Loading...

Probabilistic Framework for Risk Analysis of Buildings Under Flood and Earthquake Hazards

Aghamohammadi, Mohammad | 2024

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 57148 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Mahsouli, Mojtaba; Safaei, Ammar
  7. Abstract:
  8. This research presents a probabilistic framework for risk analysis under the dual hazards of flood and earthquake using reliability methods. In this research the term “risk” refers to the exceedance probability of the maximum loss resulting from the flood and earthquake hazards over a given time period. This methodology is implemented for the multi-hazard risk analysis of the buildings of a virtual city. The Monte Carlo sampling method is used to propagate uncertainties and compute the probability distribution of the maximum loss under each hazard. Subsequently, the load combination theory is applied to integrate the loss exceedance probabilities under the flood and earthquake hazards. Probabilistic models used in this research include hazard and risk models. Flood hazard models consist of hydrological and hydraulic models. The hydrological model outputs the flood hydrograph given the probability distribution of the maximum annual river discharge, the relationship for predicting the flood volume, and the shape of temporal variations of discharge. The hydraulic model calculates the flood depth across a network of cells in the study area by receiving the flood hydrograph and hydraulic parameters, such as Manning’s and impermeability coefficients, for various land uses and Digital Elevation Model. For this purpose, HEC-RAS software is used for two-dimensional hydraulic simulations. Due to the substantial computational effort needed to perform an extensive number of two-dimensional simulations in Monte Carlo sampling, a metamodel is developed based on the decision tree method in machine learning, using the data of 3418 simulations. The resulting metamodel has a coefficient of determination of 0.99 on the validation data and runs nearly two orders of magnitude faster than the hydraulic model. Flood risk models assess the damage measure and the repair cost by receiving the flood depth at the location of each building, along with such information as the number of floors and occupancy type. The output of the flood risk analysis is the exceedance probabilities of the maximum annual loss. In turn, the seismic risk analysis framework was developed in prior studies and adopted herein. Earthquake hazard models include the magnitude, area and location of rupture, and ground shaking intensity models. Earthquake risk models compute structural, nonstructural, and content repair costs. This calculation incorporates the ground shaking intensity at each building location, lateral load-resisting system, construction year, and occupancy type. The output of the earthquake risk analysis is the exceedance probabilities of the loss given the occurrence of an earthquake, calculated for each type of seismogenic source model. Ultimately, the load combination module computes the exceedance probabilities of the maximum annual loss under both hazards. This computation takes into account the designated time period for risk analysis, the exceedance probabilities of the maximum annual flood loss, and the occurrence rate and the exceedance probabilities of the loss given the occurrence of an earthquake for each type of seismogenic source model. The proposed framework, along with its probabilistic models, are implemented in Rtx, which is software for reliability, risk, and resilience analysis with a comprehensive library of probabilistic models for various hazards, infrastructures, and consequences. The presented framework is showcased through a case study that features a virtual city, entitled INSURER City1. The results derived from the proposed framework bear significance for decision-making on a variety of policies aimed at mitigating the risk of floods and earthquakes. For instance, the share of each hazard from the total losses offers insights into the share of investments needed to mitigate the risk of that hazard. Furthermore, considering both hazards in a single risk analysis enables the assessment of various measures in terms of their complementarity or interference in mitigating the risk associated with each hazard
  9. Keywords:
  10. Floods ; Earthquake ; Risk Analysis ; Multi-Hazards ; Monte Carlo Sampling ; Machine Learning ; Hydraulic System Analysis ; Probabilistic Seismic Hazard Analysis (PSHA) ; Probabilistic Modeling ; Multihazard Risk Analysis

 Digital Object List

 Bookmark

No TOC