Inland waterway transport: modelling the probability of accidents

In 1989 the Dutch government started the project “Safety in Inland Waterway Transport” to establish a minimum safety level and to develop a model to assess the effect and effectiveness of new safety measures. This model is called the Risk Effect Model, and calculates the integral impacts of safety measures for the entire waterway system including the risks of transporting dangerous goods. The final result of the project is a framework for evaluation, which supports cost-benefit analysis by weighing negative economical effects against achievements in safety, for different measures. In this paper we will present the methods which have been used to calculate the probability of an accident, using casuistry. In this project the probability of an accident is modelled per elementary traffic situation (a combination of several ships carrying out a ship motion produces a traffic situation). The number of accidents can be estimated by the number of elementary traffic situations multiplied by the probability of an accident per elementary traffic situation. In the paper we describe fitting procedures in order to obtain the model that “forecasts” the probability of accidents as function of waterway attributes and circumstances. We have used Generalized Linear Models (GLM), which do not need the assumption that the accident probability is normally distributed. We have used the binomial approach in the GLM models. We present the results of the fitting procedures for one group of elementary traffic situations (through-going traffic mutually). The primary governing variables appear to be visibility, wind speed, the ratio of the navigable width and the necessary width for an elementary traffic situation, and the bend radius of the waterway. The circumstances (visibility and wind speed) are more explanatory with respect to the probability of accidents than the waterway characteristics are