Aging infrastructure together with an increasing rate of extreme motivates many U.S. state departments of transportation to perform risk prediction and resource allocation analyses. However, such assessments are seldom validated with actual bridge collapse data. In this study, regional and peak flow conditions of 42 bridge collapse sites are derived and analyzed to assess the robustness of a risk assessment framework based on HYRISK, a model developed by the U.S. Federal Highway Administration to assess bridge collapse probability. HYRJSK's scour vulnerability and overtopping frequency risk factors are particularly assessed. Results suggest that risk assessment is not robust unless the analysis is specific to the (1) physiographic region's intrinsic scour risk, and (2) hierarchy of flow data for a heavy-tail distribution. The findings can be included in asset management systems to help state departments of transportation prioritize resource allocation in maintenance and replacement to reduce hazard risk.
展开▼