Abstract

As industrial facilities age beyond their design horizons, their supply chains shrink, leaving
operators of critical infrastructure with component obsolescence, dependency on monopolistic
suppliers, and unpredictable lead times. Additive manufacturing offers a route toward operational sovereignty, replacing physical inventory with on-demand digital production, yet its adoption in safety-critical environments remains poorly understood. This thesis investigates how a high- reliability organization can adopt additive manufacturing in maintaining technically complex, long-lived facilities. Drawing on the People, Process, and Technology framework and a sociotechnical perspective, it examines adoption across three dimensions: component selection, reverse engineering, and digital integration. The empirical basis is a qualitative case study of Vattenfall’s nuclear division: eighteen semi-structured interviews at Forsmark and Ringhals, plus document analysis. Findings show that procurement operates reactively, generating maintenance debt and weakening Vattenfall’s position toward suppliers. Reverse engineering relies heavily on tacit expert judgment at critical stages, resisting the near-automated narrative in the literature. Digital integration is blocked by an aggregation problem, diverging security frameworks between the nuclear sites and corporate IT, and an absence of organizational ownership. Together, these reveal that a secure digital warehouse is not a downstream task but a precondition for secondary adoption. In response, the study proposes a seven-gate flag system for component selection and an eight-stage reverse engineering workflow built around bounded codification: specifying the architecture while shielding the stages whose reliability depends on expert judgment. The central theoretical contribution is that organizational rigidity need not be overcome for adoption to proceed, but can be redirected.