Uncertainty analyses of models for high-level waste and spent fuel disposal: Results of the MICADO¹ and GLAMOR² projects

By Bernd Grambow (SUBATECH, Nantes (France)) and Pierre Van Iseghem (SCK•CEN, Mol (Belgium)

Isolation in deep geological formations (geological disposal) is now considered the safest of all long-term solutions to the problem of dealing with long-lived nuclear waste. Particular attention has been given worldwide to the high-level radioactive waste resulting from the spent nuclear fuel after discharge from nuclear power plants, as it constitutes more than 98% of the radioactivity generated in these plants. This high-level radioactive waste can be in vitrified form for those countries with reprocessing options. Alternatively, the spent fuel itself can be considered as a form of waste.

Potential disposal locations in Europe, in granite, clay or salt formations, are stable for tens of millions of years and are characterised by very slow groundwater movement. Research at international level has been on-going for more than 30 years. Large experimental data bases have been generated to simulate the long-term interaction of groundwaters with different types of nuclear waste glass and spent nuclear fuel. This includes glass with compositions similar to those produced in reprocessing and associated vitrification plants in La Hague (France), Mol (Belgium), Sellafield (U.K.) and Karlsruhe (Germany), as well as low and high burn-up spent fuels from PWR, BWR, HTR and MTR reactors.

A considerable effort has been made to develop descriptive and predictive modelling procedures. The long-term stability of the waste solids in groundwater must be deduced from short-term laboratory experiments with a duration of a maximum of a few years. As a result, the evaluation of the performance of the spent fuel and glass over different geological time periods relies largely on the development of waste/groundwater interaction models and more general safety assessment models. This allows one to assess the long-term impact of glass and spent fuel/water interactions on the overall level of risk associated with the nuclear waste repository. This assessment has to take into account the inherent uncertainties in the understanding of the processes considered in the model, the parameter uncertainties and the uncertainties in hydrogeological boundary conditions. The repositories are planned for siting in places characterised by very slow geological evolution, but the presence of waste disturbs the systems in place there and the return to natural conditions may take hundreds of thousands of years. Also uncertainties regarding the scenarios for this evolution over various geological time periods must be taken into account.

The GLAMOR and MICADO projects, which coordinated actions carried out recently by the European Commission, have assessed the uncertainties in the models and  described the dissolution processes of nuclear waste glass containers and spent nuclear fuel in a repository over geological time periods. These projects were special in the sense that a common set of existing experimental data and existing models were selected. These models were applied in each project and the results discussed. Care was also taken that participants in the projects should present the various scientific opinions.  

The GLAMOR project, coordinated by the Belgian Nuclear Research Centre, SCK•CEN, and carried out between 2002 and 2006, assessed the principal hypotheses and uncertainties of models for the dissolution of nuclear waste glass in pure water systems without repository near field materials (such as bentonite, metallic corrosion products etc.). Key European partners (SCK•CEN, CEA, SUBATECH/ARMINES) joined forces for the project with experts from leading US laboratories (SRNL, PNNL). Two reference analytical models selected by the partners were jointly evaluated using a common experimental data set. 

Some of the main conclusions reached from the GLAMOR project were as follows:  

• It was agreed that a residual dissolution rate exists for nuclear waste glass in pure water. This residual "final" dissolution rate can be very important for the assessment of the barrier function and the lifetime of the glass, because it is as much as 10,000 times smaller than the initial dissolution rate. Lifetimes for vitrified high-level waste blocks of around 100,000 years and more can now be predicted in such environments. The residual dissolution rate should be considered also in the safety assessment calculations for a geological disposal site, but there is a need for further documenting and understanding the processes behind this residual rate in order to be able to extrapolate the findings of the short-term tests in simple systems to the long-term repository conditions.  

• The dissolution rate decreasing stage that precedes the residual dissolution rate could be explained by several mechanisms included in the models. The model simulations did not allow one to decide which mechanism was dominant, but the comparison did give a clear overview of the key problems that arise when interpreting the results, and showed the weaknesses in the underlying conceptual models. This has provided a basis for further subsequent model development.

The approach followed in the GLAMOR project formed the basis for a similar coordinated project called MICADO, which was carried out between 2007 and 2010. This project was coordinated by SUBATECH Nantes (France). It consolidated the work of many European waste management agencies, technical organisations that support regulators, universities and research organisations. The organisations that participated in the MICADO project were CEA, ANDRA and IRSN from France, SCK•CEN and BEL-V form Belgium, KIT Karlsruhe, ITU and GRS from Germany, ENRESA, UPC, CIEMAT and AMPHOS21 from Spain, SKB, SSM, Studsvik and KTH from Sweden, NAGRA from Switzerland and Quintessa from the United Kingdom. Many of world's leading experts participated in the project. The general findings were shared by waste management organisations and the technical organisations that support the regulators.

A careful questioning of current experimental knowledge, hydrogeological boundary conditions and current understanding of the physico-chemical processes controlling the radionuclide release was done in order to identify key uncertainties. Separate uncertainty analyses were carried out (1) for the evolution of spent fuel for the first few thousands of years, when the container is still intact, (2) for the next few hundreds of thousands of years – a period that is likely to be characterised by hydrogen saturated conditions with hydrogen arising from container corrosion and (3) for very long subsequent time periods, which are characterised by conditions where hydrogen concentrations have diminished but the ambient condition remains one of reduction.

Some of the main findings were as follows:

• The uncertainties on the surface area of spent fuel exposed to solutions, which is a very important parameter for calculating the dissolution rate, could be strongly reduced, and ranges were agreed for the value of the dissolution rate in periods 2 (presence of H₂ – see above) and 3 (absence of H₂). The dissolution rates in periods 2 and 3 are between 0.02 and 5 µg.m^-2.d^-1, depending, amongst other things, on the value chosen for the total accessible specific surface area (ranging between 10 and 72 cm²/g).

• The occurrence of the dissolution inhibition effect of H₂ formed by container corrosion was confirmed, although the underlying mechanism is not really understood.

• The significance of the various release controlling processes and their uncertainties on spent fuel dissolution rate and the activity fluxes released by a nuclear waste repository have been assessed both by waste management and technical support organisations. Spent fuel dissolution and radionuclide release rates are important parameters in assessing performance. However, the characteristics of the disposal components (including sorption, diffusion and advective flows) also strongly influence the transfer of radionuclides to the biosphere. The results show that spent fuel is an effective isolation barrier for tens of thousands to millions of years. They also show that the instant release fraction (corresponding to the radionuclides in the different "not confined" zones of the spent fuel packages, for example the zirconia cladding layer) is the dominant dose contributor if the matrix dissolution rate is small (in the range mentioned above) and the resulting lifetime of the spent fuel is significantly longer than 10⁶ years.

• The validity and relevance of model and experimental data for long-term spent fuel behaviour has been assessed in comparison with the observations made of the reaction behaviour of naturally occurring minerals with partially similar compositions as spent fuel - such as natural uraninite.

In conclusion, this way of launching relatively low cost research has enabled us to reach a common understanding or consensus on a number of critical issues in the chemical durability of both high-level waste glass and spent nuclear fuel. Both projects also led to the identification of some key questions for further action. The findings should contribute to enhancing the credibility of our research, because they were reached jointly by international experts.

Acknowledgements

The research that led to these results received funding from the European Union's European Atomic Energy Community's (Euratom) Fifth and Sixth Framework Programmes, under the grant agreements  FIKW-CT-2001-20140 (GLAMOR Project) and 36366 (FI6W) (MICADO  project).

Bibliography:

GLAMOR: A Critical Evaluation of the Dissolution Mechanisms of high Level Waste Glasses in Conditions of Relevance for Geological Disposal,   P. Van Iseghem, M. Aertsens, S. Gin, D. Deneele, B. Grambow, P. McGrail, D. Strachan, G. Wicks, European Comission, EUR 23097 (2007)

MICADO: Model Uncertainty for the Mechanism of Dissolution of Spent Fuel in Nuclear Waste Repository, Final Report, B. Grambow, J. Bruno, L. Duro, J. Merino, A. Tamayo,  C. Martin, G. Pepin, S. Schumacher, O. Smidt, C. Ferry, C. Jegou, J. Quiñones, E. Iglesias, N. Rodriguez Villagra, J. M. Nieto, A. Martínez-Esparza, A. Loida, V. Metz, B. Kienzler,
G. Bracke, D. Pellegrini, G. Mathieu, V. Wasselin-Trupin, C. Serres, D. Wegen, M. Jonsson, L. Johnson, K. Lemmens, J. Liu, K. Spahiu, E. Ekeroth, I. Casas, J. de Pablo, C. Watson, P. Robinson, D. Hodgkinson Eoropean Commission, EUR 24597 EN 2010