A socio-historical analysis of a memory practice and corresponding regulatory work, through the lens of the “Centre de Stockage de la Manche” (CSM, La Hague, ANDRA, France).

The recent decision to revive the nuclear industry, with the construction of a series of nuclear power plants and, above all, with the launch of the CIGEO (Centre Industriel de Stockage Géologique) project, propels the issue of memory and nuclear legacy in the spotlight. However, this issue also reveals some of the most striking strategic and operational challenges facing this industry.
The constitution, the preservation and the transmission of the memory of nuclear sites and activities are based on numerous social practices, organizational processes and socio-technical devices (paper archives, digital twins, etc.). They make it possible the constitution, the maintenance and the exploration of externalized representations of the past and their availability for unspecified users in the future. The very fabric of these representations deals with a variety of problems, objectives and contingencies (NEA, 2019). They meet multiple ethical, social and cultural needs (Candau, 2005). Hence, memory appears to be at a nexus and cannot be reduced to its visible and externalized components. Its definition and comprehension must integrate social, regulatory and organizational dimensions as well as the processes and practices that have contributed to the emergence of the representation as a consensual formalization. 
Following Noiriel (2008), this paper is an attempt to sketch out the socio-history of a memory device once presented as “the ultimate bulwark against oblivion” (ANDRA, 2008) and that the ANDRA proposed to implement as the memory of the future Industrial Geological Storage Center (CIGEO).  
This memory device and the associated practices emerged mainly through the efforts of successive organizations (emanating from the French Nuclear Energy Commission, CEA, and former predecessors of the ANDRA) facing social conflicts and contestations surrounding the Centre de Stockage de la Manche (CSM, La Hague), over 50 years since its creation. During the 1990, a vague and generic injunction issued by an ad-hoc commission of experts, the Turpin’s commission (1996), emerged and triggered a regulatory work. Under the aegis of the ANDRA, this injunction became a twofold set of key information and records (ANDRA, 2008). In 2016, in a new regulatory framework dedicated to decommissioning and nuclear waste repository it was the main (but implicit reference) of the French safety regulator. During the same period, the NEA, as an international body, contributed to the dissemination and the institutionalization as a set of good practices known as Key Information Files (KIF) and Set of Essential Records (SER). More recently these practices and devices were the starting point and the framework proposed by the ANDRA to be implemented in its official request for authorization to create the CIGEO center.
Our work allows us to draw interesting conclusions about the influences of the social and organizational frameworks on the social practices and socio-technical devices that constitute the memory and the forgetting of a nuclear waste repository. ANDRA's memory were instrumental and ‘equipped’ a conflictual situation. Through its institutionalization (and the corresponding regulatory work), it has been applied, beyond the CSM waste storage, to the CIGEO project and probably to future projects.