Pitting corrosion of copper. Further model studies

The work presented in this report is a continuation and expansion of a previous study. The aim of the work is to provide background information about pitting corrosion of copper for a safety analysis of copper canisters for final deposition of radioactive waste. A mathematical model for the propagation of corrosion pits is used to estimate the conditions required for stationary propagation of a localised anodic corrosion process. The model uses equilibrium data for copper and its corrosion products and parameters for the aqueous mass transport of dissolved species. In the present work we have, in the model, used a more extensive set of aqueous and solid compounds and equilibrium data from a different source. The potential dependence of pitting in waters with different compositions is studied in greater detail. More waters have been studied and single parameter variations in the composition of the water have been studied over wider ranges of concentration. The conclusions drawn in the previous study are not contradicted by the present results. However, the combined effect of potential and water composition on the possibility of pitting corrosion is more complex than was realised. In the previous study we found what seemed to be a continuous aggravation of a pitting situation by increasing potentials. The present results indicate that pitting corrosion can take place only over a certain potential range and that there is an upper potential limit for pitting as well as a lower. A sensitivity analysis indicates that the model gives meaningful predictions of the minimum pitting potential also when relatively large errors in the input parameters are allowed for