Prediction of Long-Term Corrosion Damage in High Level Nuclear Waste Disposal Systems

Assuring the public of our ability to safely isolate High Level Nuclear Waste (HLNW) from the ecosphere over the long term represents one of the greatest technical challenges ever to face humankind. The principal difficulty is that the service life horizon is twice as far into the future (10,000 years) as recorded human history is in the past, and no direct experience exists with the degradation of engineering materials over that time period [1]. Indeed, some of the materials that are proposed for use in HLNW engineered barrier systems did not exist thirty years ago (e.g. Alloy C-22). A search of the literature reveals that no material that is currently contemplated for HLNW service has been investigated in a single corrosion study for more than a few tens of thousands of hours, or for more than a few one hundredths of one percent of the intended service life. Not only is the corrosion rate known over only a very small fraction of the service life, we cannot be sure that the mechanism of attack will remain the same as the system passes along the corrosion evolutionary path (CEP). Because the uncertainty of prediction in an empirical protocol scales with the power of each of the independent variables in the Dependent Variable (DV)/Independent Variable (IV) correlation, it is evident that the uncertainty grows rapidly over extended extrapolation. For example, the rate of propagation of a stress corrosion crack increases exponentially with potential, so that any uncertainty in the corrosion potential translates into very large uncertainty in the crack length over an extended extrapolation. The purpose of the present White Paper is to outline some of the important philosophical and practical issues that exist in the prediction of future behavior of physico-chemical systems in general and HLNW canister corrosion in particular. These issues are not just of academic interest, but will prove to have profound impact on how the public perceives, and indeed whether it accepts, YMP’s predictions of canister corrosion damage, and hence whether it feels assured that HLNW can be safely contained within the Yucca Mountain repository.