SR-Site: Oxygen ingress in the rock at Forsmark during a glacial cycle

The aim of this report is to assess the possibility for oxygen to be transported by glacial melt-water to canister positions in a final repository for spent nuclear fuel at the proposed location in Forsmark. The approach for this assessment is to combine reactive transport modelling with geological observations of present and historical indications of oxygen ingress. For safety assessment purposes a cautious approach in the modelling is required when estimating the extent of oxygen ingress. In this report, a cautious approach has been applied both in the conceptualisation of the problem and in the choice of input parameters used in the models. Oxygen consuming processes are only neglected in the modelling if they are expected to further decrease the extent of oxygen ingress. Several oxygen consuming processes have been identified, each of which may play an important role in the scavenging of oxygen along recharge flow paths in the rock. These processes include biological pathways with degradation of organic material of ground surface origin, and biotically mediated reactions with reduced rock minerals and with various materials expected to be present in the backfilled repository volume. In the absence of microbes most of these reactions may also follow abiotic pathways. Present day observations show that degradation of organic material is the most powerful oxygen scavenging process. At Forsmark, oxygen is generally depleted within a few metres under present day temperate conditions. Although biological activity is likely to exist also during different phases of a glaciation, large uncertainties exist regarding e.g. the population growth dynamics, the biotic reaction rates and the availability of organic material under the highly varying conditions expected. Microbial activity and degradation of organic material is therefore pessimistically neglected in the calculations in this report. In the absence of organic material, ferrous iron present in minerals in the rock matrix is the main source of reducing capacity. The ferrous iron occurs in different reduced minerals, of which biotite and chlorite are the most abundant. The release rate of ferrous iron from biotite is slower than for most of the ferrous minerals considered. Therefore, the ferrous iron is cautiously assumed to be comprised in biotite, as a model substance, and the release rate is calculated based on this assumption. Furthermore, the oxidation of ferrous iron by oxygen is assumed to occur in two consecutive steps; dissolution of ferrous iron from the mineral lattice into the pore solution followed by homogeneous oxidation of the dissolved ferrous ions. Direct oxidation of ferrous iron incorporated in the mineral lattice is hence also cautiously approximated by these two coupled processes. At early times, reduced minerals in the undisturbed rock matrix, which are directly exposed to the flowing water, are easily accessible from the flow paths in the fractures. In this situation, the extent of oxygen ingress along the flow path is determined by the relative rates of oxygen recharge and oxygen consuming reactions. Eventually, the reducing capacity of the rock close to the fracture becomes depleted and the remaining ferrous minerals can only be reached by diffusion in the rock matrix. With time, this diffusion resistance increases as the reducing capacity is depleted further into the rock matrix. The oxygen consuming reaction is then limited by the diffusion resistance. The extent of oxygen ingress is in this situation determined by the relative rates of oxygen recharge and diffusion into the rock matrix. Both of these situations are represented by two different models that are solved analytically in this report. The case of kinetically controlled oxygen consumption for early times is furthermore evaluated with the geochemical numerical codes PHREEQC and PHAST. Sensitivity analyses of important parameters such as the pH, availability of specific reactive mineral surfaces, oxygen content in the glacial melt water, and rock matrix transport resistance are presented for both the analytical and the numerical models. In the numerical calculations, the effect of O2 consuming reactions with reducing fracture infillings is also evaluated. The results from the oxygen ingress models in this report are presented, or can easily be expressed, as the oxygen concentration as a function of the flow-related transport resistance (F-factor) of the recharge flow paths. This allows for the coupling of the oxygen ingress models with available results from hydrogeological models that include calculated F-factors along recharge flow paths under different static conditions. In this report, recharge flow path F-factors are used for stationary 'snap-shots' of unfavourable glacial situations at Forsmark, where the ice front is assumed to halt over the repository footprint.