Description

Deep geological repositories (DGR) for radioactive waste rely on a multi-barrier system, consisting of engineered (ex. container, buffer, backfill, etc.) and natural (the host rock) components to contain and isolate the waste for hundreds of thousands to a million years. For DGRs in argillaceous rocks, the host rock plays a major role in retarding the movement of radionuclides should they break through the engineered barriers. The two agents of radionuclide transport are groundwater and gas generated from the repository. Argillaceous rocks in its undisturbed state have many desirable characteristics to retard radionuclides migration such as low permeability, low diffusivity, good sorption capacity, and in general a low fracture frequency. However, argillaceous rocks in general have lower strength compared to hard rocks, and thus are more susceptible to damage and deformation due to internal events and processes (such as excavation of the DGR, heat generation, gas generation) and external events (such as glaciation and earthquakes). The damage/deformation of argillaceous rocks can lead to an increase in permeability and diffusivity, resulting in higher rate of radionuclide transport by groundwater or gas. Currently, constitutive models to evaluate the long-term deformation/damage behaviour of argillaceous rocks are mainly based on macroscopic stress-strain observations from laboratory tests, such as triaxial tests. The assumptions used in these models need to be validated with microscopic observations, which is the novel approach being proposed here. Robust and reliable constitutive models developed in such a way would then be implemented in coupled H2M (two-phase fluid flow with mechanics) models to simulate gas injection tests with increasing injection pressure performed on Opalinus Clay (OPA) in the laboratory and at the Mont Terri underground research facility. These H2M models in turn could then be used with added confidence in the safety assessment of DGRs in argillaceous rocks.

Experimental Data

1) Laboratory uniaxial compression (UCS), Brazilian tensile (BTS) and consolidated undrained (CU) triaxial tests with pore pressure measurements followed by microscopic observations of damage and dilatancy. The tests were performed at Aachen University (RWTH) and the German Federal Institute for Geosciences and Natural Resources (BGR) at different loading orientations with respect to bedding (Figure 1). Stress, strain, strength and pore pressure evolutions as well as deformation patterns and localizations are the main output from these tests to be used for the development and calibration of constitutive models.

2) Gas injection tests at the British Geological Survey BGS.

These tests were performed by the British geological survey (BGS) using the stress path permeameter (SPP) shown in Figure 2. Gas was injected at increasing pressure into the OPA sample at a specified stress state. Fluid pressure at the injection inlet and flow outlet, radial and axial displacements, and injected and outflow gas volumes were continuously recorded. Samples with bedding parallel and perpendicular to gas injection were tested.

3) Gas injection (GT) experiment at the Mont Terri Underground Research Facility

The GT experiment is schematically illustrated in Figure 3. Gas is injected in a central borehole, and deformation and pore pressure are measured in monitoring boreholes located 1m radially from the injection borehole.

Approach

An incremental approach is adopted in HyMAR, with 3 subtasks being proposed as follows.

Subtask 1 focusses on the development of constitutive models using UCS, BTS and CU tests performed at RWTH and BGR, in conjunction with microscopic observation of damage and microfracturing. The models should be able to predict the evolution of pore pressure, the onset of dilatancy, strain localization and shear banding, and the influence of bedding orientation.

Subtask 2 consists of implementing the constitutive models developed in subtask 1 to simulate gas injection in OPA samples (at two different bedding orientation) performed by BGS.

Subtask 3 consists of adapting the model developed in subtask 2 to perform blind predictions, followed by calibration of the GT experiment at Mont Terri.

Participating Groups

• Canada: CNSC
• China: CAS
• France: ANDRA (UPC)
• Germany: BGR/BASE
• UK: NWS (Quintessa Ltd)
• USA: Department Of Energy (Lawrence Berkeley National Laboratory)
• Switzerland: ENSI

Further Information

For further information, please contact the task leader, Son Nguyen.

References

1. E.E. Dagher, T.S Nguyen and J.A. Infante-Sedano, 2021. Investigating models to represent gas transport in a swelling geomaterial, IJRMMS, 137.
2. R. Cuss, J. Harrington, R. Giot and C. Auvray, 2014. Experimental observations of mechanical dilation at the onset of gas flow in Callovo-Oxfordian claystone, Clays in Natural and Engineered Barriers for Radioactive Waste Confinement. Geological Society, London, Special Publications, 400, http://dx.doi.org/10.1144/SP400.26
3. T.S. Nguyen and D.A. Le, 2014. Development of a constitutive model for a bedded argillaceous rock from triaxial and true triaxial tests, Canadian Geotechnical Journal, d.o.i. 10.1139/cgj-2013-0323.
4. T.S. Nguyen and D.A. Le, 2014. Simultaneous gas and water flow in a bedded argillaceous rock, Canadian Geotechnical Journal, d.o.i. 10.1139/cgj-2013-0457
5. Lisa Winhausen et al.(2021). Micromechanisms leading to shear failure of Opalinus Clay in a triaxial test: a high-resolution BIB–SEM study, Solid Earth, 12, 2109–2126
6. Lisa Winhausen et al. (2022). Failure mode transition in Opalinus Clay: a hydro-mechanical and microstructural perspective, Solid Earth, 13, 901–915
7. Lisa Winhausen et al. (2023). The anisotropic behavior of a clay shale: strength, hydromechassnical coupling and failure processes, Solid Earth, 128