In earthquake-prone areas, seismic induced damage to bridge structures can result in a loss of road network functionality. The seismic risk assessment of bridges is crucial to establishing response strategies and restoring road networks, which in turn ensure the continuation of essential services such as access to hospitals and the delivery of relief assistance. In order to create an efficient management system and to optimize the decision making process, it is important to study the impact of seismic scenarios on the state of the network and the potential economic losses. The objective of this research is to generate seismic scenarios in order to evaluate the economic impacts and uncertainties of different input models on a bridge network. Four steps are proposed. The first step is the interpretation of a seismic hazard model to generate a spatial distribution of earthquake intensity for different seismic scenarios. A regression analysis is performed on ground motion prediction equations compatible with Eastern Canada, to obtain a relationship between the maximum ground acceleration and the geographic position of each bridge. Based on the Canadian Highway Bridge Code, the results are corrected using the site coefficients to account for the seismic amplification effect. The second step is to build a bridge inventory according to a classification defined by information on construction materials and structural systems. A statistical analysis is performed on each bridge class to study the different parameters contributing to seismic vulnerability. The third step is to assess the seismic performance of the bridges in terms of damage probabilities; this includes specific fragility data, respective to the different classes, based on dynamic analyses. The last stage uses the results of the previous stage to define an impact model to estimate the probable damage state, the repair cost, the priority of inspection and thus the probable state of road traffic. The results of the seismic scenarios reflect the importance of considering magnitude 5.0 in damage assessment and indicate that the standard deviation of the damage ratio can greatly influence the estimation of economic losses. The models are then integrated in a spreadsheet software tool to develop rapid seismic risk, epistemic uncertainties assessment and visualize the results on a geographic information system platform. Different input models are considered: three seismic hazard models, three fragility models and two impact models. A sensitivity analysis performed with magnitude 6,0 scenario for multi-span bridges shows that the economic loss estimate is very sensitive to the variation of these different models, where the expected increase in repair costs can reach 760 %.
Fezai, H. (Author),
Nollet (Supervisor) &
Abo El Ezz (Co-supervisor),
23 Nov 2020Student thesis: Master's thesis › Master in Engineering: Construction Engineering