An integrated bridge asset management method combining condition degradation, load rating, serviceability and risk-cost

Most road authorities only collect discrete condition data for bridge components at the overall network level. These data collected through regular inspections are used to make maintenance decisions on the bridge network in a reactive practice. In recent times, predictive modelling of condition change of bridge components using historical data has been used to forecast the future conditions, leading to a proactive management strategy. However, this approach is typically separate to the capacity assessment of bridge structures and identifying the serviceability limit state. A major objective of the research presented here is to develop a link between the two approaches by establishing the relationship between the bridge component condition and the capacity and serviceability limits of the structure. A comprehensive review of literature identified methods of predictive modelling of bridge component condition change using different approaches. The review also indicated that a key to the proposed linkage is considering specific degradation mechanisms in predicting bridge deterioration. The research program developed and executed, covered four main areas: predictive modelling of bridge component condition change with time using the level 2 inspection data collected through the road authority of Victoria, Australia (VicRoads), development of a methodology to modify the load rating calculation based on the forecast condition of bridge components for specific mechanisms of degradation, linking serviceability limit state of bridges as indicated by cracking to the condition change and calculation of the life cycle risk-cost of the bridges to aid in decision making. The research program included a literature review, VicRoads level 2 inspection data collection and processing, the development of discrete data driven models using the probabilistic Markov Chain process, and a load rating case study of a prestressed concrete girder bridge. Further research has involved developing an integrated model to link load rating and reinforcement corrosion for reinforced concrete bridges, together with the condition based data-driven model, in order to give an indication of load capacity reduction over time. This was done through a process that includes capacity reduction factor calibration and reliability analysis using Monte Carlo simulation. Also, another case study of a u-slab bridge was conducted, where Structural Health Monitoring (SHM) was used to validate a Finite Element Model (FEM). The FEM with flexural cracking was then linked with the Markov model, in order to determine cracking over time which enabled identifying the threshold for serviceability failure. An integrated probabilistic risk-cost procedure is also developed to aid in network-level budgeting. The anticipated outcomes of the project include the optimization of inspection and maintenance activities conducted by road authorities and enhancing the level of service provided by the bridges.