Risikoanalyse des Gefahrguttransportes : Unfallstatistische Risikoanalyse auf der Basis typischer Transportketten. Bericht zum Forschungsprojekt 8906/1 der Bundesanstalt für Strassenwesen BASt.

To improve the theoretical and empirical foundations of dangerous goods transport regulations, the German Bundesanstalt für Strassenwesen set up a research programme in 1992. The elements of the programme were as follows: 1. the performance of a historical risk analysis; 2. an analysis of the main weak points of the current dangerous goods transportation system; and 3. a system for continuous analysis of the risk involved in dangerous goods transport. The following study, representing the first step in this research programme, had three aims: (1) to develop a proposal for the risk model (step 3); (2) to examine and to document the existing statistical material concerning dangerous goods movement and accident data to fill any gaps in the existing data; (3) to calculate risk indicators using the historical analysis. A model was developed based on existing German and foreign studies and theoretical requirements. The main features of the model were its theoretical structure, the inclusion of different methods of data generation and the ability to regionalise the model. A method of including transport chains was also shown. This proposal was then used to analyse the data requirements of the model. The analysis of the data currently available and a comparison with the data requirements showed that at present information concerning transportation volume and performance and the regional distribution patterns in particular were not adequately covered. The situation regarding the recording of accidents was much better; most of the data required for the quantitative risk analysis were collected. Finally an explorative risk analysis for different goods (for the period 1987-1 991) was carried out, based on the current data. It was intended that the estimation of accident and loss rates could serve as input for the following phase of the programme (Step 3). The use of alternative estimations for traffic volume and performance also meant that the uncertainty deriving from the insufficient data was for the first time able to be quantified. Finally a comparison of our results with other national and international studies was carried out. (Author/publisher)