Abstract:

In order to improve and develop the accident causation theory, the four aspects of STPA and 24Model, the underlying theory, relevant definitions, analysis process and analysis difficulty, were compared and analyzed, and applied to different cases respectively. The results of the study show that both models have a strong theoretical foundation. The research object of STPA is the loss of system. It performed better at system interaction study, but the complexity is high and the research error is large, so it is more suitable for single-accident analysis. In contrast, the research object of 24Model is event in the organization. The classification of accident causes is clearer and more universal. The model is simple, the research error is small, and it is commonly used for both single-accident and multi-accident analysis. Both methods are of good extensibility. In terms of application, STPA is clear on the connection between hazards and losses. However, this method is complex and subjective on accident causation, which also lacks research on inter-organizational interactions. In contrast, 24Model is more comprehensive and clearer, but it fails in the analysis of inter-organizational interaction and design defects.

Key words: systems-theoretic process analysis (STPA), "2-4"model(24Model), accident causation theory, comparative analysis