Trust in corrosion analysis for canister materials: status and challenges

The stability of canister and geotechnical barrier materials are crucial aspects for the of long-term safety in the established concepts of a multi-barrier system for deep geological repositories of high-level nuclear waste. However, to build confidence in numerical models used to estimate the lifetime of a canister in the safety case, the right experimental data is often a key aspect.In the context of considering the long-term stability of copper as a container material for the Swedish waste management program, a study from 1986 questioned the thermodynamic stability of copper under anaerobic conditions on the basis of experimental data [1]. This gave rise to a scientific debate that continues to this day and includes various corrosion processes [2]. This prompted the Swedish waste management organization SKB to carry out another comprehensive experimental and theoretical research program to support the numerical model-based prediction of wall thickness loss in copper canister materials [3]. Even today, the investigation of potential corrosion processes that have an influence on material stability in the detection period has not yet been completed internationally. For a timely development of canister materials for the outer wall (in contact with the geotechnical barrier) a sound strategy for analyzing corrosion in an application-relevant environment should be pursued from the outset. At the current status of the German site selection act, the geochemical environment as well as the electrolyte in contact with the canister and surrounding buffer material are not defined yet. However, a change in geochemical conditions over longer time is expected and it will affect the long-term behavior of the canister and buffer materials. The controversy regarding the corrosion of copper in the Swedish disposal program makes it clear how crucial it is to carry out meaningful experiments to support safety analyses. But what experimental data is needed to support each model? How must corrosion experiments be designed and results analyzed and interpreted in order to support a model with confidence?As an impulse for further discussion, the contribution will review the controversy regarding the corrosion of copper and place it in a current context of existing knowledge on corrosion processes and their modeling in the repository context. Furthermore, own lab-based experiments in two different approaches A. in compacted Bentonite and B. in Bentonite slurries with synthetic pore water compositions are discussed in the context of corrosion under application relevant environments.[1] G. Hultquist, Hydrogen evolution in corrosion of copper in pure water, 1986, Corros. Sci., 26, 173–177[2] Entsorgungskommission, Diskussionspapier, „Diskussionspapier zur Kontroverse um die Verwendung kupferbeschichteter Behälter für die Endlagerung hochradioaktiver Abfälle“, 2021[3] Supplementary information on canister integrity issues; SKB Technical Report TR-19- 15, March 2019