Bridge Damage, Repair Costs, and Fragilities for Inland Flood Events

Flood events bring havoc to transportation assets including bridges, culverts, and roadways on an annual basis. This has been an increasing trend in the recent years, especially in Midwestern states. Millions of dollars are spent to repair or replace transportation infrastructure every year due to damage suffered from floods. With a major drive within transportation agencies to consider the risks associated with transportation asset damages, and provide optimized plans for pre-event mitigation and postevent recovery, there is a need for better estimation of the type of damages observed after inland flood events and the associated costs for the repair and recovery of these damages. Most previous studies have focused on recovery after earthquakes, which by nature imposes a different organizational and response effect on the agencies. The limited number of studies on other hazards such as coastal flooding is based on sparse survey data. This paper specifically aims to address the existing gap in the availability of data on damages and costs that are observed after actual inland flood events. For this purpose, the damage patterns associated with scour, high water levels, debris accumulation, abutment wash out, and other damages have been identified. The extreme flood events throughout a 16-year span (1998–2014) that were documented as part of the emergency relief funds have been considered. Repair tasks associated with each damage mode are identified based on actual project documents and costs are broken down by each of these tasks. This component by itself is a leap forward in collection of types of damages and associated costs for inland flood events. In addition, generalized linear models are conducted to identify the relationships between flood stage, bridge characteristics, and damage level using both Gaussian and binomial distributions of the responses. Analyses shows that the maximum water elevation at a bridge location is statistically significant in determining the probability of bridge damage. There are also many other bridge attributes including length, width, skew angle, and pier width that have a correlation to different damage modes of bridges. A series of National Bridge Inventory condition ratings are analyzed with the scour criticality rating and superstructure rating both being significant predictors in the probability of multiple bridge damage modes. This paper is the first known effort aims at identifying the type of damages associated with inland flood events, the associated repair tasks, and costs of recovery through a coordinate data collection effort from actual events and completed projects and is expected to majorly contribute to future studies aimed at assessing the resilience of transportation infrastructure under inland flooding events by providing better estimates of the recovery tasks and costs.