The Geochemical Evolution of the HLW Repository Near-Field

pdf NTB 23-02 Rev. 1 The Geochemical Evolution of the HLW Repository Near-Field(11.90 MB)

This report addresses the chemical processes occurring during the spatial-temporal evolution of the HLW repository near-field in the Opalinus Clay formation. It represents an update of a previous report (Bradbury et al. 2014) and is part of the documentation for the general licence application.

The HLW near-field is understood to be the multicomponent system of engineered barriers in the repository drifts after sealing and closure, as well as the adjacent host rock. According to the provisional design, the barriers consist of a HLW disposal canister (for spent fuel or reprocessed high-level waste), made of a carbon-steel, compacted bentonite as buffer, the drift support structure made of concrete materials, and the surrounding Opalinus Clay host rock. In the presence of water or water vapour, the various near-field components will chemically interact with one another, potentially modifying the material properties and their functionality. It is therefore required to identify and evaluate the effects of such chemical near-field interactions during the entire period of assessment, which is one million years according to regulatory requirements.

The report is structured as a series of chapters dedicated to specific topics. After a general introduction (Chapter 1), a preliminary part (Chapter 2 – Section 5.1) presents descriptions of the key properties, initial states and thermal evolution of the intact HLW near-field. The geometry and functions of the various components are described along with the expected geochemical and geomechanical evolution of the HLW repository system. This part also includes an overview of the initial states of the chemical composition of engineered barrier materials and the reacting porewater. This is followed by a discussion, based on model calculations of the predicted postclosure temperature evolution.

The following part (Sections 5.2 – 9.4) focuses on the chemical evolution and mutual interactions between the components in the repository near-field (mineralogical changes, evolution of the porewater chemistry, effects on radionuclide solubility/sorption properties and functionality of the engineered barriers, particularly the bentonite buffer). Sections 5.2 – 5.4 discuss the effects of temperature on the mineral stability, radionuclide retention and diffusivities in bentonite. Based on experimental data and reactive transport models, Chapter 6 describes the chemical interactions between the tunnel support structure, Opalinus Clay and bentonite buffer. In Chapter 7, bentonite porewater compositions and characteristics (pH, Eh, ionic strength) are modelled and discussed. Chapter 8 is devoted to the chemical interactions at the bentonite/canister interface under watersaturated conditions, with the focus on the impact of formation of Fe-silicates onto the buffer function. In Chapter 9, chemical conditions inside a failed canister are modelled, assuming fully saturated conditions. Moreover, the potential influence of boron released via dissolution of vitrified waste on the solubility of dose-relevant radionuclides is assessed. Finally, Chapter 10 provides a summary of the entire report and the main conclusions relevant to the performance assessment. These are also briefly summarised in the following paragraphs.

Compared to Stage 2 of the Sectoral Plan, the updated calculations of the thermal evolution in the HLW repository section predict higher transient temperatures for longer periods. Temperatures slightly exceeding 100 °C are predicted across the entire bentonite section over several hundred years. At 10,000 years after repository closure, temperatures in the bentonite are predicted to be in the range of ~ 55 °C, the temperatures in the undisturbed host-rock ~ 47 °C. In contrast, the scenario discussed in the previous report (Bradbury et al. 2014) assumed a shallower repository depth, and the near-field was also predicted to be close to thermal equilibrium with the undisturbed host rock. Despite the slightly higher temperatures compared to the previous study, it could be concluded, based on ample experimental evidence from laboratory tests and natural analogues, that mineralogical transformations in the bentonite due to illitisation or the supply of dissolved iron from the corroding canisters will remain limited. A consequent loss of the buffer functionality due to reduced swelling capacity and modification of radionuclide retention/ diffusion properties can be excluded.

Updated reactive transport calculations to simulate the mineral and porewater evolution at the interfaces of the cementitious drift support structure with the Opalinus Clay and bentonite indicate similar results to those presented in Bradbury et al. (2014). Specifically, alteration zones in the clay do not exceed 10 – 20 cm on either side, and the total amount of dissolved clay minerals is predicted to be small at the end of the simulation time (50,000 years). The predicted precipitation of secondary minerals will reduce porosity within both clay domains close to the interfaces. Moreover, the simulations predict that alkaline water (pH > 9) from the tunnel support moving towards bentonite and Opalinus Clay will propagate no more than 10 cm from the respective interfaces and will not move further after around 100,000 years. Eventually, near-neutral conditions (pH ~ 8) will be reached across the entire section (bentonite – tunnel support – Opalinus Clay).

The geochemical “ClaySor” model, which follows a “classical” thermodynamic approach, was applied to predict composition and characteristics of porewaters in equilibrium with compacted bentonite. This serves as a basis for determining solubility limits and sorption coefficients of considered dose-relevant radionuclides for the safety assessment. The model takes into account cation exchange reactions in the interlayer of montmorillonite, protonation/deprotonation of amphoteric surface sites and dissolution/precipitation of reactive minor minerals. Despite wide but still realistic variations of key input parameters, such as bentonite dry density, initial occupancies of amphoteric surface sites, and pCO2, the results show narrow ranges of dissolved element concentrations, Eh values (-227 to -135 mV) and pH values (6.8 to 7.5). The low sensitivity of the results is explained by robust internal chemical buffering mechanisms, specifically carbonate and Fe(II)-Fe(III) mineral equilibria in conjunction with cation exchange.

Thermodynamic equilibrium calculations were also carried out to predict the chemical conditions inside the carbon steel canister after breaching and intrusion of bentonite porewater. Two cases were calculated, representing the initial and advanced stages of canister degradation. The former case leads to a mildly alkaline, strongly reduced (pH = 8.4, Eh = -495 mV) hydrogen-rich water, while the calculation representative of advanced degradation conditions produces a nearly neutral, moderately reducing hydrogen-poor water (pH = 7.2, Eh = -222 mV).

Based on the above model, the effect of boron release due to dissolution of the vitrified waste on complexation/solubility of radionuclides, as well as on pH, was evaluated. The results indicate that the strongest borate complexes should be formed with cations of transition metals (Fe and Ni). Depending on the pH, cations of alkaline-earth metals (Mg, Ca, Sr, Ba, Ra) and of trivalent actinides (Am, Cm) could also form borate complexes to a limited extent. In contrast, monovalent (Na, Cs, K) and tetravalent to hexavalent cations (U, Np, Pu, Tc) should not form borate complexes in significant concentrations. Overall, the results indicate that the effect of borate complexation and local pH increase induced by the dissolution of borosilicate glass on the solubility limits of dose-relevant radionuclides will be negligible.