Building the bridge between safety requirements and numerical modeling: an example considering crack development of Opalinus clay in laboratory and field scales
Tuanny Cajuhi,Jobst Maßmann,Gesa Ziefle
DOI: https://doi.org/10.5194/sand-1-165-2021
2021-11-10
Safety of Nuclear Waste Disposal
Abstract:Abstract. Salt, crystalline and clay formations are under discussion as potential host rocks for storage of heat-generating radioactive waste. Each of these rocks has a different structure and composition, and consequently a different material behavior. The latter needs to be studied and evaluated with respect to the main aim: to find a place to store the waste in a safe and sustainable manner. Several requirements in the context of the safety of a repository need to be fulfilled, concerning the long-term as well as the operational phase. One key point in this matter is the integrity, which refers to retention of the isolating rock zone's containment capabilities. With the focus on some experimental and numerical investigations on the excavation influenced near-field behavior of Opalinus clay (OPA), this contribution aims to illustrate an example for the role of numerical modeling in safety assessment. Once, e.g. anthropogenic action such as excavation starts, the natural state of equilibrium in the formation is disturbed. Trying to restore it, the rock deforms (convergence) and/or releases energy in other ways such as cracking. This could lead to loss of integrity since crack nucleation and propagation can affect the mechanical stability and create paths to transport contaminants. During operation in the excavated rock, environmental changes, e.g. temperature and humidity, further affect its behavior. The understanding of these dynamic phenomena ideally needs to occur at the in situ scale; however, performing an experiment in the spatial and time scales of interest is not always possible. For this reason, the in situ problem needs to be formulated, abstracted and mathematically modeled. The interpretation of the results must take place with simplifying assumptions and complementary laboratory scale experiments can be used to improve understanding of the system. The real problem is approached stepwise, each step associated to the size of the model and its complexity. The gradually obtained knowledge is necessary to achieve a better understanding of the process and to evaluate the capacities and limitations of the models. This contribution aims at showing the basic practical steps for numerical modeling with particular focus on the preparation and interpretation of the models and results, e.g. model calibration, verification and validation. As an example, the OPA at the Mont Terri site is chosen. The material parameters are obtained either experimentally or from the literature. We choose and perform laboratory scale simulations that are related to nearly the same mechanism as in the in situ scale. To have a first impression on the latter, a simplified, large-scale numerical model is prepared. The mechanism in study is drying and wetting, which is associated with shrinkage and swelling. We analyze the pore pressure and stress development in both scales. Thus, hydraulic mechanically coupled approaches are essential. The concept of effective stress is used, which combines the contributions of the solid and fluid phases (gas and liquid). In the current modeling approach, the gas pressure remains constant (atmospheric pressure) and during drying, the liquid pressure induces capillary pressure development and decrease of saturation. The laboratory scale simulation is important to evaluate the model of choice and to assess potential numerical problems. Furthermore, it can be used to perform a sensitivity study of material and numerical parameters. This step is necessary during the development or extension of numerical models as well as to evaluate their applicability on new research questions. The simplified in situ scale numerical model is then extended. In this phase the numerical model is evaluated once again, especially with respect to its complexity. Furthermore, specific questions related to this scale are posed: overall behavior of the rock, influence of the excavation, seasonal and long-term effects. In this contribution we deal with the long-term cyclic deformation (CD-A) experiment. The CD-A experiment has been taking place in the Mont Terri Rock Laboratory since October 2019. It consists of twin niches, a closed and an open niche, subjected to either high air humidity or seasonal humidity changes leading to saturation/desaturation during summer/winter in the OPA, respectively. Several parameters are periodically or continuously measured, including relative air humidity, convergence and crack development. We attempt to transfer the knowledge and numerical models developed in the small scale to the large scale and to evaluate the possibilities and limitations of the chosen approaches by comparing the numerical and experimental results.