Rock single rough fractures: spatial continuity applications, coupled hydraulic-mechanical numerical models and their validation against experiments

Radioactive waste management is an increasingly important environmental problem facing countries and nuclear entities worldwide. Currently, the overall international consensus is to package the waste in Engineered Barrier Systems (EBS) and safely dispose of it in an underground Geological Disposal Facility (GDF) excavated in a suitable low-permeability host rock several hundred meters below the surface. The main risk for disseminating the radionuclides into the biosphere is through their transportation in subsurface fluids which, in a low permeability host rock, will be mainly achievable through the host rock’s fractures. Similar challenges are faced in containment (temporary storage or permanent disposal) or extraction of fluids in industrial applications such as CO2 and hydrogen or geothermal energy. Fractures’ ability to conduct fluids (their permeability) is dependent on their aperture (the space between the two fracture’s surfaces) which in turn is a function of the physical properties of the rock (namely its Poisson ratio and modulus of elasticity), the fracture surfaces geometries and the coupled processes acting on the fracture at each point. Coupled thermo-hydraulic-mechanical-chemical (THMC) processes are processes where inputs such as temperature, fluid pressure, mechanical stress or chemical species’ saturations, impact other THMC processes, hence termed coupled. In other words, coupled THMC processes directly interact with and influence each other, for instance the local stress field magnitudes and orientations (mechanics), the temperature (thermal), the fluid pressure (hydraulic) and the chemistry (chemical) of the fluids and rocks. Conceptual models are the simplified design of complex systems. Numerical models are computer adaptations of such conceptual models, used to assess how those complex systems function. To study the mechanisms of fluid flow through single rough fractures, numerical THMC coupled processes models are used because they simulate complex coupled THMC systems, advancing the understanding of the systems’ behaviour. These models rely on the fractures’ rough surfaces or aperture representations to simulate fluid flow with changing aperture distributions based on changes in THMC processes conditions. Single fractures’ are generally deterministically represented by acquiring the fractures’ surfaces’ roughness information which are then digitally stored. These digital files, however, can be acquired with a digital resolution that is too high (has too many data points) to create computationally efficient numerical models. One solution is to selected subsets of that original dataset to create a lower resolution model or they need to be upscaled (coarsening their resolution). Traditional upscaling methods average the points onto the model’s mesh. Hence, the methods chosen to lower the resolution of the data to create the numerical model neglect the inherent spatial correlation, or spatial continuity, of the dataset. This thesis starts by characterising a rock’s rough fracture’s surface using its inherent spatial continuity information through semi-variograms. Semi-variograms are the functions that describe correlations between the points in both distance and direction defined through a mathematical technique, which can be used to create more accurate upscaled representations of fracture models. Two Finite Element Method (FEM) hydro-mechanical numerical models were created from the aperture field data of rock fractures upscaled utilising the spatial continuity techniques and an arithmetic averaging, therefore allowing comparison of results. More accurate results were observed in the models upscaled using spatial continuity. The open-source FEM OpenGeoSys 5 (OGS-5) code utilised is uses an initial aperture distribution and the cubic law for calculating the permeability field from the aperture field. The aperture field is calculated from the normal stress acting on each element. As the normal stress on each elemental is applied, it affects the aperture of the fracture at the element level which in turn affects the fluid pressure. The fluid pressure throughout the fracture was calculated numerically using OGS. A hydro-mechanical model depicting fluid flow through the fracture amid changes in the fracture aperture due to mechanical stress was created and validated against experiments conducted in the Geo-Reservoir Experimental Analogue Technology (GREAT) cell. The GREAT cell is an apparatus capable of simulating at-depth polyaxial stresses on 200mm diameter by 200mm length cylindrical samples with fluid flow capabilities. The OGS-5 code used wasn’t intially designed to cope with fluid pressures above the confining pressure at the mesh’s element level, corresponding to the element’s aperture increase, i.e. fracture opening. This thesis incorporates adaptations to that code to allow effective stress to be negative thus accommodating fracture opening. Finally, the work presented here also attempts to find the Representative Elementary Area (REA) of the aperture field data using its spatial continuity techniques introduced. It analyses the spatial continuity of the fields upscaled at different scales using traditional upscaling methods. It uses the raw data spatial continuity (highest resolution) as the comparison standard. The assumption is that the spatial continuity parameters change as the resolution of the data decreases. The criterion for choosing the REA is that the behaviour of the spatial continuity of the standard can be reasonably reproduced (i.e. the parameters are similar) within the scope of the work intended with the upscaled data. In summary, the work contained in this thesis encompasses advances in methodology in the analysis and interpretation of rock fracture roughness and aperture fields. It provides applications to REA determination and upscaling methods for coupled THMC processes numerical modelling. Therefore, the work provides new tools to tackle uncertainty regarding processes such as contaminant transport in the disposal (e.g. CO2 and radioactive nuclides) which are crucial in applications such as seasonal fluid storage in geo-reservoirs (e.g. natural gas or hydrogen) or geothermal energy extraction.

Quintessa Ltd

Nuclear Waste Services UK

Engineering and Physical Sciences Research Council (EPSRC) Smart Pulses for Subsurface Engineering