Abstract. During the site selection process, regulated by the Site Selection Act (Standortauswahlgesetz – StandAG), the implementer has to identify adequate siting regions and has to perform long-term safety analysis for these regions. The Federal Office for the Safety of Nuclear Waste Management (Bundesamt für die Sicherheit der nuklearen Entsorgung – BASE) as the responsible federal authority has to review the implementer's long-term safety analysis. To perform this duty in the required depth, it will be necessary to recalculate important aspects of the analysis by means of numerical computer programs. In addition, this will allow assessment of the underlying uncertainties of the implementer's long-term safety analysis from a regulatory point of view. Numerical modelling requires a high degree of quality assurance. Therefore, it is of vital importance that the same problem is modelled with different computer programs and – if possible – by different teams of modelers. This strategy is known as the diverse modelling approach. The goal of this approach is to perform a cross-check of the computer programs in use as well as of the correctness of data handling, data interpretation, and model implementation. Comparing outcomes of different modelling teams also integrates complementary views and approaches, which can be beneficial when dealing with complicated problems. The diverse modelling approach forms the basis of regulatory modelling. It can also be implemented within organizations if more than one code is used to tackle the same task. The diverse modelling approach has been carried out at GRS gGmbH over
the past two decades by means of the computer programs, TOUGH2-GRS and
MARNIE(2). Both (co-)developed programs are thermohydraulic codes that
compute transport phenomena in porous media and can be coupled to
geochemical codes, see Navarro (2018) and Navarro et al. (2021) for
further details. TOUGH2-GRS and MARNIE(2) were applied in various
projects, such as preliminary safety analysis Gorleben (VSG,
Vorläufige Sicherheitsanalyse Gorleben), ZIESEL and EVEREST. A
recent example for the diverse modelling approach was the use of
TOUGH2-GRS and MARNIE for the development of indicators for the safe
confinement of radionuclides in a deep geological repository (Navarro
et al., 2019). A high degree of quality assurance is also achieved by involving
computer programs in benchmarks to compare results among different
codes for well-defined problems. For the calculation of THM (thermo-hydraulical-mechanical) processes TOUGH2-GRS has been coupled with the geomechanical computer program FLAC3D and this approach has been used in the benchmark BenVaSim (Seher et al., 2019). In this contribution selected works related to TOUGH2-GRS and
MARNIE(2) are presented. From this starting-point, ideas and concepts
for the continuation of development and quality assurance of the two
computer programs at BASE in the near future are outlined. These works
will contribute to strengthen the capabilities of BASE in the
independent review process of implementer's long-term safety analysis
within the site selection process.