ASTRID, Back to the Future: Bridging Scales in the Development of Nuclear Infrastructures

Abstract

Very few papers exist in the field of social sciences that follow and study nuclear infrastructure design projects from the inside. Such a perspective would make it possible to understand the mechanisms of their successes or difficulties at their very origin. At a time when high hopes are placed on civil nuclear energy to solve the climate issue, but when simultaneously, nuclear industry actors are facing major difficulties in a great number of flagship projects, this situated understanding of any given project’s trajectory becomes crucially important. This paper proposes an analysis, from the inside, of a project that, about ten years ago, raised great expectations from both the French and global nuclear industry, but which, in 2019, was finally halted. This project is that of a so-called fourth generation sodium reactor: the Advanced Sodium Technological Reactor for Industrial Demonstration (ASTRID). ASTRID was a new sodium-cooled fast reactor. Begun in 2010, the project’s development was suspended at the end of 2019 by the French authorities. Through an inductive study of the project in the making, conducted from 2015 to 2019, the research team was able to follow the project trajectory and carry out nearly 30 interviews with actors directly involved in the project itself. By studying ASTRID as an infrastructure development project and building on the concepts of scale and infrastructure from the literature, the ASTRID project’s halt can be understood. The project’s suspension was the result of the increasing complexity and ambiguity faced by project members and stakeholders in aligning the local and global scales as the new infrastructure was developed, and more precisely, the ASTRID project infrastructure and the global nuclear infrastructure. Our analyses show that ASTRID’s trajectory gradually drifted as a result of three misalignments between the project’s infrastructure and the global nuclear infrastructure: a temporal, social, and physical misalignment. As a result, the project identity can be described as having been vague and ambiguous. This paper sets out how such a lack of clarity impacted design practices, the project organization as a whole, and the ASTRID trajectory. In consequence, it is crucially important that lessons are learned from the project’s cessation to understand both the difficulties related to the nuclear renaissance and in terms of the field of project management in general.

Keywords:

• Sodium-cooled reactor design
• nuclear infrastructure
• project management
• scales
• alignment work

Acknowledgments

We would like to thank the ANR (French National Research Agency) and the Investissement d’Avenir program that funded the AGORAS research project in which this study has been carried out (Grant ANR-11-RSNR-0001). For their helpful comments and advice, we want to express our gratitude to the AGORAS scientific committee, to the reviewers and editors, and especially to our colleague Aditi Verma who did an increadible work. We also thank Tima Bansal for having drawn our attention to the issue of “scale.” Finally, a special and huge thank you goes to our informants for their help and patience.

Notes

^a ASTRID was the only demonstrator project, and as such, was recognized as the French active contribution to Gen IV reactor development. France is involved in two other Gen IV projects, i.e., ALLEGRO (a gas-cooled fast reactor) led by the CEA and a molten salt fast reactor led by the French National Research Center. These two concepts are at an early research stage and presented as limited contributions to Gen IV development.²

^b By this term, Midler refers to the increasing number of projects put in place to manage a growing number of activities. Projectification is, following Midler’s argument, both a process through which organizations adapt to changing environments and a transformation of activities into specific projects.

^c Studying the case of microfinance programs, Bansal and Kim show how efforts to scale them up resulted in a coarse attentional grain that prevented organizations from detecting latent issues such as regional differences. This resulted in expected negative consequences, such as “harassment by loan officers and collection agents to increase loan recovery rates and their implications in a specific cultural setting (e.g., rural India)” (Ref. 6, p. 233) leading even to suicide.

^d At the time of the study, France was engaged in two main projects: EPR and ASTRID. EPR was already in construction in Flamanville and facing many difficulties, placing it in the media spotlight. It has become a highly political object, making it very difficult to access for researchers. This difficulty was reinforced by the fact that EDF, the industrial firm leading the EPR project, was not a partner of our research program. Instead, the AGORAS consortium was composed of academic and industrial partners, including the industrial designer (ex-Areva) and the Safety Technical Support Organization (IRSN), both involved at different levels in the ASTRID project. Areva, now split between Orano and Framatome, was a key actor in the ASTRID project since it was in charge of designing the nuclear island, which allowed our research team privileged access to ASTRID design teams. The decision to study ASTRID was thus justified by scientific, political, and practical reasons.

^e These actors have a long history of collaboration, but have also experienced significant tensions in the past. Notably, Hecht ⁴ provided a very detailed analysis of what she called the “war of the systems,” which opposed the CEA and EDF in the 1960s. At stake was the choice over the technological options to build the French reactor fleet, with EDF pleading for the replication of Westinghouse PWRs and the CEA favoring a technology of French origin, the gas-graphite reactors.

^f This expression was first used in a parliamentarian report on the safety and security of nuclear installations written in 1998 [“Le contrôle de la sûreté et de la sécurité des installations nucléaires”, BIRRAUX (Claude), Député; LE DRÉAUT (Jean-Yves), Président; REVOL (Henri), Vice-Président].

^g The GIF was launched in 2000 by the U.S. Department of Energy in order to renew and stimulate research on future nuclear technologies worldwide.

^h With the main goals of Gen IV reactors in mind, more than 100 experts from 12 countries selected six concepts (among about 130 evaluated) as the most promising for research and development (R&D): gas-cooled fast reactor, lead-cooled fast reactor, molten salt reactor, supercritical water-cooled reactor, very high temperature reactor, and SFR.

ⁱ To Edwards,²¹ the notion of “force” refers to the different types of powers on which energy-based infrastructures depend, intermediate between the human body and the geophysical.

^j Data were collected by the research team involved in the research action 3 of the AGORAS project, which was coordinated by the first author. The two authors were involved in the collection of the data that are used here.

^k This is one limitation of our sample. The CEA was not part of the AGORAS project, and despite our relationships with the Areva project team, it was difficult to access project stakeholders from the CEA. The focus groups enabled meeting participants from the CEA, but not the engineers directly involved in the ASTRID project. In 2019, contacts were established with the ASTRID PMs from the CEA and interviews were planned.

^l See for example the 2007 GIF annual report, which insists on the key participation of industrial actors at https://www.gen-4.org/gif/jcms/c_43518/2007-annual-report.

^m Star defines a master narrative as “a single voice that does not problematize diversity” (Ref. 13, p. 119).

ⁿ While SFRs are not strictly French technologies (see for example the EBR I in the United States or the BN 600 and 800 in Russia), the French SFRs (and ASTRID in particular) are embedded in a typically French strategy for the management of the nuclear fuel cycle (including the reprocessing plant at La Hague) and the longstanding vision of having one day a closed nuclear fuel cycle.

^o Project identity is a conceptual label grounded in the reality of data which refers to the “multiple, contradictory [definitions and] narratives, as of convergent, stable ones”³⁷ offered about ASTRID by the different stakeholders engaged in its design. The notion of project identity is embedded in a singular temporal dynamic, or “biography” (bringing together the past, present, and future). Identity is thus never fixed; it is a process that evolves over time, with events and technological and/or organizational developments punctuating its course.

^p This expression of Super-Superphénix is quoted from an interview with a French Parliamentarian who explained to us in 2016: Let me tell you a little anecdote. I had a request for an appointment with Mr X [a proponent of ASTRID] who comes to talk to me about Generation IV. He gives me his demonstration and I tell him: ‘But it’s a Super SuperPhénix that you’re building’. – Ah no, no, not at all, it has nothing to do with it’ – ‘Well, you will go before the public, you will go before the press, you will explain that it is not Superphénix, and you will see, in public debate … I wish you a happy time!’

^q Data confirm that this ambitious deadline was suggested by senior CEA officials precisely because it would lead to SFRs as the de facto technology choice for the project.

^r “Nuclear time” refers to the timeframes of lifecycle of nuclear technologies (development, operation, dismantling), which span decades.

^s The literature on megaprojects insists on their recurrent inability to meet the schedules. But in the case of ASTRID, the situation is even more subtle: The ASTRID project can be defined as a hybrid project ³⁷ that tries to combine properties of exploitative and exploration projects, which makes the construction of the schedules itself very complicated.