This thesis focuses on the chemical and isotopic features of groundwater residing in the upper 1.2 km of sparsely fractured crystalline bedrock of the Baltic Shield. The work is based on previous understanding of the groundwater origin and its evolution in the fractured bedrock of the Äspö Hard Rock Laboratory (underground tunnel and facilities) and in two candidate coastal areas (Laxemar and Forsmark) investigated by the Swedish Nuclear Fuel and Waste Management Company (SKB) for future construction of a nuclear waste repository. In order to assess the origin and the apparent mobility of major elements and key trace elements (Cs and YREEs) in this difficult-to-reach deep environment, new (and previously published) data of SKBs hydrogeochemical monitoring programme was iteratively characterised and integrated in phenomenological models. The overall aim was to improve the integration between groundwater mixing and in situ water-rock interaction processes in deep coastal crystalline bedrock under natural and/or disturbed (i.e., dynamic) flow conditions induced by the presence of a tunnel system.

The multiple origins (glacial, marine, meteoric and allochthonous) of the fracture groundwater resulted in a large range of concentrations for dissolved major and trace elements in the studied bedrock. Dependent on the current flow conditions, the apparent mobility of dissolved elements was generally challenging to identify in the deep fractured bedrock under natural flow conditions. There, the relatively long residence time of most of the various groundwater types prevented to clearly differentiate the (apparent) fast retention processes from slow but active processes on a long-term perspective. Both processes alter the primary hydrochemical composition mainly imposed by the mixing between the dominant sources of groundwater. Nevertheless, in the particular case of YREEs, their generally low natural concentrations and predominant binding to organic colloids in most palaeo- (and modern meteoric) groundwaters (independently of the flow conditions) indicated strong active sorption onto minerals and physical filtration of organic colloids in the fractures. Together, these properties tend to minimise the mobility of dissolved YREEs and to stabilise their concentrations and fractionation patterns during the long residence time of the groundwaters.

At the Äspö HRL, an analogue (in a broad sense) of future repositories for high-level and long-lived radioactive wastes, changes in groundwater origin and salinity took place rapidly in subvertical fracture zones and progressively within the sparsely fractured deep rock domains. The changes resulted either from partial-to-substantial replacement of palaeo-groundwater by modern surface/shallow ground-water or induced dynamic up-flow of deep-lying saline groundwater. The hydrogeochemical instability near the underground facility during excavation to operational phase helped to assess qualitatively – and in some case differentiate quantitatively – the combined role of mixing, short-term and long-term reactions on the chemical composition of groundwater and the mobility of major elements and Cs within fracture zones and the sparsely fractured rock domains.

Collectively, the findings of the individual studies showed that the composition of intruded past or modern marine groundwater was likely to affect the natural retention properties/reactivity of the bedrock towards dissolved species at repository depth. For instance, the intrusion of modern seawater induced a desorption process of some dissolved species originally present on the exchange sites of the clayish fault gouge material in the fractures. This contributed to an apparent increase of the abundance level of dissolved cations naturally occurring in relatively moderate (i.e., K and Mg) and trace (i.e., Cs) concentrations in the fracture groundwaters.

The general understanding of the current hydrogeochemical conditions in deep crystalline bedrock is crucial when predicting future changes in groundwater chemistry (i.e., climatic cycles), which in turn might be of relevance to the long-term integrity of the KBS-3 repository method developed for isolating the nuclear waste from the surficial environment and biosphere.