Practices of incident reporting in a nuclear research center: A question of solidarity

•We made 28 semi-structured interviews with agents from a Nuclear Research Center.•We analyzed the interviews via a constructivist vulnerability-oriented approach.•Actual incident reporting (IR) practices are more varied than the formal ones.•IR practices reflect a pragmatic solidarity-based interpretation of the procedure.•Other IR practices tend to be contested.

Incident reporting is usually considered as an effective means to improve the safety of “at risk” socio-technical systems (e.g. nuclear plants, large industrial facilities, hospitals), as it allows implicated actors to learn from past incidents. Safety could thus be enhanced via the use of an institutionalized Incident Reporting System (IRS), enabling organizations to improve the quality of actions and reactions in case of a deviation from normality, or to prevent such deviations from happening in the first place. Yet, there is a lack of inductive analyses of actual, on-site uses of IRS. In this paper, we address this gap, using the results of 28 semi-structured interviews conducted with agents from the Belgian Nuclear Research Center (SCK·CEN). The study relies on a vulnerability-oriented Science and Technology Studies (STS) approach. Our results show that practices of incident reporting are more varied than the institutionalized ones. Indeed, actual reporting practices are to be related to specific expressions of solidarity between colleagues within a negotiated drift – a pragmatic interpretation of the reporting procedure. These results are discussed in a vulnerability-oriented perspective. Overall, the paper displays a grounded analysis of incident reporting practices which may contribute to a better understanding of how safety is co-constructed by workers, and provides opportunities for further research and concrete path of actions for practitioners.