Abstract
User cost models are used in bridge management systems to quantify, in economic terms, the potential safety and mobility benefits of functional improvements to bridges. One component of many such models is a formula to predict traffic accident risk on the basis of bridge characteristics such as roadway width and approach roadway alignment. A critical analysis of the user cost model used in AASHTO’s Pontis and Bridgit systems found that the model was overly sensitive to extremes of roadway width, yielding unrealistically high benefit estimates. A new model was developed using Florida data on bridge characteristics and traffic accidents. The new model has superior behavior and statistical characteristics on a full inventory of state highway bridges. Since it relies solely on National Bridge Inventory data items, the model is readily transferable to many different bridge management systems.