| Summary: | Many UK nuclear power stations are reaching the end of their planned lives and seeking extensions. Simultaneously, electricity demand is increasing, older coal-fired stations are required to shut and emission targets restrict like-for-like replacement. The reliability and safety of ageing nuclear stations is of increasing importance, but despite preventative efforts, significant, disruptive events still occur. Nuclear power stations were identified as Complex Socio-Technical Systems with tightly linked, complex interactions and behaviours controlled by feedback. Contemporary theories suggest that the current approach to prevention does not effectively address these characteristics. Instead, it views events as cause-and-effect chains, often focusing on prediction at the expense of developing resilience. Systems Theory, particularly Soft Systems Methodology, was used to understand the complex nature of the problem and assist in the investigation of an event analysis technique which would complement the current tools, addressing their shortfalls. System Dynamics was investigated for learning from events within the nuclear industry. It was applied in three case studies to identify the technical, cultural and organisational causes of events. Quantitative and qualitative models were built, and simulations performed. Uniquely, it was applied in an active event investigation, and Group Model Building techniques were investigated. The conclusions of each case study were compared against those of investigations using the traditional methods, assessing System Dynamics' ability and practicality in this context. In each case System Dynamics allowed for conclusions and recommendations additional to those of the current methods. It highlighted the causal influence of feedback structures within the system, which in some cases had caused a gradual drift towards vulnerability. It led to recommendations that alter interactions within the system, making it less vulnerable to unexpected variation, and therefore more resilient. This study suggests that implementing System Dynamics in event investigation is feasible and advantageous in the nuclear industry and beyond.
|
|---|
