Corrosion rate of copper canisters used for final disposal of nuclear waste, in synthetic ground water and bentonite clay pore water

In this paper we study the corrosion of copper canisters used for final disposal of nuclear waste, in synthetic ground water and bentonite clay pore water. Both waters were studied when in equilibrium either with air or low oxygen atmosphere. The purpose of the research was to study the effect of air-formed oxide films on the general corrosion of copper. The samples were oxidized in air at a maximum temperature of 100°C and maximum time of seven days. Corrosion was studied using long-term weight loss tests up to 20 months, polarization resistance monitoring (LPR) for up to 5 weeks, and quartz crystal microbalance (QCM) for up to 5 days. In the weight loss tests corrosion rates were highest in air-purged ground water, 6-8 μm/year, and lowest in nitrogen-purged pore water, less than one μm/year. In polarization resistance tests the corrosion rates were 2-4 μm/year in air-purged ground water, and less than one μm/year in air-purged pore water. Nitrogen-purging resulted in higher corrosion rates than air-purging indicating that oxygen is not the only oxidant affecting the system and that concentration of dissolved oxygen can also affect passivation. The QCM tests in ground water showed high corrosion rates in the order of 10 μm/year or more, but in pore water from less than one μm/year to 2 μm/year. The oxide films showed a slight protective effect that decreased with time in ground water but improved in pore water. In both waters the corrosion rate of oxidized samples was initially 75% of the non-oxidized ones. During the immersion the corrosion rate of oxidized samples increased to 90% in ground waters but decreased to 30-60% in pore waters when compared to non-oxidized samples. Because of the small corrosion rates the experimental errors are large, and the protective effect cannot be estimated with certainty. On the other hand, the possible effect of oxide films to increase corrosion rate can be excluded.