Reliability reallocation models as a support tools in traffic safety analysis.

One of the essential questions placed before a road authority is where to act first, i.e. which road sections should be treated in order to achieve the desired level of reliability of a particular road, while this is at the same time the subject of this research. The paper shows how the reliability reallocation theory can be applied in safety analysis of a road consisting of sections. The model has been successfully tested using two apportionment techniques - ARINC and the minimum effort algorithm. The given methods were applied in the traffic safety analysis as a basic step, for the purpose of achieving a higher level of reliability. The previous methods used for selecting hazardous locations do not provide precise values for the required frequency of accidents, i.e. the time period between the occurrences of two accidents. In other words, they do not allow for the establishment of a connection between a precise demand for increased reliability (expressed as a percentage) and the selection of particular road sections for further analysis. The paper shows that reallocation models can also be applied in road safety analysis, or more precisely, as part of the measures for increasing their level of safety. A tool has been developed for selecting road sections for treatment on the basis of a precisely defined increase in the level of reliability of a particular road, i.e. the mean time between the occurrences of two accidents.