Optimisation of preventative maintenance strategies for reinforced concrete bridges

Bridge structures have a vital role to play in the transport system.   Unfortunately, like all engineering structures they are constantly deteriorating with time due to various causes such as metallic corrosion.   A well-planned and implemented maintenance program can help to secure maximum service life at an acceptable cost.   Preventative maintenance (PM) measures can often help to prevent, or postpone, the onset and progress of deterioration.   However, there are currently significant uncertainties associated with the need for these maintenance actions and their degree of effectiveness.   The aim of this project is to develop a methodology, which will set a basis for finding optimum PM strategies while incorporating the uncertainty of their effectiveness. A review study was undertaken first to study the state of the art in the area of bridge maintenance management particularly in the UK and optimisation techniques.   Based on the findings of this study a probabilistic procedure has been developed to incorporate the uncertainties and estimate the effectiveness of various PM measures by producing probability of failure (pf) profiles related to the initiation of corrosion.   To link the pf profiles of PM with their cost, and produce optimum PM strategies, an optimisation methodology using the principle of Genetic Algorithm (GA) has been developed.   A spreadsheet - program was created using Excel XP and Visual Basic Programming 6 (VBA 6), to demonstrate the methodology and enable analyses to be performed efficiently.   This provides a tool capable of identifying optimum PM strategies for Reinforced Concrete (RC) bridges, by delaying the onset of, or stopping, the corrosion process due to the ingress of chloride based de-icing salt.   Sensitivities studies were also performed to identify significant trends and assess the relative importance of the various parameters involved. The results are promising since optimum PM strategies can be produced even though these are dependent on the input data.   In order to produce more realistic strategies and apply this methodology effectively, there is a need to enhance the quantity and quality of the available data.   It is concluded that, these strategies can be used as an aid to gain better insight into the effectiveness of PM measures, and their use in planning the protection of bridge element from chloride ingress.