Abstract for: The Use of Group Model Building for Analysing Event Causality within the Nuclear Industry

Currently the most frequently used tools for investigating and learning from industrial accidents are based on linear, reductionist models of event causality. It has been suggested that complex socio-technical systems exhibit non-linear behaviour, incompatible with these approaches. An alternative approach based on system dynamics group model building is proposed and investigated within the context of the civil nuclear industry. The success and feasibility of the approach is assessed though the analysis of a case study event by a group of industrial experts previously unfamiliar with system dynamics. A selection of the models produced is discussed. The investigation demonstrates the feasibility of the approach and its potential to provide additional insights and learning. A possible archetypal structure is also identified and described indicating the possibility of the discovery of further archetypes through this approach which could then be shared for contextual learning between organizations and industries.