Building of a model which could be used for optimization of design of nuclear waste storage, is one of biggest challenges in field of coupled processes modeling. Such a model have to take in consideration several fully-coupled processes. These processes are from field of Thermal, Hydro, Mechanical and Chemical science in short, they are described as THMC processes. Most of all, it is interaction of heat, generated by residual radiation in the waste, reacting with underground water and eventual chemical reaction in these water-solution. And all these actions being still a subject of eventual physical movements in the bedrock. In addition, model must take into consideration that values of a given process can change dramatically importance of this process characteristics for the safety of the storage. None of existing models is reaching the level of model complexity needed to address these issues. In this article is presented an object oriented model ISERIT. Model ISERIT is able to solve multidimensional task THH, where HH is for water, and water being present in 2 phases. The first phase is water vapor and the second phase is water absorbed by clay particles of the container buffer. In theoretical part of the article are defined governing equations. The governing equations are based on the conservation equations of heat and mass. The continuum equations are discretized in space by using the Galerkin finite element formulation. The time discretization is solved by implicit finite difference scheme. The main part of the article describes implementation of model ISERIT. Main structure of the model is defined with functionality of significant classes. The temperature, the water vapor concentration and the water concentration in solid phase are chosen as the three primary variables. The parameters, such as heat conductivity, heat capacity and water vapor diffusion coefficient, could be taken to be constants, or could be allowed to vary with temperature and water vapor concentration, without requiring fundamental modifications to the code. Article also describes how model can be enlarged to incorporate eventually other processes. This thanks to the fact, that model was build as object designed. At the end, article presents verification of the model against the experimental results, laboratory experiment BenchMark 1.3, as well as against full-scale experiment BenchMark 2.1.

This content is only available via PDF.