Comparing apples and oranges? An interview with GRS repository expert Jörg Mönig on the challenges posed by the site comparison
Mr. Mönig, you are an expert on the topic of disposal and Head of the Repository Safety Research Division of GRS. The German federal government has recently passed an amended Site Selection Act. What does this Act say?
Mönig: The Site Selection Act is to regulate the search for a suitable repository site in Germany. This repository is to house above all the high-active heat-generating waste from Germany under demonstrably safe conditions for a period of one million years. The Act says that this search is to be carried out on the basis of scientific knowledge and in a transparent manner.
What is the selection procedure to be like?
Mönig: The selection procedure for a repository site is to have three phases. The starting point will be a blank map of Germany, where potentially any location is possible. All host rock types occurring in Germany are to be considered equally in the selection.
In the first phase, those regions will be excluded which are out of the question due to certain exclusion criteria or which do not fulfil the minimum requirements for a repository site. For example, regions in which an earthquake may occur will be excluded. In a second phase, surface exploration and comparative analyses will be carried out. In the third phase, the locations that remain are to be explored underground. This will involve e.g. the preparation of comprehensive safety analyses and the comparison of possible sites. A site is then to be selected on this basis, also taking social-scientific and spatial-planning criteria into account.
What exactly is such a repository to look like? And how should we image the storage of the waste in the repository?
Mönig: The scientists are agreed that radioactive waste should be stored deeply underground. This is also what the Commission on the Storage of High-Level Radioactive Waste recommended. Deep underground, the waste is far removed from man and the environment. The waste is to be emplaced in a so-called containment-providing rock zone (CPRZ). This is the area of the repository which ensures containment of the waste in combination with technical barriers. The highest goal here is to confine the waste in this rock zone permanently and without requiring active measures in the future.
When deciding in favour of a repository site, one has to demonstrate scientifically that the radioactive materials - if any - will only be able to leave this CPRZ in very small amounts. In Germany, this has to be demonstrated for a period of one million years.
What influence does the host rock have on the containment of the waste?
Mönig: The host rock has a big influence on safe containment. It counts among the barriers that are to ensure that the radioactive materials will remain inside the repository. As for Germany, there are three types of host rock that come into question: salt, clay, and crystalline rock. Each has special characteristics that make it interesting with regard to the storage of hazardous pollutants.
Salt rock, for example, is highly capable of removing the heat emanating from the waste. Moreover, salt distinguishes itself in its flow characteristic. This ensures that the emplaced containers will be confined by the salt over the years. However, each host rock type also has its drawbacks. Salt rock is soluble in water, for example. Still, the host rock is only one of the barriers. The natural conditions in the repository are to be supplemented by technical barriers (e.g. suitable containers, backfill materials, and sealing structures). Hence the repository concept must be adapted to the local conditions. This means that there cannot be one single repository concept for all sites. Depending on whether one wants to build a repository in crystalline rock, in salt rock or in clay, the repository systems to be considered are therefore very different.
You have just outlined the selection procedure for a site. Whether a site is suitable or not is to be examined and demonstrated by means of safety analyses. How are such safety analyses carried out?
Mönig: To perform these safety analyses, complex computer codes are used. They simulate which geological, physical and chemical processes develop in a repository over a period of one million years. To this end, these codes are fed with all available data relating to the site. You need to have a good knowledge of the site to do so. An even then will there be a number of things that are only roughly known. Therefore the simulation codes have to be capable of handling uncertain data, too.
The model calculations that calculate the release of nuclides and the associated consequences are referred to as long-term safety analyses. The analyses are part of the so-called safety case. The methods behind this safety case are largely the same internationally. A safety case contains amongst other things the results of the safety assessments and the long-term safety analyses as well as a discussion of the robustness and reliability of the repository and its design, and it describes the confidence level of the safety analyses.
In this procedure, how can sites in different host rock types be compared? Is that at all possible?
Mönig: The results of the safety analyses are typically dose intervals. They indicate how high the possible radiation exposure around the repository will be in the future. However, merely comparing dose intervals will not get you very far. Here, only the quantitative arguments would play a role. But there are always also qualitative arguments for or against a repository site.
In the VerSi research project, GRS has examined methods for comparing repository sites of different host rock types. We did this as an example for clay and salt. For the direct comparison of two sites we have developed two fundamentally different methods: a qualitative, verbally argumentative method and a quantitative method that is based on probabilistic long-term safety analyses. The two methods led to different results. Nevertheless, they supplement each other very well. We are currently looking at whether these methods can also be applied to crystalline rock as host rock type.
Still, things remain difficult. It is a bit like comparing apples and oranges. In the end it will therefore be a big challenge to simply pass a judgement like "Repository system A is better than repository system B".