Improving scour risk management for Iowa by utilizing improved flood frequency estimates and modified HYRISK

Two approaches can be utilized for determination of the design flood discharge depending on the presence of the USGS gages in the streams that bridges are built on. For bridges on gaged streams, the annual peak discharge data can be used in order to estimate the design discharge. For ungaged sites however, such data is not available and the only possible method is the use of regional flood frequency models that are developed based on estimations of gaged data. The high prediction errors associated with such models along with the fact that climate change and urbanization can also undermine accuracy of the estimated discharges, motivated this research. It was assumed that the most recent available flood prediction model was used by Iowa DOT for determining design discharge at the time of bridge construction. In this regard, the estimated bridge design discharge was compared with the latest estimate of the same flood event. The results showed that as the basins get larger, the estimated discharges are more reliable. It was also concluded that bridges built before 1980s are more prone to experience an increase in their estimated discharges. Floods and resulting scour are responsible for about half of bridge failures in the United States. Catastrophic consequences of bridge failures along with guidelines from the Federal Highway Administration (FHWA) motivated the development of scour assessment tools. HYRISK is one of the available tools for network-level scour analysis and is developed by the FHWA for prioritizing bridges based on their expected scour risk. This study proposed three major modifications to improve and customize HYRISK estimations for Iowa. Soil erodibility was incorporated into the HYRISK along with a modified failure cost calculation accounting for scour countermeasure installation cost rather than bridge reconstruction that was originally being considered. The modified HYRISK was used to estimate the annual cost of scour risk for Iowa DOT bridge network and also the damage to the affected bridges by the 2008 flood in Upper Mississippi River basin. The results were significantly different from original HYRISK estimations and were in line with the actual annual expenditure on scour maintenance program and also the reported damage from the 2008 flood