Control of very serious reactor accidents

This paper is concerned with very serious accidents, i.e. with potentially several thousand worldwide casualties in the longer term. Chernobyl showed that effective action can be taken to reduce the release of activity following a very serious reactor accident, but at a human cost. In this paper we examine the merits of advance planning for such control action. There is an argument that a very serious reactor accident will not happen in the UK and, therefore, that planning for one is unnecessary. Certainly, the likelihood of a disastrous accident is extremely low and the risk to any individual in the UK is small. Nevertheless, it is clear that the risk is very widely distributed, and this in turn gives rise to a public concern which may not be taken lightly. One intention of this paper is to promote a wider debate of the issue of preplanned on-site action for mitigating the consequences of very serious reactor accidents. Means of controlling such accidents are indicated in the paper, but these are preliminary proposals only. A serious attempt to pre-plan such actions will involve considerably more study than the authors have devoted so far. (author)