THM Modelling of expansive clays incorporating uncertainty quantification
Conference Paper
Overview
Additional Document Info
View All
Overview
abstract
The design of repositories for high level nuclear waste involves simultaneous changes in the thermal, hydrologic and mechanical fields, what generally requires the use of coupled Thermo-Hydro-Mechanical (THM) formulations The THM formulations are generally composed of a number of balance equations for porous media; a set of constitutive equations; and equilibrium restrictions as components of this process To simulate these behaviours, numerical models are implemented with a number of parameters describing the intrinsic thermal, hydraulic and mechanical behaviours The parameters are often determined by using deterministic approaches by fitting a model over experimental observation However, certain uncertainties exist over determining the parameter value, due to which the model output would not fully capture actual observations This work explores a methodology of capturing these uncertainties using probabilistic calibration methods, particularly using Bayesian paradigms and Markov chains The results of using such methods are then contrasted with the original numerical results and the actual observations The temperature response was in agreement with the model response, although it failed to capture the temporal evolution accurately The model predicted the trend of water content, but shows variation with the actual values The uncertainty in thermal conductivity has no influence; the model response is under predicted The uncertainty in predicting the dry density at the initial stages persisted even after propagating the variation in both thermal and hydraulic parameters This suggests that there could be some other parameters whose variation has to be taken into consideration Correlation coefficient between the output response at different locations and time are shown
name of conference
International Symposium on Energy Geotechnics SEG-2018: Current Advances and Challenges Associated with Geological Disposal for High Level Nuclear Waste: From Laboratory to Field Scale Tests and Modeling Presentations
