Estimating expected rates of transport fatalities requires quantitative risk assessment of widely differing scenarios. These scenarios can range from incidents that occur many times per year, with typically single fatalities, through to rare but catastrophic events where hundreds of lives may be lost in a single incident.
A problem for transportation system risk managers and regulators is how to prioritise improvement effort to reduce risk. Efficient deployment of effort requires a valid understanding of the systemic risks across the whole transportation mode. Commonly available data sets are discussed, along with pitfalls, and how those data sets can be analysed and integrated to develop a coherent and valid assessment. The methods are illustrated by a national rail risk assessment case study.
The novel feature of this approach is the integration of differing quantitative risk assessment (QRA) methods to build an overall system-wide quantitative risk assessment. This poses issues of consistency and reliability, which need to be overcome.
The case study describes how this approach was applied to develop a New Zealand national rail risk assessment covering national freight operations, passenger services, industrial operators and the tourist and heritage sector. A distinctive feature of the assessment is that it covered a wide range of incident probabilities, including some incident types that have never occurred on the New Zealand rail system. Another novel feature of the assessment is that it addresses both natural hazards and system operational risks. The case study describes the types of data sources accessed, the QRA methodologies chosen, and how those selections were influenced by factors such as the frequency of incidents, the rate of fatalities per incident, data quality and the pace of rail system safety improvements.
The case study also discusses responses to the finding that a largely unrecognised natural hazard arising from tsunami poses a significant rail transportation risk with potentially catastrophic outcomes.