Jean-Marie Gras

Long term evolution of spent nuclear fuel in long term storage or geological disposal. New findings from the French PRECCI R&D program and implications for the definition of the RN source term in geological repository

This paper aims to give a brief overview of the wide research undertaken in France in order to elucidate the potential long term evolution of spent nuclear fuels in long term storage or geological disposal. Scientific key issues related to the potential long term evolution in closed system, in presence of and oxidative phase and in presence of water are presented as well as the anticipated trends. A particular emphasis is put on the major outcomes of this research which is a new definition of the radionuclides source term for the geological disposal: we estimate that we have now to allocate a higher fraction of the radionuclides inventory to the so-called instant release fraction which is instantaneously released in presence of water.

A New Approach to the RN Source Term for Spent Nuclear Fuel under Geological Disposal Conditions

Under the geological disposal conditions, spent fuel (SF) is expected to evolve during the 10,000 years while being maintained isolated from the biosphere before coming in contact with water. Under these circumstances, several driving forces would lead to the progressive intrinsic transformations within the rod which would modify the subsequent release of radionuclides. The major mechanisms are the production of a significant volume of He within the UO 2 lattice, the accumulation of irradiation defects due to the low temperature which avoids any annealing, the slow migration of radionuclides (RN) within the system by (i) the α self-irradiation-induced athermal diffusion and (ii) locally the building-up of internal mechanical stresses which could turn the pellets into powder. However, the current RN source terms for SF have never accounted for this evolution as they are based on existing knowledge of the fresh SF. Two major mechanisms were considered, the leaching of the readily available fraction (one which was supposed to be instantly accessible to water), and the release of RN through alteration of the UO 2 grains. We are now proposing a new RN source term model based on a microscopic description of the system in order to also take account of the early evolution of the closed system, the amplitude of which increases with the burnup and is greater for MOX fuels.

Key Scientific Issues Related to the Evolution of Spent Nuclear Fuel in Long Term Dry Storage and Geological Disposal

The anticipated long term evolution of spent nuclear fuel as well as the remaining scientific key issues are presented for the various boundary conditions that can be encountered in long term dry storage and geological disposal. Spent fuel is expected to evolve significantly in closed system conditions which are representative of long term dry storage and the first stages of geological disposal. The mechanical evolution of the grain boundaries, the fate of helium and the evolution of the RN location within the pellet are the three major questions to be addressed which could significantly modify the physical and chemical state of the fuel. In addition, mechanisms and kinetics of fuel alteration by water in deep geological repository are still to be more deeply understood, in particular the inventory of the instant release and the radiolytic dissolution processes, to get a robust and reliable source term.

Consequences of the Anticipated Long-Term Evolution of Spent Nuclear Fuel for the Assessment of the Release Rate of Radionuclides.

The research conducted in the framework of the French research project on spent nuclear fuel (SNF) long - term evolution (PRECCI Project) has enlightened the potential significance of spent nuclear fuel intrinsic evolution in closed system for the assessment of radionuclide (RN) source term in long-term storage or geological disposal. Beyond others, alpha self-irradiation enhanced diffusion and evolution of the grain boundaries cohesion are two major processes which have to be accounted for in view of the RN source term models development. Accounting for these processes, operational models are developed, the aim of which is to quantitatively define the RN release rates from SNF in long-term storage or geological disposal. They distinguish basically an instantaneous contribution (IRF in geological disposal) and a time-dependent contribution (matrix oxidation or alteration). RN inventories associated to these two different processes have to be modeled since they are time-dependent due to the RN diffusion within the pellet. The present paper details the models that are developed in France in terms of assumptions, conservatism and robustness. It comes out from this work that for the instant release fraction, we have to consider a much higher instant release fraction than classically assumed (5–6% in geological disposal) in particular for geological disposal.