Evolution of Repository Conditions—Swedish Repository at Forsmark

The Forsmark site located in the municipality of Östhammar, Sweden, was chosen for the construction of a future KBS-3 repository. It is estimated that 1.2 × 104 tonnes of spent nuclear fuel will be transported to the final KBS-3 DGR, which would require approximately 6000 Cu containers to dispose of the high-level nuclear wastes [6].

Mark I Mark II KBS-3 1050 mm 4 8 3 5 m m 562 mm 2 5 1 2 m m

Figure 1.7: Comparison of the (a) KBS-3, (b) the original NWMO reference (NWMO Mark I) and (c) the current Canadian reference (NWMO Mark II) used fuel container designs [10].

1.1.3.1 Thermal Performance of the Cu Container

The thermal performance of a Cu container is determined by various factors, such as the thermal properties of the rock, container, tunnel spacing, and the degree of saturation of the highly compacted bentonite buffer surrounding the container [16,24]. The following discussion is focused on the thermal evolution of the container surface after repository closure.

The average background temperature of the host rock is approximately 11 °C at the proposed repository depth of around 500 m in Forsmark [10], which is considerably lower than the temperature on the surface of a Cu container. The residual heat from a container originates from the radioactive decay of the spent fuel waste form, leading to a transient increase of the container surface temperature to a peak of 85 °C after ~10 years of emplacement [10]. One pessimistic calculation indicates that the surface temperature of a container could reach a peak above 95 °C, provided that the bentonite buffer remained unsaturated [7,10]. The container surface temperature will decrease to 40 °C after 1000 years of containment, and to an average temperature of 20 °C after 10,000 years. Eventually, the container surface temperature will reach the host rock background temperature in approximately one million years due to slow heat dissipation into the buffer and host rock, Figure 1.8 [10]

Figure 1.8: The anticipated evolution at the waste container surface temperature in a KBS-3 repository [7]. Image courtesy of SKB.

1.1.3.2 Bentonite Buffer Saturation Time

The saturation of bentonite buffer is influenced by the wetting/drying from the rock and backfill and the heating induced by the waste container. Model calculations have been made by Åkesson et al. for various sets of conditions and assumptions [26]. In general, the saturation time of the buffer is anticipated to vary between a few tens to a few thousands of years depending on its position in the repository. More recent modelling by Sellin et al. [27] using an hydraulic conductivity of the rock matrix is 10−13 m/s, indicated that the saturation time would be more than a thousand years in most deposition holes.

1.1.3.3 Redox Conditions and Groundwater Sulphide Concentrations

After an initial short-lived oxic phase due to the disturbance caused by excavation and operation of the repository, reducing conditions will be re-established and then prevail [10]. Once reducing conditions have been established, sulphide will be the predominant corrosive agent that poses a long-term threat to the durability of the Cu containers. The concentration of sulphide in groundwaters is controlled by a balance between its production by microbial sulphate reduction and its consumption by oxidation and precipitation with metals. In particular, of the sulphide that will be produced, the vast majority (> 99 %) will be present as precipitated mackinawite (Fe1+xS) [28]. Based on the simulation by King and Kolář, the sources/production of dissolved Fe (II) in the repository included siderite (FeCO3), biotite, and pyrite (FeS2) present in the backfill, buffer materials and rock layers [28]. Whereas the consumption of Fe (II) is attributed to the irreversibly precipitation as Fe1+xS if the concentrations of dissolved Fe (II) and SH− exceed its solubility product [27]. In addition, calculations suggested that sulphide would also be consumed by mineral oxides in the geosphere [27].

The distribution of groundwater sulphide concentrations in the Forsmark site has been thoroughly researched and evaluated, Figure 1.9 [10], which shows the predicted sulphide concentration ranges from 1.2 × 10-7 mol/L to 1.2 × 10-4 mol/L. The majority of values are at least one order of magnitude lower than the maximum sulphide concentration.

1.1.3.4 Peak Sulphide Fluxes at the Container Surface

Sulphides will be produced in remote locations where the microbial sulphate reduction reactions are robust. Dissolved hydrogen (H2), methane (CH4), and organic carbon, are potential reductants that may be used by SRB to produce sulphide [7]. The total sulphide concentration in the vicinity of the container is expected to be substantially lower than the bulk sulphide concentration in the groundwater due to the presence of other barriers to transport, such as the bedrock and buffer materials.

The rate of sulphide-induced Cu corrosion will be determined by the mass transport of sulphide from the remote locations where it is produced to the container surface, leading to the establishment of a sulphide concentration gradient. According to Fick’s first law of diffusion, the diffusive flux is proportional to the concentration gradient [29]. In other words, the sulphide flux (mol/(cm2⸱s)) at the container surface becomes one of the vital factors that control Cu corrosion under anoxic conditions.

The phases of repository evolution are of relevance to the sulphide flux at the container surface [10]. When the bentonite buffer is unsaturated, sulphide may be transported via the gas phase. Once bentonite is fully saturated, the sulphide flux at the

Figure 1.9: Groundwater sulphide concentrations used for assessing the long- term safety of a KBS-3 repository. Image courtesy of SKB (technical report TR-19-15 [6]).

surface of a container is expected to be very low due to the large diffusive transport resistance in the bentonite buffer [7]. It is reported that the peak sulphide flux to the surface under saturated conditions will be less than 10-10 mol/(m2⸱s). This peak sulphide flux of ~ 10-11 mol/(m2⸱s) is calculated for the highest sulphide concentration in the groundwater (1.2 × 10-4 mol/L) [10].

Even under the worse case scenario when it is assumed the bentonite has been eroded in the vicinity of the container and the sulphide concentration in the groundwater (1.2 × 10-4 mol/L) is at its highest, a peak sulphide flux of only 10-9 mol/(m2⸱s) is obtained [10].