ABSTRACT
This paper presents a probabilistic method for predicting future performance deterioration of coastal flood defence structures based on the asset condition grading system and deterioration rates. The deterioration curves are used to obtain transition probabilities through regression based nonlinear optimisation algorithms. The Monte Carlo simulations are then used to evaluate the future performance of the structure on the basis of the estimated transition probabilities with consideration of the settlement of coastal defences. Results from the case study show that the proposed method can provide an effective predictive model for various situations by using available condition grading data. The proposed model also gives useful information on time-variant probability of failure of the coastal flood defences.
