In the following, the development of the database and the process understanding gained since previous studies for the demonstration of disposal feasibility for high-level waste (Nagra 2002) and the more recent documentation of Stage 2 of the siting process (Nagra 2014h) are discussed with regard to specific topics relevant for post-closure safety.

The following list describes the evolution of the database and the system understanding compared to the reference studies mentioned above. Aspects that were confirmed by the new data and studies are indicated with a green checkmark (ü). Significant changes (geological situation, con­cepts and parameters) compared to the earlier investigation phases are indicated with a red excla­mation mark (!).

• ü The Opalinus Clay host rock: Compared to the reference studies mentioned above, the database has been substantially expanded (Fig. 2‑3). Key barrier properties and processes have been sub­stantiated. The site-specific datasets also confirmed that only minor differences exist between the siting regions, as expected in SGT Stage 2.

  • ü Drilled host rock thicknesses are in the expected range (~ 100 – 120 m). ► Section 4.2.6

  • ü The lithological variability is small compared to other Mesozoic formations of Northern Switzerland, and the general sequence of subfacies trends can be correlated over all boreholes. A new and consistent subfacies scheme was developed, linking the observed subtle variations and trends to the depositional environment and its temporal evolution. ► Section 4.2.6

  • ü The vastly extended database confirms the mineralogical composition with con­sistently high clay-mineral content and a small lateral variability in all the siting regions. ► Section 5.2.3

  • ü Porosity shows only a slight depth dependency and only minor differences between the different siting regions. ► Section 5.3.2

  • ü The extended database allows for an improved understanding of the relationship between lithology, mineralogical composition, porosity and pore-space architecture. ► Sections 5.3.3, 5.3.4

  • ü The degree of tectonic deformation in the Opalinus Clay is low, with the majority of recovered drill cores free of any natural discontinuity. ► Sections 4.3.4 and 5.5.4

  • ! The geomechanical properties of the Opalinus Clay were constrained based on a large number of triaxial tests. Compared to previous campaigns, new testing protocols have been established and their robustness was demonstrated before the deep borehole (TBO) campaign. The intact strength of the Opalinus Clay was found to be significantly higher than estimated in SGT Stage 2, and post-peak strength (rele­vant for faulted Opalinus Clay) is approximately comparable with previous estimates. No relevant differences were observed with current depth or between siting regions. ► Section 5.5.3

  • ! Many new microhydraulic fracturing tests and additional sleeve reopening tests have significantly improved the database on stress magnitudes in the Opalinus Clay. The results demonstrate a com­paratively stable gradient between the minimum (S_hmin) and maximum (S_Hmax) horizontal stress with depth, irrespective of the siting region. Whereas in earlier studies, the dominant stress regime was assumed to be strike-slip faulting across all three siting regions (S_Hmax > S_v and S_hmin ≈ S_v^, where S_v is the vertical stress), the dominant stress regime in the Opalinus Clay is now found to be normal faulting in NL (S_v > S_Hmax > S_hmin), and normal faulting to strike-slip faulting in ZNO and JO. The average of all three principal stress magnitudes (mean stress) is lower than estimated in SGT Stage 2. ► Section 4.4.4

  • ü The high self-sealing potential of the Opalinus Clay was confirmed by laboratory and field tests and can be related to its consistently high clay-mineral content in combination with a minimum effective stress greater than 2 to 3 MPa. Special hydrotests in the RHE1 borehole have reaffirmed the conceptual and quantitative understanding of the role of effective stress in self-sealing and directly demonstrated a rapid reduction of trans­missivity when injection was stopped. The test results also provide additional confidence that experience with self-sealing from the Mont Terri rock laboratory can be transferred to the siting regions. ► Sections 5.7.2, 5.7.4

  • ü Numerous laboratory and hydraulic packer tests confirm the very low hydraulic conductivities of both undisturbed and fractured Opalinus Clay. ► Sections 5.6.2, 5.6.3

  • ü Long-term monitoring systems installed in new boreholes show subhydrostatic hydraulic heads, i.e. fluid underpressures, in the Opalinus Clay, compared to the values in the over- and underlying aquifers. This observation is consistent with the long-term trend in the Benken borehole and confirms the very low hydraulic conductivities. ► Section 5.6.5

  • ü A substantial number of laboratory measurements extended the database on diffusion properties and allowed an improved understanding of the influence of the clay-mineral content and ionic strength (salinity) of the porewater on diffusive transport. ► Sections 5.4.2, 5.8.2

  • ! Porewater investigations based on dense and complementary datasets, including a substantial number of analyses based on direct sampling techniques (squeezing, advective displacement), improved the porewater model and demonstrated site-specific differences. These feed into the parametrisation for radionuclide transport (diffusion and sorption). ► Section 5.4

• ! Confining units: An important aim of the investigation campaign addressed the characteri­sation of the confining units across the siting regions. Key issues were resolved, and new insights were gained with the SGT Stage 3 campaign:

  • ! Lower confining units – Staffelegg Formation: In SGT Stage 2, the Beggingen Member was treated as a potential release path for radionuclides. In contrast, the intensive hydro­geological investigation programme showed only slightly elevated hydraulic con­ductivities, including in fractured intervals. ► Section 4.5.3.9

  • ü Lower confining units – upper Klettgau Formation: The investigations confirmed that the Gruhalde Member in the upper part of the Klettgau Formation is characterised by low hydraulic conductivities in all the siting regions. The central and lower parts of the Klettgau Formation are more variable (see next point). ► Sections 4.5.3.9, 4.5.3.10

  • ! Lower confining units – downward delimitation: The lower confining units are bounded by the Keuper aquifer in all the siting regions. This local aquifer is internally complex and comprises three units, which may act as potential flow paths. These are (from top to bottom): Seebi Member, Gansingen Member and Ergolz Member. Facies, thickness and hydrogeological relevance of these units vary across the three siting regions. Their site-specific significance is better captured by the new boreholes. Consequently, the down­ward delimitation of the lower confining units is site-specifically tied to the uppermost unit with hydrogeological significance. In ZNO, the investigations confirmed the existence of a laterally continuous active Keuper aquifer in the Seebi Member, as sus­pected in SGT Stage 2. In Nagra (2002), the Keuper Group below the Seebi Member was treated as potentially part of the lower confining units – this is not supported by the new data. In JO, the new data point towards a laterally continuous Keuper aquifer related to the Gansingen Member. In NL, the enhanced hydraulic conductivities are limited to locally occurring sandstone channels of the Ergolz Member. ► Sections 4.2.4, 4.5.3.10

  • ! Lower confining units – potential contribution of the units below the Keuper aquifer to the barrier function in NL: The new investigations indicated that potential flow paths in the Keuper aquifer in NL relate to locally occurring sandstone channels of the Ergolz Member. Where these sandstone channels are absent or isolated, the underlying rocks (clay-mineral- and anhydrite-rich Bänkerjoch Formation) may act as an additional trans­port barrier. ► Sections 4.2.4, 4.5.3.10, 4.5.3.11

  • ! Upper confining units – clay-mineral content of units directly overlying the Opalinus Clay: The new investigations show that the Dogger Group directly above the Opalinus Clay is more clay-mineral-rich and contains fewer "hard beds" below the «Herrenwis Unit» in NL compared to JO and ZNO. The corresponding thick stack of clay-mineral-rich deposits in the eastern part of NL is interpreted to be the result of sediment accu­mulation by bottom currents. ► Section 4.2.7

  • ! Upper confining units – hydraulic significance of "hard beds" in the Wedelsandstein and Passwang Formations: In SGT Stage 2, the "hard beds" were treated as potential water-conducting features (i.e. potential release paths for radionuclides). The focused investigation programme did not indicate any elevated hydraulic conductivity. ► Sections 4.5.3.2, 4.5.3.5, 4.6

  • ! Upper confining units – «Herrenwis Unit»: The anomalous seismic facies detected on 2D seismic lines in NL in SGT Stage 2 was identified as an isolated carbonate platform, which is now designated as the «Herrenwis Unit». This platform developed on top of the positive relief formed by the directly underlying argillaceous units, which is probably the result of bottom currents (see above). On top of this swell, the conditions for coral growth were more favourable in comparison to the surrounding areas. Hydrogeological observations in the «Herrenwis Unit», and the fact that it is embedded in clay-mineral-rich units, indicate that it does not provide a relevant flow path connected to the biosphere. ► Sections 4.2.7, 4.5.3.7, 4.6

• ! Independent evidence for slow transport from natural tracers in porewater and hydro­geochemistry of groundwater in deep aquifers: The database was substantially expanded. This relates not only to the number of investigated boreholes but also to the significantly enhanced vertical resolution of data and to important analytical progress (e.g. ⁸¹Kr dating of groundwater) and the improved and more integrated data analyses and interpretation:

  • ! Porewater tracers were investigated in all new boreholes and complement the hydro­geological dataset of the aquitard units. The profiles of the natural tracers indicate differences in the thickness of the zone dominated by diffusive transport. ► Section 4.6

  • ! In combination with the hydrogeochemistry of the deep groundwaters, including the ⁸¹Kr model ages (the first application of this method in Northern Switzerland), tracer pro­files contain crucial information about the mechanisms and rates of transport in the aquifer – aquitard system over geological timescales. The availability of numerous tracer profiles across the siting regions allows the identification of regional trends and differences in the palaeohydrogeological evolution. In a plot of δ¹⁸O vs. δ²H, all pore­waters of the Opalinus Clay plot on the right side of the Global Meteoric Water Line. The shift away from the Global Meteoric Water Line is largest at NL, indicating a particularly large component of old porewater in this siting region and a lower influence by meteoric components. The differences in the overall shapes of the tracer profiles between the siting regions are mainly related to the presence/absence of the Keuper aquifer, the hydrogeo­chemical composition of the groundwater in the upper bounding aquifers (Hauptrogen­stein, Malm) and the duration of the diffusive interaction, which differs for each aquifer between the siting regions. Overall, this indicates that the most stable conditions of the aquitard – aquifer system are found in NL compared to ZNO and, even more so, compared to JO. ► Sections 4.5.6, 4.6

  • ! Groundwater from the Malm aquifer in NL is characterised by a uniform composition of stable water isotopes, showing a substantial shift to the right of the Global Meteoric Water Line. The shift is more pronounced than was previously known from boreholes in Northern Switzerland. The large dataset including ⁸¹Kr model ages now makes it possible to robustly substantiate earlier hypotheses regarding a marine component in the Malm groundwater and to obtain quantitative constraints on groundwater residence times. These waters characterise a quasi-stagnant flow field. In comparison, the data from the northwestern part of ZNO evidence a more active flow system with markedly younger ⁸¹Kr model ages. ► Sections 4.5.5.2, 4.5.6

  • ü The hydraulic activity of the Hauptrogenstein and its classification as a deep aquifer was confirmed by the investigations in the BOZ2 borehole. The base of the Hauptrogen­stein is thus regarded as the upper bound of the low-permeability confining units in JO. As expected from the Riniken borehole, the thickness of the Hauptrogenstein decreases eastwards. ► Sections 4.2.7, 4.5.3.6, 4.5.6

  • ! Groundwater samples from the Keuper aquifer are characterised by contrasting chemical types and a large range of residence times. This reflects the variable hydraulic properties (a lithologically complex aquifer), which were expected. As a result of the new litho­logical and hydraulic data, the differences between the siting regions could be better captured (see point above discussing the lower confining units). Groundwater geo­chemistry and ⁸¹Kr model ages provided further quantitative information on the activity of this flow system. The hydrogeochemical dataset supported by lithological arguments points to a quasi-stagnant flow system in NL. ► Sections 4.5.5.4, 4.5.6

  • ü As expected from previous studies, the most active flow system was observed in the Muschelkalk aquifer. The cold-climate isotopic signature known from earlier boreholes was confirmed for the ZNO and NL siting regions. The trends in hydrogeochemical com­position and residence times agree with the flow paths from the hydrogeological model. ► Sections 4.5.4.5, 4.5.5.5, 4.5.6

• ! Tectonic setting and fault pattern: Compared to previous phases, the structural geological knowledge base was substantially expanded, including 3D seismic reflection surveys in all the siting regions and investigations in numerous boreholes. Generally, the previously established tectonic concepts were confirmed. However, differences among the siting regions could be established:

  • ! Coherent, comparable and consistent seismic processing and interpretation allowed a detailed and robust comparison of the fault inventory in the three siting regions at the seismic scale. Differences between the siting regions were demonstrated and integrated into the larger-scale tectonic framework. In JO, an extended zone with increased faulting occurs in the Top Muschelkalk Group in the southeastern corner (the "Brugg Struktur­zone"). In NL, deformation in the Mesozoic strata is focused above structures related to the Konstanz – Frick Trough, leaving a large area free of seismically mappable faults in-between. In ZNO, the nearby transtensional Hegau – Bodensee Graben predominantly controls the structural inventory, resulting in a higher number of seismically mappable faults compared to JO and NL. ► Section 4.3.4

  • ü Extensive age dating of calcite cements associated with regional fault zones supports a main orogenic phase of the eastern Jura Fold-and-Thrust Belt between 14 and ~ 8 Ma. This new quantitative dataset acquired in Northern Switzerland is consistent with earlier understanding about the timing of folding and thrusting in the Jura. ► Section 4.3.5

  • ü Seismic interpretation confirmed the influence of inherited faults (e.g. faults related to the Konstanz – Frick Trough) on subsequent strain localisation. This indicates that future deformation is likely to preferentially occur along inherited faults (i.e. zones of weakness) rather than as new faults. ► Sections 4.3.4, 4.3.6

  • ü Stress indicators in the new boreholes confirmed the dominant present-day orientation of the maximum horizontal stress from SGT Stage 2, i.e. a NNW-SSE direction, approxi­mately perpendicular to the Jura Fold-and-Thrust Belt. ► Section 4.4.2

  • ! The degree of fracturing in the Opalinus Clay is lower compared to over- and underlying formations in all the siting regions. All recent boreholes show no evidence for any secondary décollement within the Opalinus Clay. ► Sections 4.3.4, 4.3.6, 4.3.7, 5.5.4

• ! Long-term geological evolution: Compared to earlier phases of investigations, the database and process understanding have been substantially expanded. Furthermore, the uncertainties related to future erosion have been addressed using a hybrid-probabilistic approach.

  • ü The acquired hydraulic data confirm previous understanding that a residual overburden thickness of 200 m can be considered reliable for efficient self-sealing in the Opalinus Clay (see above). Moreover, dedicated investigations in the shallow Lausen borehole demonstrated that a significant increase in hydraulic conductivity of exhumed Opalinus Clay is limited to the uppermost 20 to 30 m and showed that the increase is associated with exhumation-related fracturing and a marked increase in porosity. ► Sections 5.6.3, 5.7.2, 5.7.4, 6.5.2 and 6.5.3

  • ! The timing of past fluvial incision during the Early Pleistocene was refined with an extensive dating campaign using multiple geochronometers. The new data indicate a possibly younger age of the Höhere Deckenschotter and thus later fluvial incision after Deckenschotter deposition than previously assumed. ► Section 6.4.1.2

  • ! The timing of glacial overdeepening and the filling history was also substantially refined by dating the sediments from the drill cores. Several new ages reveal that part of the overdeepening formation is older than previously suggested, i.e. older than Marine Isotope Stage 6. ► Section 6.4.1.4

  • ü The depth of glacial overdeepening in Northern Switzerland was substantiated by 2D seismic data and drilling. The maximum depth remained in the expected range. The new data from the lower Aare Valley confirm the previous assumption that bedrock type exerts a strong influence on incision depth. ► Section 6.4.1.4

  • ! The uncertainties related to future erosion have been addressed using a hybrid-probabi­listic approach allowing for a quantitative estimate of future erosion, including uncertain­ties. In the previous assessment deterministic cases (plausible, optimistic, pessimistic) were derived based on the evaluation of the Quaternary past. The new concept covers a broader range of evolutions and evaluates the remaining repository overburden thickness over the periods under consideration in a probabilistic way. It shows that repository excavation during the period under consideration is extremely unlikely. ► Sections 6.4.3, 6.4.4

  • ü The densified seismic monitoring network seamlessly records all seismic events with a magnitude > 1 in Northern Switzerland and allows improved location of hypocentres, and in turn an improved understanding of regional seismotectonic activity. For example, it is possible to associate a sequence of micro-earthquakes occurring between 2014 and 2019 to the Neuhausen Fault, previously known from outcrop mapping and reflection seismic data. This and other micro-earthquake sequences further to the east confirm ongoing tectonic activity in the Hegau – Bodensee Graben system. ► Section 6.2.3

  • ü The extended observation period of global navigation satellite system permanent stations provides an improved quantification of recent deformation and confirms that present-day horizontal motion in Northern Switzerland is small. ► Section 6.2.3