Predicting bridge deck performance using Markovian models

The life-cycle cost analysis of highway bridges requires reliable prediction of their performance in order to estimate agency and user costs over the entire service life. State-of-the-art bridge management systems, such as Pontis, have adopted Markov-chain models for predicting the future condition of bridge components, systems, and networks. Markov chains reflect the stochastic nature of bridge deterioration through the use of transition probability matrices. These matrices are developed based on the assumption that bridge inspections are performed at predetermined and fixed time intervals (i.e. constant inspection period). This paper evaluates the impact of this assumption on the performance prediction of bridge deck systems using field condition data obtained from the Ministére des Transports du Québec (MTQ). Transition probability matrices are developed for the different elements of deck systems and adjusted for the variation in the inspection period using the Baye's rule. Comparing the predicted performance before and after adjustments indicated that the variation in the inspection period may results in a 22% error in predicting the service life of bridge deck systems.