The reversibility for repositories is frequently quoted in the media, and opponents to nuclear energy or to waste burial often justify their opposition by the impossibility of reversibility in deep geological formations, as an echo to the supposed absence of reversibility in everything associated with nuclear energy in general.
Reversibility responds to various motivations: in case of error or of unpredicted events, it must be possible to intervene; inversely, it must also be possible to recycle certain materials many years after disposal; due to a potential lack of confidence, it must finally be possible to progress in a stepwise approach in order to ensure that every generation may orient the process.
That approach leads to a wide definition of reversibility no longer restricted to demonstrating that it is possible to remove a package technically (“retrievability”), but rather encompassing the possibility to modify the process at each step throughout the lifetime of the repository.
Confronted with this definition for reversibility, the preliminary design options of repositories selected for the Meuse/Haute-Marne site in France appear at this stage to be sufficiently flexible to allow further progress in the feasibility study of a reversible repository.
Four main areas of study need to be emphasised:
• It is necessary to further the knowledge concerning the evolution of a repository: What phenomena control the evolution of the repository? What characteristics of the different components do they affect?
• It is necessary to define reversibility levels and their specific characteristics while modulating them, if necessary, according to the different waste types;
• It is necessary to define a monitoring programme in close relationship with the phenomena modelling ensuring the representation of the repository evolution over time and the analysis of the safety conditions during each phase;
• Finally is it necessary to examine the effects of the different states of the repository on the implementation of technologies: for example, what characteristics or what state of the components may complicate package retrieval, in other words make the proposed technological means inadequate to intervene at each reversibility level?