Technical Report NTB 93-09
Kristallin-IConclusions from the regional investigation programme for siting a HLW repository in the crystalline
The current Nagra strategy envisages disposal of high-level and long-lived, intermediate-level radioactive waste in a deep geological repository. The crystalline basement of Northern Switzerland was identified as a potential host rock for such a repository and a three-phase geological characterisation plan was defined and initiated in 1981. Kristallin-I involves a synthesis of the results from the recently completed Phase I (regional scale) field investigations. This synthesis is very much based on the repository concept presented in Project Gewähr 1985 (PG'85; Nagra 1985). PG'85 was judged by the Swiss authorities in 1988 to have demonstrated the basic feasibility of construction of such a repository and showed that the concept introduced for the disposal of vitrified high-level waste (HLW) could provide sufficient assurance of safety. The geological database available at that time was relatively limited, however, and the authorities set requirements for demonstrating more convincingly the availability of suitable siting possibilities in Switzerland (Standortnachweis). Work since 1988 has also been carried out in sedimentary formations, in parallel to the crystalline host rock project documented here.
The main aims of Kristallin-I are:
- Updating and supplementing the PG'85 studies, with particular emphasis on the expanded geological database now available and progress made in understanding mechanisms and processes relevant for performance assessment.
- Documenting completion of Phase I and integrating results in order to provide the basis for selection of potential areas for a siting demonstration project.
- Providing updated technical input for waste management planning and setting future research priorities.
This report provides a concise overview of the entire project. It is supported by two technical synthesis reports, one of which integrates the results of the geological studies (Thury et al. 1994) while the other documents the associated repository performance assessment (Nagra 1994). These synthesis reports are, in turn, supported by large numbers of more specialist publications, many of which have been released in the Nagra Technical Report (NTB) series.
Selection of the crystalline basement of Northern Switzerland as a potential host rock was based on criteria including good long-term predictability of behaviour, potentially suitable hydrogeological characteristics, low seismicity, absence of exploitable natural resources and favourable rock mechanical/engineering properties. Regional characterisation during Phase I involved extensive seismic surveys and drilling 7 deep (1300 – 2500 m) boreholes, which were analysed in detail. Geological mapping and hydrogeological/hydrochemical surveys were carried out in the research region and in the nearby Southern Black Forest where the crystalline basement outcrops.
Progress since PG'85 is particularly evident in the development of the geological understanding of the basement in the areas of interest and the representation of this understanding in a geo-database for performance assessment. In particular:
- A much improved tectonic model of the region has been developed allowing two potential siting sub-regions (Areas West and East) to be delineated and the understanding of the distribution of major faults within these areas to be formalised in a statistical model.
- A new suite of hydrogeological models has been developed based on a synthesis of more extensive hydraulic measurements. They also directly consider the structural model developed and are constrained by regional geochemical and isotopic measurements.
- Great effort has gone into detailed characterisation of all water-conducting features observed in the deep boreholes and the resulting information has been synthesised into model representations of three classes of such features. These descriptions are supported by new literature studies to define the important sorption and matrix diffusion properties of these water-conducting features.
- Hydrochemical and isotopic data have been synthesised on a regional basis to provide a reasonably consistent picture of groundwater evolution due to rock/water interaction. This not only shows general groundwater flow patterns in the basement but also forms the basis for the definition of reference waters for the two potential siting sub-regions.
- Analysis of possible long-term geological evolution has underpinned the two bounding scenarios presented in PG'85 – particularly with regard to the quantification of expected uplift, erosion and movement along various sizes of faults in the areas of interest over time periods of up to a million years in the future.
In addition, the geosynthesis report provides an integration of all the key information produced to date which, taken as a whole, enhances confidence in the concepts developed in the various sub-disciplines (tectonics, hydrogeology, geochemistry, etc.).
Based on the expected geological structure, a conceptual repository layout has been defined in which "panels" of HLW emplacement tunnels are set out in low-permeability basement between sub-vertical, regional faults. For example, for the reference waste inventory (resulting from a planned 40 year lifetime of the existing 3 GW(e) of installed nuclear power plants), 3 panels in blocks with areas of ~ 0.3 km2 would be sufficient. At a reference depth of about 1000 m below surface, emplacement tunnels would be overlain by at least 100 m of low-permeability host rock, several hundred metres of higher-permeability basement and hundreds of metres of sediments. The repository layout also includes silos for disposal of long-lived intermediate-level waste, but this has not been further analysed within Kristallin-I.
Alternative nuclear power production scenarios and the option of direct disposal of future arisings of spent fuel are considered in Swiss waste management strategic planning but these options are also not discussed in any detail in the Kristallin-I synthesis.
As presented in PG'85, vitrified HLW in a stainless steel fabrication container would be encapsulated in a massive steel canister, which is emplaced horizontally in tunnels backfilled with highly compacted bentonite. The methodology and databases used for assessing the performance of this disposal concept have been developed somewhat since PG'85 with:
- Adoption of a rigorous scenario development and analysis methodology which documents system understanding in a more complete and traceable manner and allows the numerous calculations performed to be structured in a more logical manner.
- Improvement in conceptual model development to better represent the systems studied -particularly for the structure of the water-conducting features in the host rock which provides a basis for modelling radionuclide transport through the host rock.
- Extension of the capabilities of the model chain used for calculations - in particular explicitly accounting for near-field radionuclide retardation in bentonite and allowing consideration of non-linear sorption and colloid transport in the far-field.
- Improved and better justified databases for radionuclide release and transfer calculations (e.g. solubilities, sorption coefficients, etc.).
- Rock mechanical and thermal calculations to support conceptual layout studies which show how relatively small blocks of low-permeability basement can be utilised.
- Implementation of procedures which improve quality assurance (e.g. fully traceable calculation output, formalised peer review).
- Extensive investment of effort in model verification and validation, often as part of international studies, using natural analogues and via specific experiments.
- Updated reviews to put performance assessment results in perspective by comparison with other repository analyses and evaluation of other potential environmental hazards.
An analysis incorporating moderately conservative model assumptions and data (designated the Reference Case) indicates that peak radionuclide releases, due predominantly to Cs-135, would be more than two orders of magnitude below the regulatory guideline and 3 – 4 orders of magnitude less than the natural background in Switzerland. Effectively no releases would be expected within the first 10,000 years and the small releases which would occur would peak only after more than 100,000 years.
Overall system performance is dominated by the near-field – the critical roles of the host rock are physical protection of the engineered barriers and ensuring relatively low groundwater fluxes through the repository. In the Reference Case, a very conservative description of the features in which groundwater flows through the basement is assumed due to the limited database available – more realistic descriptions imply that radionuclide retardation in the geosphere could also play a major barrier role.
Uncertainty in the long-term behaviour of the repository system is taken into account by analysing a range of alternative scenarios. In particular, a Robust Scenario is defined in which bounding estimates of the effects of parameters and processes are obtained by making very conservative assumptions. Even the very pessimistic combinations of conservative parameters and model assumptions within this scenario do not lead to doses above the regulatory guideline.
The results of the Kristallin-I project demonstrate that the crystalline basement remains an attractive option for a HLW repository host rock. An exploration strategy has thus been devised for Phases II and III of site characterisation from the surface and underground, respectively. This initially involves high resolution seismic surveys and an array of angled boreholes in order to further clarify key questions concerning the lateral extent of suitable blocks within the crystalline basement. Given the date specified for the next major milestone in the HLW programme – Project Entsorgungsnachweis by the year 2000 – It is prudent to minimise the risks of delay or of geological complications by focusing Phase II on the location of an existing deep borehole which suggests favourable local conditions – I.e. Leuggern or Böttstein.
The methodology and databases developed within Kristallin-I form an infrastructure for analysis of other types of waste which would be included in a deep HLW repository (spent fuel and various types of long-lived intermediate-level waste) and, indeed, also for assessment of the potential of the Opalinus Clay as an alternative host rock. The Kristallin-I Reference Case forms a benchmark for comparison of the expected performance of different waste type/host rock combinations.
The documentation of system understanding within the scenario development procedure allows potentially important areas of uncertainty to be identified, which serves as a basis for deriving a more detailed list of research objectives. The sensitivity analysis within the Kristallin-I performance assessment allows the relative importance of individual parameters to be evaluated, thus forming a basis for assigning research priorities. The Kristallin-I technical documentation also forms the basis for deriving suitable information packages for the general public. As the HLW programme moves into a site characterisation phase for both crystalline basement and Opalinus Clay options, informing the population in the investigation areas becomes increasingly important. To promote public acceptance, the general procedures used should be as transparent as possible. The Kristallin-I project was planned with such output in mind and it is intended that it will provide a template for future studies of this type.
In terms of long-term development of a national HLW strategy, this confirmation that crystalline rock is sufficiently promising to merit further local exploration, and the selection of the preferred sites for this exploration are the key end-points of the Kristallin-I synthesis. It is important to note that the specific location chosen – even in the event that very positive results are achieved in the exploration – is unlikely to be the exact location of a deep repository. On a local scale, optimisation of shaft access etc. would certainly lead to new shaft areas; on a regional scale the long times available for potential implementation mean that comparison with alternative sites in crystalline host rock is not excluded; on a national scale, the option of disposal in clay is available and on an international scale the attractiveness of joint projects is obvious. Although Kristallin-I and the following Project Entsorgungsnachweis 2000 will certainly not lead to immediate or even early implementation of a HLW repository, the goal of providing to the Swiss public a convincing and complete demonstration that we can, if necessary, dispose of HLW within our own borders remains a key element in our long-term waste management strategy.