Reproduction of natural corrosion by accelerated laboratory testing methods

Various laboratory corrosion tests have been developed to study the behavior of glass waste forms under conditions similar to those expected in an engineered repository. The data generated by laboratory experiments are useful for understanding corrosion mechanisms and for developing chemical models to predict the long-term behavior of glass. However, it is challenging to demonstrate that these test methods produce results that can be directly related to projecting the behavior of glass waste forms over time periods of thousands of years. One method to build confidence in the applicability of the test methods is to study the natural processes that have been taking place over very long periods in environments similar to those of the repository. In this paper, we discuss whether accelerated testing methods alter the fundamental mechanisms of glass corrosion by comparing the alteration patterns that occur in naturally altered glasses with those that occur in accelerated laboratory environments. This comparison is done by (1) describing the alteration of glasses reacted in nature over long periods of time and in accelerated laboratory environments and (2) establishing the reaction kinetics of naturally altered glass and laboratory reacted glass waste forms