Electronic Thesis and Dissertation Repository


Doctor of Philosophy




Dr. Jungsook Clara Wren


This thesis presents a newly developed mechanism and predictive model for the corrosion of Alloy 800. The Fe-Cr-Ni Alloy (Incoloy 800) is mainly used for steam generator (SG) tubing in CANDU and PWR reactors and is a candidate material for the proposed Canadian Supercritical Water Reactor (SCWR) in which it will be exposed to extreme conditions of high radiation flux and large temperature gradients. The influence of gamma radiation and water chemistry conditions on the corrosion behaviour of Alloy 800 are studied in this work. Ionizing radiation creates reducing (•eaq, •H, •O2-) and oxidizing radiolysis (•OH, H2O2, O2) products that affect the redox chemistry, controlling corrosion. Water chemistry conditions including pH, temperature and redox agents can significantly influence the corrosion kinetics. A systematic study of Alloy 800 corrosion was carried out to investigate the effect of these solution conditions. This analysis was used to develop a mechanistic model that takes into account both metal dissolution and oxide formation during the corrosion of Alloy 800. This model is designed to predict the effect of different variables on the corrosion behaviour of Alloy 800 in extreme environments where direct corrosion measurement is nearly impossible.

A series of electrochemical experiments and corrosion tests along with post-test surface analyses were performed in order to gather information on the composition and thickness of the oxide formed during corrosion and the metal cations dissolved in the solution. This combination of electrochemical measurements and surface analyses provided a highly-detailed understanding of Alloy 800 corrosion, allowing a mechanism to be proposed. The proposed mechanism can explain the corrosion behaviour of Alloy 800 in a variety of environments and temperatures, including aqueous and steam corrosion.

The principles behind the proposed mechanism were used to develop a model to account for both oxide formation and metal cation dissolution. The model was used successfully to model oxide thickness on pure iron, the Co-Cr alloy Stellite-6 and Alloy 800 in neutral and moderately alkaline aqueous solutions. The modeled results correlate well with experimental data. Using the model, it was possible to predict the time-dependent corrosion potential in environments where direct measurements are not possible.