The Importance of Considering the Service Environment When Studying and Predicting the Performance of Corrodible Structures

It goes without saying that when studying the corrosion behaviour of a component or structure, the experimental conditions should reflect the service environment to which the object will be exposed. However, all too frequently, “accelerated” conditions are used, involving applied potentials, elevated temperature, high solute concentrations, excessive strain or strain rates, etc., which complicates the prediction of the in-service behaviour or component lifetime. At best, it is necessary to extrapolate the results of these accelerated laboratory measurements to more realistic conditions, ideally based on a mechanistic understanding of the processes involved. At worst, accelerated laboratory tests may suggest corrosion processes that are not feasible or relevant to the service environment, potentially disqualifying a given material or design from consideration that would otherwise provide acceptable behaviour in service. Examples of the need to properly take into account the service environment and the potential negative consequences of accelerated testing are given for the case of the corrosion behaviour of nuclear waste container materials. For example, the use of bulk solutions to study the corrosion of copper by sulfide in the laboratory involves high sulfide fluxes and leads to localized corrosion and stress corrosion cracking mechanisms that are not possible under actual repository conditions. Similarly, accelerating the effects of γ-irradiation using high absorbed dose rates runs the risk of changing the mechanism of radiation-induced corrosion. Above all, it is imperative to develop a sound mechanistic understanding of the underlying corrosion mechanisms in order to confidently apply the results of short-term laboratory observations to the prediction of the long-term performance of nuclear waste containers.