- Technical Report NTB 02-06Download
The present report is designed to provide readers with the necessary information to perform, if they so wish, independent checks of the results of the evaluation of the "assessment cases" described in the Safety Report (Nagra 2002c) for Project Entsorgungsnachweis. It also describes the conceptual models and corresponding codes for the near field, geosphere and biosphere that were used in the safety assessment to evaluate the assessment cases, including the reasons why they are considered adequate for their intended purposes, and the operational elements and procedures that were used to manage the required large number of calculations. Models and codes not described in this report are the supporting models used to derive parameter values for the near field, geosphere and biosphere codes, and to support model assumptions. These include, for example, groundwater flow models, mechanistic models of sorption, temperature evolution models, waste dissolution models, etc., and are described in the Project Entsorgungsnachweis reference reports.
An assessment case is a specific set of assumptions regarding the broad evolution of the repository and its environment, the conceptualisation of individual features, events and processes (FEPs) relevant to the fate of radionuclides within the disposal system and the parameters used to describe these FEPs. In the safety assessment, a broad range of assessment cases is analysed in order to illustrate the impact of various detrimental FEPs and uncertainties on the level of safety provided by the disposal system. The assessment cases are defined, the underlying reasoning documented and the results of their analysis presented in the Safety Report. In the interests of transparency, the Safety Report presents these descriptions in a mainly qualitative fashion, without exhaustively documenting all relevant formulae and data. The present report complements the Safety Report with a comprehensive description of models, codes and data and thus provides traceability within the safety assessment. The two reports together satisfy the assessment principle (see Chapter 2 of the Safety Report) that the development of the safety case and its results should be documented in manner that provides both transparency and traceability.
In the safety assessment for Project Entsorgungsnachweis, assessment cases are divided into a number of groups, according to the issues or uncertainties that they address. The main part of the present report focuses, in turn, on groups of assessment cases that explore:
- the consequences of particular scenario, conceptual and parameter uncertainties, where this range can be bounded with reasonable confidence on the basis of available scientific understanding,
- more speculative "what if?" possibilities that are considered in order test the robustness of the disposal system,
- various design or system options, and
- different stylised possibilities for the characteristics and evolution of the surface environment ("biosphere").
Overviews are given of the conceptual models underlying the assessment cases and the various assumptions and simplifications that are made in order to arrive at sets of mathematical equations and input parameters so that each case can be evaluated using corresponding computer codes, with references to appendices that give detailed descriptions of the codes and the equations that they solve, as well as tables containing the data used and the data source. In the appendices, descriptions are given of the capabilities of each code in terms of the phenomena (or "Super-FEPs") that they address. This is important within the FEP management procedure to allow an evaluation of the suitability of a given code to address a given Super-FEP (Nagra 2002d). It is shown that the codes used are sufficiently versatile to evaluate all the required assessment cases. There are a few safety-relevant phenomena that the codes are not "qualified" to evaluate, but, in all cases, either significant effects can be ruled out by supplementary studies (e.g. for criticality), effects are intrinsically favourable to safety and can be conservatively neglected (e.g. transport resistances in the SF / HLW near field), or parameters (e.g. canister breaching time) can be chosen to ensure that calculations err on the side of pessimism.
In addition to the models and codes used to analyse the assessment cases, simplified insight models are used to examine particular aspects of system performance and sensitivity to key system properties and model assumption. The insight models are described in Chapter 9.
The deterministic evaluation of assessment cases is complemented by probabilistic calculations to build further system understanding and, in particular, to indicate the performance of the system for parameter combinations not analysed by deterministic calculations. The computational tool used to sample input parameters from probability density functions (PDFs) and the PDFs themselves are described in Appendices 2 and 3, respectively.
Except for a few cases, no justifications are given in the present report for the model assumptions made and no final or intermediate results are presented, except where this is considered to help understanding. For a full presentation of results and a justification of assumptions, the reader is referred to the Safety Report and the reference reports.
It is, however, noted that all assessment cases are evaluated using, to some extent, pessimistic or conservative conceptual assumptions, parameters and model simplifications, resulting in calculated doses that should be regarded as upper bounds for, rather than predictions of, expected doses. To assess the degree of bias introduced by such assumptions, the modelling approaches used to evaluate each assessment case are examined systematically in Chapter 10. It first summarises the various assumptions and simplifications made in order to arrive at sets of mathematical equations and input parameters so that each case can be quantitatively evaluated, and then it assesses these assumptions and simplifications in terms of their degree of realism, pessimism or conservatism. This information is then used as a tool to systematically assess the bias that has been introduced into the models ("bias audit").