Data-Driven Life-Cycle Cost Analysis of Corrosion-Affected Elastomeric Bridge Bearing Assemblies Using Survival Models and Monte Carlo Simulation

Abstract

Elastomeric bridge bearing assemblies are critical structural components that facilitate load transfer and accommodate movements in bridge systems; however, their long-term performance can be susceptible to corrosion-induced deterioration, particularly in aggressive environments. Despite their widespread use, the economic implications of such degradation remain insufficiently quantified using data-driven and probabilistic approaches. This study presents a data-driven life-cycle cost analysis (LCCA) method that integrates survival analysis with probabilistic cost modelling to evaluate the long-term economic performance of corrosion-affected elastomeric bridge bearing assemblies. Inspection and element-level condition rating data from the FHWA National Bridge Inventory database are used to characterize deterioration behaviour. A nonparametric Kaplan-Meier estimator is first applied, followed by parametric survival modelling using candidate distributions including Weibull, lognormal, log-logistic, and hypertabastic models. The best-fit model, selected based on statistical criteria, is integrated into a Monte Carlo based LCCA to capture uncertainties in replacement timing, cost variability, and economic discounting. Results show that corrosion-driven replacement behaviour dominates long-term cost outcomes, with substantial variability across plausible scenarios. Sensitivity analyses indicate that discount rate assumptions and service-life uncertainty are the primary drivers of cost variability. The proposed framework provides a data-driven and risk-informed basis for bridge asset management and maintenance planning.