New interpretations of test observations for initiation of corrosion of steel reinforcement in concrete

A widely used criterion for the likelihood of initiation of corrosion of steel reinforcement in concrete is chloride concentration. Herein, recent test observations using samples of concrete extracted from practical structures claimed to support this notion are re-examined. The test results for the measured acid-soluble chloride concentrations arise from the electrochemical test conditions. These allowed the consumption of water in the initial cathodic reaction over several days, before the resulting local acid-soluble chloride concentration in the concrete was actually measured. The measured value was therefore not that at the point of initiation. Re-interpretation of the reported observations for initiation shows that they exhibit two-cohort probability distributions, similar to those known for the size of air-voids in concretes. The reported poor correlation between the location of rust spots and air-voids in the concrete at the steel–concrete interface may be attributed to only ‘visible’ air-voids being recorded, whereas actual concretes are known to contain a much wider distribution of void sizes. The reinterpretations given herein support the role of air-voids at the steel–concrete interface governing the likelihood of reinforcement corrosion initiation, but not the notion of a critical chloride concentration.