made of uranium and plutonium fuel rods, and is surrounded by arrays of rods containing U-238 to be converted into plutonium. This region of the reactor is termed the "blanket." Liquid sodium is relatively heavy and does not moderate neutrons very much. Further, sodium is an excellent heat-transfer medium. Results thus far show little corrosion caused by the coolant in operating components such as piping, pumps, or valves. In spite of their advantages LMFBRs have not been economically successful. Part of the reason is that the manufacturing, processing, and handling of plutonium-bearing materials is very costly. Further, the systems tend to be large and their construction requires high material and labor costs. The design is very interesting for the future should uranium ultimately become scarce. In addition, some recent safety experiments have shown these plants to have significant safety features that may make them easy to license and operate. NUCLEAR SAFETY As the reactor operates, a large inventory of radioactive isotopes accumulates and represents a unique hazard. Any accident releasing a sizeable portion of these materials would be very serious--as demonstrated by the \TChernobyl\t accident. A fundamental objective of nuclear reactor design is to prevent accidents that could allow the escape of \Tradioactivity\t. In order for fission products to reach the environment several barriers must be overcome. For an LWR, the first barrier is the fuel cladding that contains the fuel as well as the fission fragments. The cladding material is a high-strength alloy of zirconium capable of withstanding high pressures and high temperatures, well beyond normal reactor conditions. The second barrier is the pressure vessel, which is exceedingly strong, but which does have numerous penetrations for the cooling water to enter and exit the vessel. The third barrier is the containment building, which is a large, reinforced concrete building designed to withstand substantial pressure. In order for any barrier to be breached, the system must first become overheated. There are two possible ways for this to occur. The first is for the fission rate to grow too rapidly for the coolant to remove all of the energy being created. The second is for the coolant system to fail and lose the ability to cool the fuel. Excessive fission energy production is monitored by numerous sensors throughout the core region; if they detect a rapid rate of growth in the fission process, the control rods are automatically lowered into the core to absorb the fission products. The reactor shuts down. There is another important design element that protects against possible nuclear runaways. The heating up of a local region of the reactor would cause the nearby cooling water to boil, thereby reducing the water density through the creation of bubbles, or voids. It is a safety requirement of U.S. reactors that creation of coolant voids must, by itself, reduce the fission rate. The amount of water that surrounds a fuel rod is carefully adjusted so that a void reduces neutron moderation and hence reduces the fission rate. This property is not required in the Soviet Union and is the fundamental reason for the disastrous Chernobyl accident. The basic design of LWRs makes them safe against nuclear runaways. It is not physically possible for a light-water reactor to undergo a rapid power excursion, thus an LWR cannot explode like an atom bomb. The greatest threat to reactor safety is the loss of coolant accident, or \TLoca\t. The fission process itself ceases if a reactor loses its cooling water because the reactor goes subcritical. However, the fuel continues to heat up due to the stored thermal energy as well as from the decay heat of radioactive fission products. Without any coolant the cladding heats up and ultimately melts. Safety systems have been designed and installed in plants to prevent the clad from overheating by providing emergency cooling water. Such systems are collectively known as emergency core cooling systems, or ECCs. All such systems have multiply redundant pathways for introducing water into the vessel under high-pressure or low-pressure conditions. The state of the system will depend upon the exact cause of the loss of coolant. In the event that one of the large coolant pipes ruptures, the system would depressurize rapidly and emergency water could be injected under relatively low pressure. Conversely, if a very small line broke, or there was a small leak in a large line, the system might depressurize slowly. Thus, it is necessary to have the ability to provide emergency coolant to the system while at high-pressure. The design of safety systems begins by hypothesizing a number of different failures and then developing systems to mitigate the consequences of these failures. Such failures are known as design basis accidents. In order to obtain a license, a plant must show it is protected against the class of design basis accidents. The broad areas of concern include accidents within the plant as well as accidents involving the handling of radioactive spent fuel. Initiating events must include hardware failures, operator failures, and external events such as tornadoes. REACTOR ACCIDENTS In spite of the installation of safety systems, there have been notable accidents at nuclear plants. The first serious nuclear accident occurred (1957) in a British weapons production plant, Windscale 1, an air-cooled graphite reactor. During heat-up operations that were conducted to heal defects in the graphite, it caught on fire and melted some of the fuel cladding. Volatile fission products, most notably iodine and cesium, were released into the environment around the plant. The Windscale plant and its twin unit were shut down and taken out of service. Three Mile Island The most serious U.S. commercial reactor failure occurred on Mar. 28, 1979, at the Three Mile Island (TMI) reactor near Harrisburg, Pa. The TMI-2 accident began as a small break \TLoca\t in which a valve stuck open, allowing coolant to escape from the vessel. The emergency core cooling system (ECCS) operated as designed and provided makeup water for the core. Unfortunately, the operators misinterpreted the information available to them in the control room and shut off the ECCS for several hours. The decay heat from the core boiled off the available water in the vessel, and without adequate cooling, the cladding and fuel started to melt. Before the operators resumed the flow of emergency coolant, a sizeable portion of the core, about one-half to one-third, melted. The molten fuel and cladding dropped into the bottom of the vessel, which was full of water. This water was adequate to quench the molten material. The vessel itself maintained its integrity and kept all of the debris contained. A sizable amount of gaseous fission products escaped from the vessel through the open valve into the containment building. The containment functioned as the ultimate barrier and prevented a release into the local environment. The small amount of activity that did escape was carried by coolant water that leaked out the valve into the containment and then overflowed into an auxiliary building where the gases leaked into the environment. The releases were almost entirely noble gases (such as xenon), which are chemically inert and not retained within the human body. The health effects of the accident proved to be quite small, and virtually undetectable against the normal incidence of background radiation. Chernobyl The accident at Chernobyl Unit 3 in in the \TUSSR\t in April 1986 has been the most serious of all nuclear accidents to date. The reactor involved was a water-cooled, graphite moderated reactor, which is used simultaneously for power and plutonium production. The fuel is contained in fuel rods which reside in pipes through which the coolant flows. The design is known as the RBMK. As water in the coolant pipe begins to boil, the effect on the fission rate is positive rather than negative (in contrast to the LWR). The boiling creates a void that allows an increase in the number of neutrons that escape into the graphite moderator and become thermalized. When the thermal neutrons return toward the fuel they encounter less neutron-absorbing hydrogen. As a result, a void will lead to an increase in the heating of the water, which in turn creates more voids, and a self-reinforcing nuclear runaway begins. This characteristic of the RBMK was known to nuclear analysts in the Soviet Union as well as the West. The design is not licensable in the Western nations. The Soviet Union accepted the risks because they did not have the technology to build large pressure vessels until the 1960s. One of the design basis accidents analyzed at all nuclear plants is the "station blackout." The name refers to the loss of an independent electric energy supply to a nuclear plant. Normally, a plant uses large amounts of electricity to power motors, valves, fans, etc., within the plant itself. For a large plant this requirement might be as large as 5% of the plant generating capacity. The output of the plants own generator can be used during normal operations to provide this electricity. However, when the reactor shuts down, an independent source of electricity is required. A station blackout refers to the loss of the independent supply. All plants are required to have auxiliary systems, such as batteries or diesel generators, to protect the plant in case of a station blackout. It is also a requirement that a reactor must shut down if there is a loss of off-site power, in order to reduce total plant electric energy needs. The accident at Chernobyl occurred while plant operators were conducting an experiment to see how long they could extract electricity from a generator as a reactor shut down. The operators knew that instrumentation would detect a loss of electric supply to coolant pumps and would automatically start the plant emergency core cooling system. To prevent this from happening and ruining the test, the operators disconnected the ECCS, in violation of standing safety rules. Further, the operators wanted to start the test from low-power conditions, but had trouble adjusting the power level appropriately. To assist in maneuvering the power level they withdrew more control rods than would be allowed during normal operations. By the time the test had begun, the control rods were almost useless. They had a long distance to travel before they could affect the fission rate, and the mechanical design was such that it took them too long to travel into the core and stop the accident. The operators began the test by stopping the flow of steam to the turbine. Without removal of energy the coolant temperature increased to the point where boiling began and voids formed in the coolant channel. The fission rate then accelerated, producing more energy and heating the coolant further. The operators detected the power excursion, but were unable to insert control rods in time to prevent a massive surge of energy that ruptured the fuel rod cladding and distorted the fuel channels. The ECCS was unavailable to cool the core and reduce the accident consequences. Some uncertainty remains as to the sequence of events that led to the reactor catching on fire. The generally accepted scenario is that molten fuel came in contact with coolant water and reacted to generate huge volumes of steam that ruptured the piping. Subsequently, a second explosion occurred due to a chemical reaction of the incoming water and the hot metals and graphite in the core. The combination of events broke the barriers containing the fission products and allowed them to escape into the environment. The RBMK design does not include a Western-style containment building, thus once the fission products escaped from the cooling system they had easy access to the environment. About one-half of the radioactivity released emerged in the first few hours of the accident. The remainder escaped over the next ten days as the reactor burned. Only when the core was smothered in sand and various neutron-absorbing materials did the release cease. Post-accident calculations indicate that a western containment building would have prevented much of the large release of radioactivity. Kent F. Hansen Bibliography: American Nuclear Society, The Safety of Next-Generation Power Reactors (1988); Bennet, D.J., and Thomson, J.R., The Elements of Nuclear Power, 3d ed. (1990); Johnson, J.W., Insuring against Disaster (1986); Knief, Ronald N., Nuclear Energy Technology (1988); Marples, David R., Chernobyl and Nuclear Power in the \TUSSR\t (1986); Smyth, Henry D., Atomic Energy for Military Purposes, new, encl. ed. (1989).