Glass chemistry development strategy for Hanford high-level waste

Developing glass compositions for Hanford high-level waste (HLW) is complicated by several factors. Hanford HLWs are stored in 177 underground storage tanks and have a wide range of compositions, some of which are not well-known. Further, decisions about retrieving, blending, and pretreating Hanford tank wastes are not final. A global-local strategy for developing and qualifying waste glass compositions in this uncertain Hanford environment has been formulated by Pacific Northwest National Laboratory (PNNL) over the last 10 years. The global phase of the strategy explores a wide region of glass compositions to cover the widely varying and uncertain compositions of Hanford wastes. Local phases of the strategy focus on narrower glass composition regions corresponding to well-known individual waste types. The global and local phases of the strategy rely on statistically designed experimental studies, which are used to develop property-composition models as functions of glass composition (and temperature for some properties). Property limits implemented via the property-composition models (and their corresponding uncertainty equations) are then used to identify one or more qualified glass composition regions that meet product quality and processing constraints. Optimum glasses (with maximum waste loadings to minimize glass production and disposal costs) within the qualified region(s) are then developed using the property-composition models. The steps of the strategy are discussed, and applications of the strategy at PNNL are briefly summarized.