Probabilistic assessment of seismic vulnerability and retrofitting decisions using bayesian analysis for reinforced concrete structures

Abstract

Seismic vulnerability assessment of existing reinforced concrete structures remains a critical challenge in earthquake-prone regions, where uncertainties in material properties, structural capacity and seismic demand significantly influence decision-making processes. This study introduces a robust Bayesian framework for the probabilistic seismic assessment and retrofitting of reinforced concrete (RC) building and bridge foundations. The methodology synthesizes prior information from design codes, historical evidence and expert insight with in-situ measurement data to iteratively refine the probabilistic characterization of vital structural parameters. Utilizing Markov Chain Monte Carlo (MCMC) sampling techniques, this study elucidates the derivation of posterior distributions for foundational geotechnical parameters, notably soil bearing capacity, fragility curve parameters and peak ground acceleration can inform risk-based retrofitting strategies. A case study reveals that Bayesian updating reduced the failure probability from 23.2% to 4.3% post-retrofit, with a benefit-cost ratio of 7.58, validating the economic efficiency of the proposed approach. The framework provides engineers with a rational, probabilistic tool for continuously updating structural safety assessments as new data becomes available, ultimately enhancing resilience in earthquake-prone communities. This research advances the broader discourse in performance-based earthquake engineering (PBEE) by proposing an applicable framework that systematically incorporates both epistemic and aleatory uncertainty into seismic risk quantification.