First, we must ask the question: ‘Why are nuclear reactors hazardous, and in exactly what way are they hazardous?’ The answer to this question (Meneley, 1999) should underlie the rationale for all analysis of potential failures and the means for mitigating those hazards.

10.3.1 Hazards of solid fuel reactors

A typical fuel pellet made of sintered uranium dioxide (melting point approximately 2800°C) is shown in Fig. 10.5. Millions of such pellets are located in an operating power reactor. Almost all the reaction products of fission — fission products — are trapped inside this pellet. The second impor­tant fact is that most of the heat energy of the fission process is produced inside this same pellet. Heat is removed by flowing coolant (usually high — pressure water).

It is apparent that either increasing the rate of heat production (i. e. increasing the rate of fission) or decreasing the rate of heat removal (i. e. decreasing local water pressure or flow) threatens to increase the tempera­ture of the fuel pellet and therefore bring it closer to melting temperature. If and when the pellet melts it will release essentially all of its fission prod­ucts that are volatile at the mixture temperature — some of which are highly radioactive. These fission products represent the main hazard of nuclear reactor operation.

High-pressure water presents an obvious hazard due to the possibility that a pipe might break and release the water, and so might lead to over­heating of the fuel pellets, if emergency water supply were not available. Other initiating events that reduce the heat removal rate (e. g. loss of forced circulation) add similar hazards.

Returning to the fission process itself, and recalling that the process involves a chain reaction, sheds light on yet another hazard of fission reac­tors. We all know that a chain reaction involving successive generation of fissions is at the heart of this technology. We also know that a few (less than 1% of the total) of the next-generation neutrons essential to keep the chain reaction going at a constant rate are emitted after a slight delay. This is known as the ‘delayed neutron fraction’. Further, to increase reactor power we must manipulate controls so that the number of neutrons in each suc­cessive generation is slightly larger than the number in the previous genera­tion. It is important to control this increasing neutron population to a very low rate so that engineered control systems can return the excess number of neutrons per generation to zero once again, when the desired neutron density (proportional to the reactor power level) is reached.

A serious hazard may arise if and when the excess number of neutrons in successive generations approaches or even exceeds the number of delayed neutrons in that generation; in such a case the rate of multiplication becomes very much faster. If this number exceeds the delayed neutron fraction, the dominant rate of power increase becomes inversely proportional to the time between successive fissions, and the delayed neutrons are left behind. This is known as the ‘prompt critical’ state. Different reactors have different characteristic times; they range between about 1 millisecond in a thermal neutron reactor design to less than a microsecond in a fast neutron reactor. In every case of abnormal operation when a ‘prompt critical’ state can occur it is vital to ensure that either inherent characteristics or highly reliable engineered systems will act to return the reactor to a non-self-sustaining condition — that condition is known as the ‘shut down’ state.