Quantifying Conservatism of Performance Assessment Calculations by Sorption Model Reduction: Case Study on Near Field Cs

  • PDF / 426,202 Bytes
  • 8 Pages / 595 x 841 pts Page_size
  • 61 Downloads / 173 Views

DOWNLOAD

REPORT


Quantifying Conservatism of Performance Assessment Calculations by Sorption Model Reduction: Case Study on Near Field Cs Migration in Callovo-Oxfordian Clay

Eef Weetjens, Evelien Martens and Diederik Jacques Institute Environment, Health and Safety, Belgian Nuclear Research Centre SCKā€¢CEN, B-2400 Mol, Belgium

ABSTRACT Implementation of full thermodynamic models in performance assessment (PA) calculations (large domain and very large timescales) is practically unfeasible due to excessive computational times. The complex competitive sorption processes of radionuclides are often simplified. In this paper, such simple sorption models (i.e. linear K d and Langmuir isotherms) were compared with more complex thermodynamic models for a reduced geometry and relatively short timescales (compared to traditional PA calculations). Within the context of a safety case the value of this study is twofold. Primarily, it provides support information on how to choose adequate parameter values in a consistent simplified analysis for compliance with safety criteria. For the cases studied in this paper it became clear that a linear model is sufficient to represent sorption, but a proper choice of the Kd values is critical. Secondly, the comparison of results from these compliance calculations with more realistic analyses demonstrates quantitatively the safety margin implicitly present in PA calculations through model abstraction.

INTRODUCTION In classical performance assessment (PA) calculations, the often complex geochemical transport behaviour of radionuclides in the near field of a geological repository is (necessarily) simplified in view of the extreme timescales that are to be computed. In the Belgian approach, compliance calculations tend to demonstrate safety through the use of safety functions, in which radionuclide release and transport processes through engineered barriers and surrounding geological layers are described with relatively simple but conservative models, hence, using a limited number of parameters. The objective of this paper is to test whether using more complex and more realistic modelling approaches provide added value and whether it is required to include processes currently not fully accounted for in PA. For that purpose, a case study is defined in the PAMINA project (see below) in which a comparison is made between models with different levels of complexity in describing sorption processes for a small-scale model, looking at the near field of a geological repository in Callovo-Oxfordian clay (COX clay) for a timeframe up to 50000 years. This case study focuses specifically on the migration of 135Cs, released from vitrified waste packages. Simulations consider three different sorption models: (i) a linear Kd approach, (ii) sorption according to a Langmuir isotherm, and (iii) cation exchange models for clay and bentonite. One as well as two dimensional simulations are performed for a radial geometry using different numerical tools. Whenever conceptual models are equal, results are verified.

This work consti