Long-term gas migration through clays cannot be simulated by conventional two-phase flow models alone owing to the presence of material deformation. In this article, an extended twophase flow model that incorporates mechanical effects is proposed. The model allows the formation of preferential pathway and considers the relation between pore moisture and pore deformation. It was carried out with the intention of avoiding the complexity of a fully coupled thermal, hydraulic and mechanical modelling. In the new model, porosity, permeability, swelling pressure and pathways formation threshold depend on the water saturation. The model is validated on different gas injection experiments with controlled flow rate and controlled pressure. Some experiments are well known in the literature; some are new. In each case, an inverse approach is used to identify the model parameters. The results confirm that, depending on the type of bentonite (MX80, Avonlea, KunigelV1), modelling the gas migration could require the existence of a pressureinduced saturation-depending preferential pathway. In laboratory-scale experiments, the model leads to an accurate evaluation of the long-term gas migration trends, including not just the gas migration stage but also the water re-saturation level. In a field-scale experiment, the behaviour of the model in a realistic context is revealed.

The Radioactive Waste Management Funding and Research Center (RWMC) is leading a research programme to evaluate the gas transport mechanisms through a TRU (TRans-Uranium) waste disposal facility in Japan and acquire information on gas migration properties. In this paper we describe a series of laboratory gas injection tests using the bentonite adopted for use in Japanese TRU disposal, as well as an attempted visualization of the gas migration path inside the bentonite used as a buffer. By building a conceptual model from the results of these tests, the characteristics of gas migration through to breakthrough for bentonite can now be better understood in the context of Japanese TRU disposal. Thanks to the outcome of this research project, advanced knowledge may be applied to the conceptualization of TRU waste disposal.

Bentonite is one of the safety-critical components of the engineered barrier system for the disposal concepts developed for many types of radioactive waste. However, bentonite — especially the swelling clay component that contributes to its essential barrier functions — is unstable at high pH. To date, results from laboratory tests on bentonite degradation have been ambiguous as the reaction rates are so slow as to be difficult to observe. As such, a key goal in this project is to examine the reaction of natural bentonites in contact with natural hyperalkaline groundwaters to determine if any long-term alteration of the bentonite occurs. Ophiolites have been identified as sources of hyperalkaline groundwaters that can be considered natural analogues of the leachates produced by some cementitious materials in repositories for radioactive waste. At the Zambales ophiolite in the Philippines, widespread active serpentinisation results in hyperalkaline groundwaters with measured pH values of up to 11.7, falling into the range typical of low-alkali cement porewaters. These cements are presently being developed worldwide to minimise the geochemical perturbations which are expected to result from the use of OPC-based concretes (see Kamei et al., this conference, for details). In particular, it is hoped that the lower pH of the low-alkali cement leachates will reduce, or even avoid entirely, the potential degradation of the bentonite buffer which is expected at the higher pH levels (12.5 and above) common to OPC-based concretes. During recent field campaigns at two sites in the Zambales ophiolite (Mangatarem and Bigbiga), samples of bentonite and the associated hyperalkaline groundwaters have been collected by drilling and trenching. At Mangatarem, qualitative data from a ‘fossil’ (i.e. no groundwater is currently present) reaction zone indicates some alteration of the bentonite to zeolite, serpentine and CSH phases. Preliminary reaction path modelling suggests that the zeolites could have been produced as a product of smectite reaction in the hyperalkaline groundwaters. Although not included in this calculation to date, the CSH phases identified are completely consistent with reaction of clays with hyperalkaline groundwaters, as seen at other sites worldwide. At the Bigbiga site, an active hyperalkaline groundwater/bentonite reaction zone (at the base of the bentonite deposit) has recently been identified and a drilling campaign is planned for late autumn 2010.

No abstract

No abstract