Concepts of decision making during an early phase of a nuclear accident

Full text: The final decision in case of a nuclear accident is, in principle, simple: 'To take or not to take the protective action?' Nevertheless the process of decision making, which leads from the initial information of the event to the final decision, is not straightforward. The goal is to present different concepts of this process rather than analyse various methods and tools, and try to assemble different parts into the puzzle, which gives a clear picture. The postulates which should be always kept in mind at all stages of process: 1. this is an early stage of the accident (the first 24 or 48 hours) - time is important, therefore do not lose time thinking about longer term actions. 2. Avoid deterministic health effects and high risks of stochastic health effects should be minimized. 3. Make it simple. Do not get bogged down with too many details (although they might seem important) in order to keep clear scheme of thinking. The process will be split into three categories of expertise: nuclear technology issues, radiation protection issues, optimization of all factors. The first two categories are more technical, the third category belongs to the real decision maker, usually not an expert but a politician. A) Nuclear technology issues. They cover areas as design basis accidents, severe accidents phenomenology, probabilistic safety analyses, reactor engineering, thermohydraulics, etc. These experts should provide the answer to the following question: 'When the release is going to happen and what the source term would be?' At the onset of an event and at a first glance one would say: 'Only a small fraction of initiating events lead to a core melt and even smaller fraction could be associated with radioactive release to the environment and further less could lead to a large release to the environment'. If the initiator could be a large contributor to the core damage frequency, then the parameters to watch before the core melt (or serious degradation of fuel cladding) are inventory of the primary coolant, capability to maintain this inventory, its temperature and pressure. First derivatives of three parameters, if available, might give the indication about the speed of the processes. After the core melt the pressure in the containment and the hydrogen/water ratio in the containment atmosphere may give the indication, if the containment is going to fall. The data about the inventory of radionuclides released to the containment atmosphere could be deduced from the samples of containment atmosphere, if available, from the readings of high range radiation monitors in the containment or assessed from the time the core was uncovered. It should be noted that before the release, the role of nuclear technology experts is very important. After the release has started, their rote is slightly less important, because one can rely on radiation monitor readings. In the case of breach of containment the assessment of the termination of release is straightforward, it will go empty until the pressure drops to external pressure. B) Radiation protection issues. The radiation protection specialist (RP specialist) needs to work closely with the nuclear technology expert. The rate of release to the environment is very difficult to assess if it is not just the design leakage of the containment. The RP specialist could calculate what would be the opening in the containment to produce such a release, which can cause the intervention level (e.g. for evacuation) to be exceeded in the downwind sector and compare it to the actual penetration size (assuming failed penetration pipe). The RP specialist deals with weather in close co-operation with the meteorologist. lt is particularly important not to just assume the weather (wind speed, wind direction, Pasquill-Gifford stability class) at the time of calculation, but to assume (anticipate) the probability of change of these parameters in a short time period. If the weather is unstable, e.g. the wind changes the direction following some pattern (wind shift). The local breeze can shift the direction during the day even the weather is stable. When the clouds are disintegrating (evaporating) the stability class can change for many classes in just one hour. The wind forecast in the potentially affected area for the 6-hrs intervals or shorter is appreciated, if available. The determination of stability class is defined with different methods, which do not lead to the same final result, therefore new concepts using SODAR were proposed. The strategy of radiation monitoring teams how to perform plume tracking should be known in advance. When the measured data are received from the field monitoring teams a sort of check an accuracy and quality of data should be made. In this operation the data an which team sent the Information should not be lost. One should try to visualise the plume with contour mapping. Corrections of source term could be anticipated. The outcome expected from the RP specialist is the recommendation for protective actions based on radiation protection principles (i.e. Intervention levels) only. C) Optimization of all factors. The final decision maker who performs the optimization (or at least evaluates it) should take into account all significant factors not just the radiation protection principles. The persons at this stage are not specialists, but usually politicians or experienced civil servants, whom should be given proper advice from other non-radiological specialists. (author)