Reducing Uncertainty in Performance Prediction

Final Report Summary - REDUPP (Reducing Uncertainty in Performance Prediction)

Executive Summary:
In a spent nuclear fuel repository, the engineered barriers are designed to protect humanity from the radiological risks associated with the waste for at least 100 000 years. In Sweden and Finland, the spent fuel will be protected by copper canisters, bentonite clay and 500 meters of crystalline bedrock. Water is not expected to enter the canisters for a very long time; however, when it does, the spent nuclear fuel will start to dissolve, resulting in a release of the radioactivity into the surrounding environment. Therefore, the rate with which the spent nuclear fuel dissolves is central to the safety assessment of the spent fuel repository. In the REDUPP project, this question is investigated through dissolution experiments using analogue materials. There is today an understanding and a general consensus regarding what main processes influence spent fuel dissolution. Spent fuel mainly consists of uranium dioxide, which is nearly insoluble in the absence of oxidants, such as oxygen. However, since it contains radionuclides, it is better the slower the dissolution is. Our present knowledge tells us it will take several millions of years to dissolve all the spent nuclear fuel in the repository, providing the surrounding repository environment keeps oxidants away from the spent fuel. This estimate is based mainly on laboratory experiments using real spent nuclear fuel, and analogue materials, such as uranium dioxide and similar oxides.

Unavoidably, laboratory experiments are of much shorter duration than the time analysed for the safety assessment of the repositories. In order to increase the confidence in the extrapolation of these dissolution rates far into the future, a deeper understanding of the dissolution process is required. An interesting and confounding observation, and recognized problem, is that when results from laboratory dissolution experiments are compared with studies of natural systems, the experimental studies indicate dissolution rates that are orders of magnitude higher. This problem is addressed here by investigating how the surface of a dissolving solid changes during the course of dissolution. It is at the sold-fluid interface that dissolution takes place, and by using the experimental equipment available today, very fine details of the changing surface can be imaged and analysed. Coupling the analysis of the bulk solution with the studies of the surface, it is possible to identify which sites at the surface are most reactive during the dissolution process. Another issue that is being investigated is the effect of using real, natural groundwater instead of the simulated groundwater compositions that were used previously in the experiments.

The dissolution experiments show that the dissolution rates rates slow down as dissolution proceeds. Dissolution of grain boundaries plays an important part in the first few days of dissolution, resulting in a relatively fast initial dissolution. A strong crystallographic control was observed, leading to faster dissolution for grain boundaries with high misorientation angle. Modelling using Ab Initio Molecular Dynamics, combined with atomistic thermodynamics, predicts dissociative chemisorption of water on the UO2 surface, leading to a hydroxylated surface. Thus, the first step in the dissolution process is described from first-principles. Overall, the results show the effects of surface properties, especially grain boundaries and surface defects, on measured dissolution rates. Changes in the chemical composition of the fluid will affect the dissolution rate, as will the crystallographic structure of the exposed surfaces. The surface properties are expected to change over time, as could the fluid composition. Thus, these parameters need to be carefully considered when conducting laboratory experiments and when applying these dissolution rates to analyses that cover many thousands of years.
Project Context and Objectives:
Please see the attached document, which contains the full Final publishable summary report.
Project Results:
Please see the attached document, which contains the full Final publishable summary report.
Potential Impact:
Please see the attached document, which contains the full Final publishable summary report.
List of Websites:
Website: www.skb.se/REDUPP

Contact details:
Dr Lena Z. Evins (REDUPP Coordinator)
lena.z.evins@skb.se
+46-(0)8-579 386 88

Final Report Summary - REDUPP (Reducing Uncertainty in Performance Prediction)

Related documents

Other documents

Share this page

Download