This document provides a critical review, on behalf of SSI, of the models employed by the Swedish Nuclear Fuel and Waste Management Co (SKB) for dose assessment in the renewed safety assessment for the final repository for radioactive operational waste (SFR 1) in Forsmark, Sweden. The main objective of the review is to examine the models used by SKB for radiological dose assessment in a series of evolving biotopes in the vicinity of the Forsmark repository within a time frame beginning in 3000 AD and extending beyond 7500 AD. Five biosphere models (for coasts, lakes, agriculture, mires and wells) are described in Report TR-01-04. The principal consideration of the review is to determine whether these models are fit for the purpose of dose evaluation over the time frames involved and in the evolving sequence of biotopes specified. As well as providing general observations and comments on the modelling approach taken, six specific questions are addressed, as follows.
- Are the assumptions underlying the models justifiable?
- Are all reasonably foreseeable environmentalprocesses considered?
- Has parameter uncertainty been sufficiently and reasonably addressed?
- Have sufficient models been used to address all reasonably
- Are the transitions between biotopes modelled adequately (specifically, are initial conditions for developing biotopes adequately specified by calculations for subsiding biotopes)?
- Have all critical radionuclides been identified?
It is concluded that, in general, the assumptions underlying most of the models are justifiable. The exceptions are a) the rather simplistic approach taken in the Coastal Model and b) the lack of consideration of Ôwild’ foods and age-dependence when calculating exposures of humans to radionuclides via dietary pathways. Most foreseeable processes appear to have been accounted for within the constraints of the models used, although it is recommended that attention be paid to future climate states when considering these processes. Parameter uncertainty has been addressed in detail for the existing models but numerous detailed recommendations are made for the improvement of several specific parameter selections. It is also recommended that two semi-natural ecosystem models (for forest and dry meadows) are included in the study – this would require further parameter identification and collection. Transitions between developing biotopes need more careful explanation and reasoning and EDFs (Ecosystem Dose Factors) for all ecosystem types should be presented to demonstrate more clearly that all critical radionuclides have been identified.