Road safety management activities include screening the network for sites with a potential for safety improvement (Network Screening), diagnosing safety problems at specific sites, and evaluating the safety effectiveness of implemented countermeasures. It is important that these activities be both efficient and methodologically sound, since resources would otherwise be wasted on unnecessary treatments for safe elements and elements deserving of treatment would be left untreated. The state-of-the-art methodologies for conducting these activities make use of statistical models to predict expected accident frequencies using traffic volumes and other site characteristics as the input to the models (known as Safety Performance Functions or SPFs). CDOT’s research and safety engineers are in the forefront of national efforts to develop methods using SPFs to screen large networks to find sites with a potential for safety improvement. CDOT has previously developed SPFs to identify freeway and rural roadway segments that have the potential for accident reduction. This report documents the data collection, modeling efforts, and findings of a research project to develop SPFs for ten categories of intersections. Implementation: The development of SPFs for ten intersection categories was successful. It is recommended that data for additional sites be collected as they may become available. Additionally, as more years of crash and traffic data become available, these data too can be added to the dataset in order to continually add up-to-date information. The models can be recalibrated to apply to these additional years of data. When several additional years of data are available, it may be desirable to calibrate a new set of original SPFs. (Author/publisher)
Abstract