ABSTRACT

For many concrete infrastructural works a service life of 80, 100 or 200 years is required. To convince owners and authorities that these requirements can be met probability-based models for service life predictions have been developed. These models are based on theoretical and experimental laboratory studies. Many of these models focus on the probability of chloride-induced rebar corrosion. For a check of the reliability of one of these models, i.e. the DURACRETE model, predicted chloride profiles have been compared with chloride profiles measured in five marine concrete structures. Lessons learned from these existing structures in view of the reliability of numerical service life predictions are presented in this paper.