Nuclear Waste Glass Durability as a Function of Groundwater Composition

The complex nature of groundwater solutions makes it difficult to predict how groundwaters of different chemistry influence glass dissolution, for example, if a granite groundwater from one potential GDF site contains Mg, but another does not, how is it possible to make predictions of long-term glass behaviour based on a generic granite groundwater? To achieve a detailed mechanistic understanding, a step-wise approach using simplified solutions is necessary. Examining individual ionic species, and building a model that is capable of describing how each individual ion, and how different combinations of ionic species together, influence the durability of glass is advantageous since a more detailed analysis of long-term behaviour can be developed, underpinned by a thorough geochemical understanding of the dissolution process.

To provide information to build a geochemical model for the dissolution of the UK's nuclear waste glass (including the compositions MW, Ca/Zn, POCO, BUTEX and ISG) as a function of groundwater ion species, this project will utilise a multi-technique approach involving:

i) Dynamic flow-through experiments with glass powder (SPFT tests);
ii) In-situ surface analysis techniques (e.g. vertical scanning);
iii) Long-term static powder dissolution tests;
iv) Long-term static monolith dissolution tests on timescales of 1 to 3 years with periodic sampling and surface characterisation (e.g. VSI, EDX, TOF-SIMS, X-ray synchrotron techniques, TEM in collaboration with partners at ORNL);
v) Geochemical and thermodynamic modelling of all solution data acquired (e.g. PHREEQC, GEMS, application to the GRAAL model) to determine the alteration layer formation mechanism, to identify the geochemical controls on dissolution in different solutions and to predict the behaviour of UK HLW glass in a wide range of groundwaters. It is anticipated that this modelling work will be performed, in part, at the National Nuclear Laboratory during a placement.