Socio-Technical Systems-Theoretic Analysis of Industrial Processes

Over the last decades, hazards and potential accidents (near miss) for new types of industrial accidents increased severely. Safety management systems must be focused on considering different inter- and intra-organizational elements to analyze accidents in view of the interactions among different organizational levels. Traditional safety risk methods are rooted in linear causal analysis of incidents and accidents. Systems theory can help dealing with the limitations imposed by such reductionist analysis techniques, when coping with increasingly complex systems. This study adopts the Systems-Theoretic Accident Model and Processes (STAMP) as a new accident causation model based on systems theory. STAMP is used to gain a larger understanding of the system at hand, and it is complemented with network analysis, obtained semi-automatically from the relationships identified through the model. This joint analysis is meant to identify critical system agents, and and drives to the definitionof leading indicators which could improve decision making. This model is focused on a demonstrative use case referred to an industrial plant for oil and gas production. The paper provides firstly a hierarchical safety control structure, and a preliminary agent network analysis where network metrics are used to study technical, human, and organizational elements.