Prediction of Service Life and Evaluation of Probabilistic Life-Cycle Cost for Surface-Repaired Carbonated Concrete

Concrete structures are economical and typically durable when exposed to a variety of environmental conditions. However, carbonation results in a reduction of the durability of reinforced concrete members because it damages the passive film surrounding the reinforcement, which accelerates corrosion processes and may ultimately lead to premature failure of the members. Predicting service life is complex because it depends strongly on changes in materials and environmental conditions. Existing carbonation models predict the service life based on deterministic theories. In this study, deterministic and probabilistic methods are applied to study concrete carbonation in the presence of repair materials using the maintenance periods and repair cost according to the coefficient of variation (COV) of the carbonation depth of each repair material. Water-based paint, organic alkaline inhibitor, inhibiting surface coating, and corrosion-inhibiting mortar (IM) were used as repair materials. An accelerated carbonation experiment using 20% CO2 was performed for 5 days, and then the repair materials were applied on the concrete surface. Then the samples were put back in the carbonation chamber and carbonation depth was measured after 7, 14, and 28 days. Based on the COV value, the carbonation depth and maintenance periods were predicted. These were used as parameters for the probabilistic life-cycle cost (LCC) model. Results showed that carbonation inhibition was best when the repair was done using IM. Cost results obtained from deterministic and probabilistic models were compared. When the probabilistic model is applied, the repair cost is evaluated as a curve, unlike with the deterministic model. The probabilistic model reduces the maximum cost by 50% compared to the deterministic model. As the COV decreased (indicating better quality concrete), the probabilistic model results approached those of the deterministic model evaluation. (c) 2020 American Society of Civil Engineers.