In this special issue, we, the invited co-editors, have sought to assemble for the general readership of this Journal an introduction to the present state of the art in the treatment of probability and uncertainty in the radiological assessment of underground radioactive waste disposal for the long-term period after closure of the repository (the 'post-closure' period). Typically, such assessments make estimates of performance over time periods of up to thousands or even millions of years into the future.
Particularly during the past 15 years or so, national and international research and development programmes in radioactive waste management have apportioned substantial funds to developing assessment tools, techniques and procedures. A considerable literature has grown around these efforts and the disposal sites and facilities that have been considered.
For high level radioactive wastes, repositories at depths typically of many hundreds of metres are envisaged. At present, no such facility has been licensed although possibilities are being considered on the basis of underground research laboratories. For instance, those operated under the aegis of the Commission of European Communities (CEC) and situated in a Boom Clay layer beneath Mol in Belgium, see Fig. 1, and in granite under the Canadian Shield at Lac du Bonnet operated by Atomic Energy of Canada Limited (AECL), see Fig. 2, or in bedded salt at the Waste Isolation Pilot Plant (WlPP) site in New Mexico under the control of the United States Department of Energy (USDOE), see Fig. 3.
Shallow disposal facilities for low level wastes have, however, been operating for some decades, as for instance at Drigg, Cumbria, in the UK, see Fig. 4, and at the Centre de la Manche, in France, whilst deeper facilities for such wastes are currently being licensed in Finland and Sweden, for instance the SFR repository, shown here in Fig. 5.
Considerable effort has been expended in the assessment of disposal concepts and in rehearsals of assessment as, for example, in the CEC PAGIS programme, 1 the Swedish KBS-32 and SKI Project 90, 3 also the Swiss Project Gew/ihr. 4 Rehearsal of fully probabilistic assessments have been undertaken as in