The safety performance of complex and dynamic systems can be assessed by a probabilistic risk approach and a deterministic scenario approach. Based on a series of experiences with tunnel fires, public governance decision-making on risk acceptance and awareness in major infrastructure and transport developments, decision makers and designers are faced with limitations of the quantitative approach.

This paper focuses on the deterministic scenario approach. Scenarios of the deterministic approach describe accident processes of potential or actual sequences of events in time, in a specific context and operating environment. This paper focuses on an important difference between the deterministic scenario approach and the quantitative approach, in the notion of time. Not only time, but also accelerations and decelerations in time can become explicitly clear in scenarios. In the scenarios, also critical safety decisions and actors at different decision-making levels can be identified. Besides the physical scenario, accidents are also located in a policy decision-making scenario. This is all explained by the case study of the train collision at Amsterdam CS on the 21st of May 2004.