Development of an upscalable HM model for representing advective gas migration through saturated bentonite
Introduction
Many Geological Disposal Facility (GDF) concepts for radioactive waste expect gas to be produced through mechanisms including corrosion of metals (e.g. H2), radioactive decay (e.g. Rn) and radiolysis of water (H2), as discussed by Rodwell et al.1 It is therefore important to understand the migration of gas (by molecular diffusion and bulk advection) that has been generated within the repository context, both in terms of bulk gas (potential GDF and host rock pressurisation) and trace gas transport. For disposal concepts involving bentonite as a barrier within the GDF and/or clay host rocks, a discrete gas phase could form if the rate of gas generation exceeds the rate of diffusion of the gas into the clay and/or the rate of dissolution of gas into the pore water. Transport of water, bulk gas and dissolved gases through multi-phase flow processes is well understood for traditional porous media. However, models using only these processes are unable to represent the complex behaviour of gas flow in water-saturated clays, where gas migrates through the creation of additional flow paths.2,3 These emergent gas pathways are observed and conceptualised by Marschall et al.4 via the processes of pathway dilation and development of tensile fractures. Such processes can change the hydraulic and mechanical properties of the medium suggesting strong hydro-mechanical (HM) coupling in the system. It is therefore important in terms of a GDF safety case and performance assessment (PA) context to develop a sufficient understanding of these gas migration processes, since they could affect the ability of clay buffers and seals to perform as required by the safety case.
Two gas flow experiments carried out by the British Geological Survey (BGS) have been considered as part of Task A of the DECOVALEX-2019 (D-2019) project, whereby gas is injected into a small fully-saturated compacted MX-80 bentonite cylindrical sample under constant volume conditions. The first is a one-dimensional gas injection test5 in which gas is injected from one end of the sample and outflow measured at the opposite end. The second experiment is a point injection gas test6 where gas is injected through a steel rod into the centre of the sample and outflow measured at the outer radial surface. These experiments are analysed to build a conceptual understanding of the key processes involved in such gas tests and then used to inform the numerical model presented herein. One of the important findings for modelling these experiments has been to include an accurate representation of the gas injection pressure boundary condition, which is tightly coupled to the rate at which gas is able to enter the clay. It was therefore found to be necessary to explicitly model the gas injector mass balance, to allow proper comparison to the experimental data.
A bespoke fully-coupled hydro-mechanical model has been developed7, 8, 9 to help conceptualise and understand the D-2019 Task A experiments, which represents gas flow through dilating capillaries with a separate permeability to water. This approach follows conceptually from previous work10,11 whereby gas migration is modelled through a predefined network of capillaries within MX-80 bentonite. Modelling the gas using a single-phase flow approach was found to be sufficient to capture the key gas migration features from these laboratory-scale experiments, but does not preclude using the model in conjunction with conventional multi-phase flow approaches. To enable the model to be applied in future to upscaled geometries it has been developed using a continuum modelling approach. This was one of the main objectives of the model since upscaling (in space and/or time) may be necessary if the model is to be applied in a PA or GDF safety case context. For this reason, development of the model has also focussed on representing the key deterministic features from the experiments rather than reproducing fluctuations (e.g. gas “burst-type” behaviour) relating to pathway instability as observed in the experimental data.
The setup and results of the D-2019 Task A gas flow experiments are described in Section 2 and the formulation of the model is given in Section 3. Results are compared with these experiments in Section 4, where the MX80-8 dataset from Harrington and Horseman12 is used as an additional dataset outside of D-2019 Task A to build confidence in the modelling approach. This is followed by some discussion in Section 5 of conceptual and modelling uncertainties, and consideration for possible future application to radioactive waste disposal systems.
Section snippets
One-dimensional gas flow test – Experimental details and observations
This experiment is a one-dimensional gas injection test performed on a fully-saturated bentonite sample. Within D-2019 Task A, it is referred to as the ‘Stage 1A’ experiment. The experimental regime consists of an initial saturation period using deionised water followed by the main phase of gas testing with helium. Full details of the experiment are given in Daniels and Harrington.5
The cylindrical sample is prepared from compacted MX-80 bentonite (diameter: 59.59 mm; length: 119.88 mm; mass:
Conceptual model
A conceptual understanding of the main processes and phenomena that determine flow of gas through clay is important for informing the numerical model. With the objective to develop an upscalable model for PA or GDF safety case applications, the focus of the model is to be as simple as possible and be capable of reproducing only the key deterministic features from the experiments. Thus, the ambition is not to reproduce every fluctuation or stochastic variation observed in the data from these gas
Main modelling results
The modelling results for the one-dimensional gas flow and point-injection experiments are shown in Fig. 13 and Fig. 14, respectively. Table 4 lists the parameters used for both models. Note that for the purposes of these graphs, mass flow rates (mol s−1) for the experiment have been determined from the experimental data.
For the one-dimensional gas test, the injection pressure, flow rates and stress measurements from the experiment are all extremely well captured by the model (Fig. 13).
Conclusions
A novel upscaled model for representing gas flow through dilating capillaries has been developed and applied to three experiments. The model differs from conventional multiphase flow models of the advection of gas in clay in that it directly represents the development of gas pathways in the clay, rather than assuming that gas displaces water in pre-existing ‘pore space pathways’. Net gas saturations in the overall clay volume are very low, consistent with experimental observations, in contrast
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This work was funded by Radioactive Waste Management Limited (RWM; https://www.gov.uk/government/organisations/radioactive-waste-management) as part of participation in the international DECOVALEX project (https://decovalex.org/).
DECOVALEX is an international research project comprising participants from industry, government and academia, focusing on development of understanding, models and codes in complex coupled problems in sub-surface geological and engineering applications; DECOVALEX-2019
References (22)
- et al.
Gas network development in a precompacted bentonite experiment: evidence of generation and evolution
Appl Clay Sci
(2017) A multiple front propagation model for gas migration through clay
Eng Geol
(1999)- et al.
Gas Migration and Two-phase Flow through Engineered and Geological Barriers for a Deep Repository for Radioactive Waste. A Joint EC/NEA Status Report Published by the EC, European Commission Report EUR 19122 EN
(1999) - et al.
Evidence for gas-induced pathways in clay using a nanoparticle injection technique
Mineral Mag
(2012) - et al.
Final Report of FORGE 3.2.1: Key Gas Migration Processes in Compact Bentonite. FORGE Report D3.33
(2014) - et al.
Characterisation of gas transport properties of the Opalinus Clay, a potential host rock formation for radioactive waste disposal
Oil Gas Sci Technol
(2005) - et al.
The Response of Compact Bentonite during a 1D Gas Flow Test. British Geological Survey Open Report
(2017) - et al.
RWM Coupled Processes Project: First Annual Report for DECOVALEX-2019. Quintessa Report to RWM QRS-1612d-R1 v1.2 (Independent Peer Review Organised by RWM)
(2018) - et al.
RWM Coupled Processes Project: Second Annual Report for DECOVALEX-2019. Quintessa Report to RWM QRS-1612d-R2 v2.1 (Independent Peer Review Organised by RWM)
(2019) - et al.
RWM Coupled Processes Project: Third Annual Report for DECOVALEX-2019. Quintessa Report to RWM QRS-1612d-R3 v1.2 (Independent Peer Review Organised by RWM)
(2020)