Restoration models for quantifying flood resilience of bridges

Bridges are the most vulnerable assets of our transport networks. They are disproportionately exposed to and hit by multiple natural hazards, with flooding being the leading cause of bridge failures in the world. Their performance is constantly challenged by the combined effects of natural hazard stressors, e.g. flash floods, exacerbated by climate change, ageing, increasing traffic volumes and loads. Bridges are vulnerable to scour and other flood-related impacts, such as hydraulic forces and debris accumulation. In order to assess and quantify the resilience of flood-critical bridges and subsequently deploy bridge resilience models aiming at building resilience into transport networks, it is essential to use reliable fragility, capacity restoration and traffic reinstatement metrics and models. It is surprising that, despite the importance of bridges and their high vulnerability to hydraulic actions, there are no available recovery models. The latter can help quantify the pace of post-flood capacity and functionality gain for facilitating well-informed decision making for reliable prioritisation and efficient allocation of resources in transport networks. The main barrier is the nature and complexity of recovery actions, which encompass engineering, operational, owner resources and organisational challenges, among others. This paper, for the first time in the international literature, aims at filling this gap by generating a set of reliable recovery models that include both bridge reinstatement (traffic capacity) and restoration (structural capacity) models based on a detailed questionnaire that elicits knowledge from experts. Recovery models are then presented and validated for spread and deep foundations for a typical reinforced concrete bridge, including restoration task prioritisation and scheduling, inter-task dependencies, idle times, durations and cost ratios for different damage levels, as well as the evolution of traffic capacity after floods.