Quantifying Condition Indicators for Concrete Structures

Authors

Tina H. Johnsen, Mette R. Geiker and Michael H. Faber

Abstract

The present paper introduces a risk-indicator based approach to inspection and maintenance planning. Its applicability is illustrated on an example considering probabilistic deterioration models relevant for reinforcement concrete highway bridges subjected to chloride induced corrosion.

Probabilistic models for Non Destructive Evaluation (NDE) are formulated representative for two type of inspection methods 1) half-cell potential mappings as inspection method for the indication of initiated corrosion and 2) visual inspections used as indicator of progressed corrosion of the stirrups.

Probabilistic modeling is applied both for quantifying the quality of the inspection method and to modeling the deterioration of the concrete. The data applied in the probabilistic modeling are both experience based and from in-depth inspections covering half-cell potential values, observations from excavations, and chloride profiles. For statistical analysis the data are classified reflecting similar concrete quality and environmental exposure.

The quality of the performed inspection method may be adequately quantified by means of the probability that the considered procedure indicates deterioration given that deterioration is actually present and in addition to this the probability that the procedure indicates deterioration given that there is no corrosion. It is shown how the optimal choice of the critical potential reading may be formulated as a decision problem integrated with the decision problem of inspection and maintenance. The optimization of inspection and maintenance strategies has been undertaken for two simplified decision models illustrating the merits of preventive (low-price) repairs carried out at an early state compared to extensive (costly) repair carried out at a later state.

Published in/by

Concrete International, ACI, 25(12), December 2003, pp. 49-56.