Sorption and matrix diffusion in crystalline rocks – In-situ and laboratory approach

The geological disposal of spent nuclear fuel may represent a good solution after a thorough risk assessment. In Finland and Sweden, the host rock for geological disposal is crystalline rock and the repository is based on the Swedish KBS-3 multi-barrier design. When radionuclides from spent nuclear fuel are potentially released into the bedrock, they are then transported by advection along water-conducting fractures. The retardation of radionuclides can occur via molecular diffusion into stagnant pore water and/or via chemical retardation onto mineral surfaces in the rock matrix. When assessing the risk of spent nuclear fuel deposition, it is important to understand the transport behaviour of radionuclides in the bedrock, that is, at the rock–groundwater interface. Most parameters, such as the diffusion coefficient, the distribution coefficient and the rock porosity, associated with radionuclide migration in the rock are obtained from laboratory scale experiments. In order to apply results from laboratory experiments to the full-scale nuclear fuel deposition, in-situ experiments are performed. For example, the Swiss National Cooperative for the Disposal of Radioactive Waste (Nagra) have been conducting extensive in-situ experiments at the Grimsel Test Site (GTS). During the first in-situ test, tritiated water (HTO), 22Na, 134Cs and 131I, as well as non-radioactive isotopes 127I and 23Na, were circulated in a borehole interval isolated by packers 70 cm apart from each other. The second ongoing Long Term Diffusion (LTD) experiment primarily uses the same radionuclides as well as the non-radioactive element selenium. This thesis presents the laboratory analysis of HTO and iodine from an in-situ diffusion experiment and supporting laboratory studies aimed at determining the sorption and diffusion of caesium and selenium on Grimsel granodiorite (GG). Caesium sorption was studied through batch sorption experiments using crushed rock, while selenium diffusion and sorption relied on batch and block scale experiments using Kuru grey granite (KGG) and GG rock blocks. HTO and iodine diffusion was modelled using the time domain diffusion (TDD) method among in in-situ rock blocks as well as selenium in a laboratory using rock blocks. The outleaching method proved successful for analysing non-sorbing radionuclides from the connected pore network of the GG. TDD modelling of the results lead to an apparent diffusion coefficient of 3 × 10-10 m2/s for both HTO and iodine. No significant difference between the in-situ and laboratory diffusion coefficient was detected. Caesium sorption stood at 0.107 ± 0.003 m3/kg on GG at a 10−8 M Cs concentration. Sorption was highest on biotite at Kd = 0.304 ± 0.005 m3/kg, explaining the in-situ diffusion results of caesium which followed the biotite veins in GG. The sorption of selenium was significantly overestimated when the determination was conducted on crushed rock using the batch sorption method compared to studies conducted on intact rock. The outleaching method proved successful in the analysis of non-sorbing radionuclides, while flexible TDD modelling proved to be a quite useful tool in handling measured data from both block scale experiments and in-situ experiments of weakly and non-sorbing tracers.