Geological long-term evolution: Erosion

pdf NAB 24-15 Geological long-term evolution: Erosion(54.04 MB)

This report supports the geosynthesis by presenting scenarios of future erosion relevant for site comparison and the safety case. Ensuring repository safety requires that barrier integrity is maintained for 100,000 years for low- and intermediate-level waste (L/ILW) and for one million years for high-level waste (HLW). These barrier properties are depth-dependent, sustained as long as a minimum residual cover remains above the Opalinus Clay. Over time, erosion driven by rock uplift and climate processes reduces this cover. This report addresses two main questions: (i) the expected remaining overburden thickness in each siting region after these timeframes, and (ii) the probability of erosional excavation after one million years.

The erosion assessment builds on understanding from Stage 2 of the Sectoral Plan for Deep Geological Repositories (SGT Stage 2), which considers past landscape evolution as a predictor of future trends. The landscape around the siting regions is shaped by progressive fluvial incision from the Aare and Rhine Rivers, development of low-relief cuesta topography, and deep glacial erosion driven by Alpine foreland glaciations. These processes are largely influenced by glacial/interglacial cycles and long-term rock uplift caused by Alpine tectonics, with generally low erosion rates. Past reorganisation of the Rhine and Danube drainage basins has caused significant episodic incision events; baselevel drops of similar magnitude are not expected in the next million years.

In SGT Stage 3, a comprehensive scientific field investigation programme was launched to analyse the Quaternary evolution of Northern Switzerland, covering the upper 300 meters of the subsurface with seismic data, drill cores, and new geochronology. Numerical models and literature reviews further developed understanding of erosion dynamics. One important outcome was that various anthropogenic CO₂ emissions may result in a delay of the next full glacial for several 100 kyr.

Erosion scenarios developed in SGT Stage 2 were refined to to better incorporate and illustrate uncertainties. This includes a new model cascade that allows a combination of numerical simulations and probabilistic calculation cases to estimate future repository overburden and excavation probabilities, thereby integrating expert judgment for quantifying the uncertainties of model parameters. The erosion model resembles a cascading system: Model A (fluvial incision) provides input to Model B (evolution of local topography), and both feed into Model C (deep glacial erosion). Sensitivity analyses were conducted to gauge the impact of parameter uncertain­ties.

Model A simulates fluvial incision using the stream power incision model (SPIM) based on rock uplift, baselevel drop, erodibility, and drainage scenarios. After one million years, the most likely scenarios indicate that the provisional disposal areas in all siting regions remain well below the fluvial erosion base. However, in some scenarios, the Aare – Rhine River system may incise down to the level of the repository in JO. Accordingly, the JO site depends on the preservation of the additional cover given by the local topography, which is not necessary for Nördlich Lägern (NL) or Zürich Nordost (ZNO). Model B incorporates a 3D geology and simulates local river and hillslope responses to the fluvial incision (output of Model A), accounting for stable drainage networks and alternative channel configurations, particularly in JO. Results indicate that, while JO benefits from additional topographic cover, alternative drainage configurations may reduce the over­burden, resulting in a disadvantageous reduction of remaining overburden thickness after one million years. Model C evaluates potential deep glacial erosion by combining fluvial incision (Model A) and drainage scenarios from Model B. Key inputs include depth distributions of glacial overdeepenings, bedrock erodibility, future glaciation scenarios, as well as the spatial occurrence and percentage of glacial overdeepenings within the siting regions.

The results show significant differences in remaining overburden thickness across the three siting regions at the locations of the provisional disposal areas. The most likely scenarios show at least 400 m of remaining overburden thickness above the repository after one million years at all the siting regions. In NL, nearly 600 m of overburden thickness is most likely. At the 5 – 95% probability range, between ~ 125 and 410 m of overburden thickness might be expected in JO, between ~ 350 and 670 m in NL and between ~ 220 and 560 m in ZNO. An excavation of the repository within one million years was estimated to be highly unlikely in all three siting regions. This estimated likelihood is about a magnitude lower for the provisional disposal area in NL than for JO and ZNO.

The erosion assessment shows that, with the geological evolution expected for the period under consideration, the overburden thickness will remain large enough to ensure robust self-sealing within the Opalinus Clay. Regarding less likely erosive events, the NL siting region is the most robust, particularly because of the greater depth of the repository. JO is the least robust, especially towards possible future local river diversions.