Arbeitsbericht NAB 22-47
Reference porewaters for SGT Stage 3 of the Opalinus Clay for the siting regions Jura Ost (J), Nördlich Lägern (NL) and Zürich Nordost (ZNO)
Abstract
The scope of this report is to define a reference porewater composition for the host rock Opalinus Clay in the three siting regions Jura Ost (JO), Nördlich Lägern (NL) and ZNO (Zürich Nordost) as a constraint for deriving bentonite porewater compositions, sorption coefficients and solubility limits for radionuclides used in performance assessment calculations. This work is performed in the context of the Sectoral Plan, Stage 3 (SGT Etappe 3), that provides an in-depth evaluation of the three regions remaining from the pre-selection process of Stage 2.
A single composition for a reference porewater for the Opalinus Clay is defined for the siting region Zürich Nordost (ZNO) and Nördlich-Lägern (NL), abbreviated ZNO-NL. A reference porewater of lower salinity is defined for Opalinus Clay of the region Jura Ost (JO). These reference porewaters are based on data derived from the recent deep drilling boreholes Trüllikon-1-1 (TRU1-1, 2019/2020), Marthalen-1-1 (MAR1-1, 2020), Bülach-1-1 (BUL1-1, 2019), Stadel-3-1 (STA3-1, 2021), Stadel-2-1 (STA2-1, 2021), Bachs-1-1 (BAC1-1, 2022), Bözberg-1-1 (BOZ1-1, 2020) and Bözberg-2-1 (BOZ2-1, 2021) as well as from the geoscientific boreholes drilled in Benken (1998/1999) and Schlattingen (Schlattingen-1, 2011), and some few older boreholes (only partial information), all penetrating the Mesozoic strata and ending in Permo-Carboniferous sediments or basement rocks. The regions are therefore well constrained in terms of expected salinities and salinity gradients.
The borehole geochemical depth profiles for ZNO-NL share a similar stratigraphy of the clay-rich confining units, a salinity maximum at the top of Opalinus Clay and/or at the base of the overlaying clayey strata, and a small decrease towards the base of Opalinus Clay (in boreholes BUL1-1, STA3-1 and BAC1-1 no or only minor impact of a Keuper aquifer). Common is also the presence of a bounding aquifer in the Malm, but an aquifer of low salinity in the Keuper is only consistently present in the ZNO region. A Keuper aquifer is lacking at BUL1-1 and STA3‑1 and is bearing similar or only moderately reduced salinities at STA2-1 and BAC1-1 in the NL region, where steeply decreasing chloride concentrations in the Staffelegg Formation are lacking. Chloride concentrations in the Opalinus Clay for ZNO are similar at the TRU1-1 and MAR1-1 locations, and somewhat lower at the Benken location, with an overall range of approximately 6'500 – 10'000°mg/L for the most saline parts. Chloride concentrations for NL are highest at BUL1-1 (11'500 – 14'000°mg/L, depending on the method), but cover a range of 6'000 – 10'700 mg/L for STA3-1, STA2-1 and BAC1-1. There is a distinct and systematic difference between chlorinities obtained by porewater squeezing and advective displacement, the latter being more saline by approximately 10 – 25%. These two independent methods sample directly porewater aliquots from core samples and form one of the pillars for defining the reference porewater compositions.
The borehole geochemical depth profiles for JO feature distinctly less saline porewaters compared to ZNO-NL. BOZ1-1 located in the southern part of JO has chloride concentrations near 3'000°mg/L in the Opalinus Clay and upwards (no aquifer sampled), and it is decreasing to 1'000°mg/L at the position of the Keuper aquifer (visible as a modest dip in water stable isotope composition, but too low hydraulic conductivity for sampling). The profile at BOZ2-1 in the northern part is bound by distinct aquifers in the Hauptrogenstein and the Keuper, with maximum chloride concentrations approaching 2'000°mg/L at the top of the Opalinus Clay and decreasing towards the bounding aquifers. The systematic difference between chlorinities from squeezing and advective displacement (the latter being more saline) is present but distinctly smaller compared to data form the ZNO-NL regions.
The geochemical model for the reference porewaters is very similar to a previous model (SGT Etappe 2) with fixed chloride and sulphate concentrations, a prescribed partial pressure of CO2, a measured exchanger composition (as initial estimate) and multiple mineral saturation constraints. An average chloride concentration of 8'500°mg/L is adopted for the reference porewater for the ZNO-NL regions, and a sulphate concentration of 2'250°mg/L constrained by advective displacement experiments and squeezing experiments. A partial pressure of CO2 of 10−2.2°bar is imposed as expert judgement on multiple lines of evidence, and the obtained model pH is 7.1. The reference porewater for JO is derived in an analogous fashion but it is distinctly less saline and more sulphatic relative to chloride, with an imposed chloride concentration of 3'000°mg/L, and a sulphate concentration of 2'600°mg/L. A partial pressure of CO2 of 10−2.2°bar is imposed as expert judgment, and the obtained model pH is 7.3.
Redox condition and some minor and trace components (F, Si, Al, Fe, Ba, Mn) are constrained by mineral saturation of stoichiometric end-member phases (fluorite, quartz, kaolinite, pyrite, siderite, barite, rhodochrosite). While calcite, dolomite, pyrite, quartz, kaolinite (and possibly fluorite) are ubiquitously present and reasonable choices as equilibrium phases, siderite, barite, and rhodochrosite are proxies, and are observed as solid-solutions rather than end-member phases. The concentrations of Fe, Ba and Mn may therefore be overestimated in this model and are likely just an upper limit for concentrations. Celestite is used in the model as a solubility constraint for Sr concentration, but it is not inferred to be necessarily present as a controlling phase. It is not ubiquitously present in Opalinus Clay (confirmed as traces in some samples, searched for but not found in many others), but it is at/very near saturation or slight supersaturated in squeezing and advective displacement aliquots.
The calculated redox potentials (ZNO-NL: −2.76°pε [−163°mV EhSHE] at pH = 7.07; JO: −3.06°pε [−181°mV EhSHE] at pH = 7.34) are based on the given sulphate activity and mutual pyrite – siderite equilibrium. Because of the siderite equilibrium, this is also tied to the carbonate system (calcite – dolomite – alkalinity – PCO2 – pH). The redox potential is therefore shifting with pH and is not an independent quantity.
Uncertainties related to the partial pressure of CO2, mainly based on previous work, are such that a variation of a log-unit from 10−1.8°bar to 10−2.8°bar result in a pH range from 6.9 to 7.4 (for ZNO-NL), and correlated shifts in alkalinity, with other components undergoing little change. This uncertainty, including the correlated parameters, is addressed for ZNO-NL and JO with two porewater variants calculated with a PCO2 of −1.8 and −2.8 log-bar units. Apart from well-constrained chloride concentrations, it is not possible to provide exact bounds on uncertainties of other components because the uncertainty range of the underpinning measurements is not accurately known (complex laboratory procedures and known/suspected experimental artefacts). A substantial uncertainty related to sulphate concentration arises from seemingly contradictory constraints (measurements of aqueous extracts vs. advective displacement and squeezing), and this is addressed with a more sulphate-rich porewater variant (with approximately twice the sulphate concentration), and accordingly lower concentrations of Ca, Mg and Sr in order not to be supersaturated with respect to gypsum/anhydrite and/or celestite.
The reference porewaters were requested for a temperature of 25°°C for subsequent radionuclide solubility and speciation calculations. The use of the latest version of the PSI Chemical Thermodynamic Database 2020 was a pre-requisite, but comparative calculations to commonly used older versions and the PHREEQC database are also performed and show only small differences for most components. The potential effects of temperature on the porewater composition are addressed in a separate section. Similarly, an alternative modelling approach using a combination of selected hydrous silicates as equilibrium phases in lieu of a prescribed partial pressure of CO2 is also presented.