Methodology and tools for source term assessment in case of emergency - astrid project (EC 5th framework programme)

Full text: Following the reactor accident in Three Mile Island most western European countries with nuclear power reactors started a huge safety upgrade work to improve not only reactor safety but also the introduction of mitigating systems that would minimize consequences of a severe reactor accident. Further to this also emergency preparedness arrangements were made; alarm criteria, on-site and off-site emergency structures and plans to protect population and environment. Following also the Chernobyl accident in 1986, many countries started the development of dispersion codes to calculate and better predict the consequences of a radioactive release. However, follow-ups of the Chernobyl accident in the mid 90ties revealed the need for an earlier start of assessing the actual severity of an accident to efficiently succeed in emergency response actions. By looking into the NPP, at the state of f fission product barriers and critical safety systems, the magnitude of a potential radioactive release could be predicted in a timely manner to allow emergency response to be executed even before the occurrence of a release. This is the perspective in which the development of ASTRID methodology and tool should be regarded. By focussing more an similarities than differences the ASTRID methodology aims at a solution that could be acceptable for several reactor types as well as reactor containments, different stages of technical designs, but also for use at on-site as well as off-site emergency centres. So, the scope of the ASTRID methodology outlines the common FP barriers and interrelated critical safety functions, that is, plant process parameters of importance to address, for such an assessment of the likely future state of the failed NPP and the resulting source term. The methodology maps out relevant process parameters and indicators, what and how to calculate and a structured way to summarize and conclude on potential source term and likely time projections. In a way this is components in a source term assessment toolbox where single tools could be substituted over time. To suite different user situations, both on-site and off-site as well as size of staff, priority and work order is likewise addressed. The output from such an assessment is intended to, firstly give bases for decisions on necessary urgent protective actions pre-release, and, secondly an input for the sophisticated dispersion calculation codes. It consists of basically two different products; a source term description (amount, composition, release height, heat content, time scale); reports of different kinds (non-technical action orientated reports and technical expert reports). The importance of facilitating cross-competence communications among EP players is very much a question of efficiency. The advantage of harmonized European methods would increase confidence in what is communicated among EP organizations but also between countries. In the ASTRID methodology there is no kind of preference with regard to deterministic or probabilistic ways of making the assessments though alarm criteria and plant parameters are physically based. Actually, combinations might be wise to reach the objectives of timeliness in assessments. (author)