DEVELOPING A SAFETY PERFORMANCE FUNCTION TO PREDICT THE LIKELIHOOD OF WEATHER RELATED CRASHES ON WISCONSIN BRIDGES
Motl, Alexandria R.
MetadataShow full item record
As a northern tier Midwestern state, Wisconsin typically faces challenging travel conditions, particularly during winter months. Weather conditions can be especially problematic on and around bridges. Pavement conditions and bridge deck surfaces can be prone to friction issues. Bridge locations are generally prone to ice formation in winter weather conditions. Therefore, there is a need to address bridges that are especially prone to friction related issues in winter weather. Understanding what factors contribute to weather related crashes on bridges will help decision makers understand where countermeasures can be most useful. Therefore, a Safety Performance Function (SPF) was developed to predict the likelihood of weather related crashes on bridges. Variables determined to be significant predictors of these crashes include: the Average Daily Traffic crossing the bridge (log transformation was used in the SPF); the number of directions of travel carried by the bridge; the facility over which the bridge crossed (a water facility, another roadway, or a railroad); the number of lanes on the road under the bridge; the bridge length (the log transformation was used in the SPF); the year of construction; and the bridge sufficiency rating. Variables that were not considered significant predictors for these crashes include: bridge deck material; bridge width; bridge area; the number of lanes on the bridge; and the railing type on the bridge itself and the approach to the bridge. The SPF was used to predict the number of weather related crashes on bridges over a five year period. Bridges were ranked in order of the highest likelihood to lowest. This list can be used to select bridges to receive safety countermeasures, such as de-icing and anti-icing operations, as well as High Friction Surface Treatment (HFST) applications. It can also be compared to other prioritization ranking lists to ensure that a selected bridge meets multiple criteria.