The starting point for the radiological consequence analysis is the set of calculation cases derived from the preceding steps of the safety assessment methodology, specifically performance assessment and safety scenario development.
The reference safety scenario corresponds to the expected initial conditions and evolution of the repository system. Uncertainty in geological characterisation leads to alternative safety scenarios or conceptualisations and to most of the calculation cases that relate to physical phenomena (Section 7.3) such as the presence of an undetected fault, or the presence of water-conducting sedimentary features within the confining geological units.
In the reference safety scenario, the repository system performs its safety functions as designed, even considering wide parameters ranges. The reference conceptualisation of the CRZ conservatively assumes that the Keuper aquifer is hydraulically active and forms a significant advective transport path below the repository. An alternative conceptualisation considers an inactive Keuper aquifer below the HLW repository section (see Section 7.2.2). Deterministic calculation cases for the reference safety scenario cover the reference case with best estimate parameters22 and a biosphere according to present-day climate, three different biospheres that account for alternative climatic conditions and geomorphology, and one case where all the sensitive parameters23 are set to pessimistic bounding values. These cases are complemented with a probabilistic calculation case (Section 8.2).
In the two alternative safety scenarios, a hydraulically active geological unit within the CRZ and an undetected fault in the vicinity of the repository, are considered (Section 7.3). Conservative assumptions about the properties (e.g. parametrisation, geometry) of these geological units and features lead to upper bounds of potential radiological consequences. Calculation cases for the alternative safety scenarios consider different hydraulically active units within the CRZ and different orientations and locations of the undetected faults (Section 8.3).
To test the repository system robustness, several “what-if?” cases are postulated and analysed. The first category of “what-if?” cases relate to physical phenomena but that are shown to be hypothetical, i.e., de-facto impossible (Section 7.4.1). This category contains cases postulating either undetected or newly generated or re-activated faults with unrealistically high transmissivity. They are treated very similarly to the cases of the alternative safety scenario of an undetected fault, but with hypothetical parametrisations. The first category further contains cases analysing the excavation of the repository by erosive processes. Even though excavation of the repository by erosive processes is deemed to be hypothetical within the time period for assessment, the analysis is required by the regulatory guideline ENSI 33/649 (ENSI 2018). The second category of “what-if?” cases postulates the hypothetical degradation of the pillars of safety (Sections 7.4.2 and 8.4.2).
Finally, calculation cases considering future human action safety scenarios (Sections 7.5, 8.5) complement the analysis, as required by regulation (Section 2.4).
An important conservatism throughout the consequence analysis is that radionuclides that leave the modelling domain of the geosphere (via deep aquifers) are instantaneously transferred to the regional aquifer of the biosphere. Thus, additional delay with decay and further dispersion is neglected.
Best-estimate values for parameters are used whenever there is enough knowledge or arguments to do so (e.g., as for diffusion coefficients of the Opalinus Clay). If this is not the case, conservative values are assumed (e.g., no complete containment for L/ILW). Also, care is taken to avoid potential risk dilution, which introduces further conservatism (e.g., by assuming instantaneous failure of HLW disposal canisters after 10,000 years). ↩
Sensitive parameters are one result of the uncertainty and sensitivity analyses with the probabilistic calculation case. ↩