Operations at the Dounreay establishment have resulted in some soil and rock becoming contaminated. The site is now entering a phase in which it is to be decommissioned and restored, which is likely to take fifty years or so. The United Kingdom Atomic Energy Authority (UKAEA) is managing this process, and, amongst other issues, is addressing the question of the contaminated land. Ultimately, decisions will have to be taken as to the extent and character of the remediation of contaminated land. However, the long timescales for the decommissioning of the site mean that it is appropriate to maintain the contaminated land in a safe and controlled condition for the meantime, and make these decisions when the decommissioning of the site is further progressed.
In support of this strategy, UKAEA has produced a Safety Case for the management of the existing contaminated land during the period of decommissioning and site restoration. A key element of the Safety Case is an assessment of the likely levels of contamination and comparison with relevant targets and standards. The 'Contaminated Land Safety Assessment' uses a site-specific environmental model to calculate concentrations in the contaminated land areas and surrounding environment for a range of times. This information has been used to evaluate possible dose rates to various potentially exposed site worker groups, as well as potential dose rates to members of the public with access to the surrounding land and the foreshore.
The Contaminated Land Safety Assessment (CLSA) necessarily involves using a stylised representation of the environment as the basis for the models considered. For example, the characteristics of media are averaged over a defined area in a way that means small scale (metres or less) heterogeneity is not directly represented. This approach is well-established for environmental modelling, as accurate information on the fine detail is usually not available, and modelling such a level of detail directly is likely to be inappropriately resource-intensive for the purposes of safety assessment. It is also the case that there are other competing uncertainties that cannot be resolved, which dominate the uncertainties associated with detailed features. Furthermore, the key endpoints (doses and risks to people) are intrinsically averaged spatially, due to the fact that people inherently move around.
Given the emphasis on safety assessment through the calculation of potential doses using cautious assumptions, the CLSA model was not designed to reproduce observed groundwater concentrations. Nevertheless, comparisons were made in the original study and demonstrated agreement, but only in very general terms.
It is now considered to be of value to investigate further some of the key sources of uncertainty with respect to groundwater transport and undertake some additional comparisons with observed groundwater concentrations. Although the intention is not to seek to validate the model, measurement data provide an important point of reference for the model results, and a degree of agreement will build confidence in the safety assessment study.
This report describes an analysis of uncertainties in the CLSA model, and the undertaking of a series of sensitivity calculations to address these sources of uncertainty. It focussed on the key sources of parameter uncertainty with respect to groundwater transport. This will enable UKAEA to have confidence in the appropriateness of the model, and to be fully aware of the key uncertainties and issues associated with the dataset describing the contaminated land and the modelling approach adopted.
For more information please contact James Penfold.