||This article needs additional citations for verification. (February 2008)|
A loss-of-coolant accident (LOCA) is a mode of failure for a nuclear reactor; if not managed effectively, the results of a LOCA could result in reactor core damage. Each nuclear plant's emergency core cooling system (ECCS) exists specifically to deal with a LOCA.
Nuclear reactors generate heat internally; to remove this heat and convert it into useful electrical power, a coolant system is used. If this coolant flow is reduced, or lost altogether, the nuclear reactor's emergency shutdown system is designed to stop the fission chain reaction. However, due to radioactive decay the nuclear fuel will continue to generate a significant amount of heat. The decay heat produced by a reactor shutdown from full power is initially equivalent to about 5 to 6% of the thermal rating of the reactor. If all of the independent cooling trains of the ECCS fail to operate as designed, this heat can increase the fuel temperature to the point of damaging the reactor.
- If water is present, it may boil, bursting out of its pipes. (For this reason, nuclear power plants are equipped with pressure-operated relief valves and backup supplies of cooling water.)
- If graphite and air are present, the graphite may catch fire, spreading radioactive contamination. This situation exists only in AGRs, RBMKs, Magnox and weapons-production reactors, which use graphite as a neutron moderator. (see Chernobyl disaster.)
- The fuel and reactor internals may melt; if the melted configuration remains critical, the molten mass will continue to generate heat, possibly melting its way down through the bottom of the reactor. Such an event is called a nuclear meltdown and can have severe consequences. The so-called "China syndrome" would be this process taken to an extreme: the molten mass working its way down through the soil to the water table (and below) - however, current understanding and experience of nuclear fission reactions suggests that the molten mass would become too disrupted to carry on heat generation before descending very far; for example, in the Chernobyl disaster the reactor core melted and core material was found in the basement, too widely dispersed to carry on a chain reaction (but still dangerously radioactive).
- Some reactor designs have passive safety features that prevent meltdowns from occurring in these extreme circumstances. The Pebble Bed Reactor, for instance, can withstand extreme temperature transients in its fuel. Another example is the CANDU reactor, which has two large masses of relatively cool, low-pressure water (first is the heavy-water moderator; second is the light-water-filled shield tank) that act as heat sinks. Another example is the Hydrogen Moderated Self-regulating Nuclear Power Module, in which the chemical decomposition of the uranium hydride fuel halts the fission reaction by removing the hydrogen moderator. The same principle is used in TRIGA research reactors.
Under operating conditions, a reactor may passively (that is, in the absence of any control systems) increase or decrease its power output in the event of a LOCA or of voids appearing in its coolant system (by water boiling, for example). This is measured by the coolant void coefficient. Most modern nuclear power plants have a negative void coefficient, indicating that as water turns to steam, power instantly decreases. Two exceptions are the Russian RBMK and the Canadian CANDU. Boiling water reactors, on the other hand, are designed to have steam voids inside the reactor vessel.
Modern reactors are designed to prevent and withstand loss of coolant, regardless of their void coefficient, using various techniques. Some, such as the pebble bed reactor, passively slow down the chain reaction when coolant is lost; others have extensive safety systems to rapidly shut down the chain reaction, and may have extensive passive safety systems (such as a large thermal heat sink around the reactor core, passively-activated backup cooling/condensing systems, or a passively cooled containment structure) that mitigate the risk of further damage.
The Three Final Defenses 
A great deal of work goes into the prevention of a serious core event. If such an event were to occur, three different physical processes are expected to increase the time between the start of the accident and the time when a large release of radioactivity could occur. These three factors would provide additional time to the plant operators in order to mitigate the result of the event:
- The time required for the water to boil away (coolant, moderator). Assuming that at the moment that the accident occurs the reactor will be SCRAMed (immediate and full insertion of all control rods), so reducing the thermal power input and further delaying the boiling.
- The time required for the fuel to melt. After the water has boiled, then the time required for the fuel to reach its melting point will be dictated by the heat input due to decay of fission products, the heat capacity of the fuel and the melting point of the fuel.
- The time required for the molten fuel to breach the primary pressure boundary. The time required for the molten metal of the core to breach the primary pressure boundary (in light water reactors this is the pressure vessel; in CANDU and RBMK reactors this is the array of pressurized fuel channels; in PHWR reactors like Atucha I, it will be a double barrier of channels and the pressure vessel) will depend on temperatures and boundary materials. Whether or not the fuel remains critical in the conditions inside the damaged core or beyond will play a significant role.
See also 
- Containment building
- Nuclear power
- Pressurized water reactor
- Nuclear fuel response to reactor accidents
- Nuclear accidents in the United States
- Nuclear safety in the U.S.
- Nuclear meltdown
- "DOE fundamentals handbook - Decay heat, Nuclear physics and reactor theory, vol. 2, module 4, p. 61". Retrieved 2 August 2009.
- Peterson, Otis G. (2008-03-20). "Patent Application 11/804450: Self-regulating nuclear power module". United States Patent Application Publication. United States Patent and Trademark Office, Federal Government of the United States, Washington, DC, USA. Retrieved 2009-09-05.