OR/18/049 Introduction

Tamayo-Mas, E, Harrington, J F, Brüning, T, Kolditz, O, Shao, H, Dagher, E E, Lee, J, Kim, K, Rutqvist, J, Lai, S H, Chittenden, N, Wang, Y, Damians, I P, Olivella, S. 2018. DECOVALEX-2019 project: Task A - modElliNg Gas INjection ExpERiments (ENGINEER). Nottingham, UK, British geological Survey. (OR/18/049).

In 1999, Rodwell et al. stated “there are few problems in geoscience more complex than the quantitative prediction of gas migration fluxes through an argillaceous rock formation”. To understand this statement, it is necessary to appreciate why argillaceous materials (which include clays, claystones and mudrocks) differ from other clastic sedimentary rocks. Key factors in this respect include the sub-microscopic dimensions of the interparticle spaces, the very large specific surface of the mineral phases, strong physico-chemical interactions between water molecules and surfaces, very low permeability, generally low tensile strength, a deformable matrix, and a very pronounced coupling between the hydraulic and mechanical response of these materials. It is therefore necessary to consider these properties when defining the behaviour of these materials (both natural and engineered) in order to successfully represent flow in such systems.

With this in mind, the processes governing the movement of repository gases through engineered barriers and clay-rich host rocks can be split into two components, (i) molecular diffusion (governed by Fick’s Law) and (ii) bulk advection. In the case of a repository for radioactive waste, corrosion of metallic materials under anoxic conditions will lead to the formation of hydrogen. Radioactive decay of the waste and the radiolysis of water are additional source terms. If the rate of gas production exceeds the rate of gas diffusion within the pores of the barrier or host rock, a discrete gas phase will form (Wikramaratna et al., 1993; Ortiz et al., 2002; Weetjens and Sillen, 2006). Under these conditions, gas will continue to accumulate until its pressure becomes sufficiently large for it to enter the surrounding material.

In clays and mudrocks, four primary phenomenological models describing gas flow can be defined, Figure 1: (1) gas movement by diffusion and/or solution within interstitial fluids along prevailing hydraulic gradients; (2) gas flow in the original porosity of the fabric, commonly referred to as two-phase flow; (3) gas flow along localised dilatant pathways, which may or may not interact with the continuum stress field; and (4) gas fracturing of the rock similar to that performed during hydrocarbon stimulation exercises.

Figure 1    Conceptual models of gas flow (after Marschall et al. 2005), BGS © UKRI.

There is now a growing body of evidence (Horseman et al., 1996, 2004; Harrington and Horseman, 1999, 2003; Angeli et al., 2009; Harrington et al., 2017a and b) that in the case of plastic clays and in particular bentonite, classic concepts of porous medium two-phase flow are inappropriate and continuum approaches to modelling gas flow may be questionable, depending on the scale of the processes and resolution of the numerical model. However, the detail of the dilatant mechanisms controlling gas entry, flow and pathway sealing are unclear and the ‘memory’ of such features within clay may impair barrier performance, in particular, acting as preferential flow paths for the movement of radionuclides.

As such, development of new and novel numerical representations for the quantitative treatment of gas in clay-based repository systems are therefore required, and are the primary focus of Task A in the current phase of the DECOVALEX project. New numerical techniques provide an invaluable tool with which to assess the impact of gas flow on repository layout and therefore design of any future facility. In addition, the same processes and mechanisms described in such models are of direct relevance to other clay-based engineering issues where immiscible gas flow is involved e.g. shale gas, hydrocarbon migration, carbon capture and storage and landfill design.

The Task is split into four stages each building on the previous, representing an incremental increase in complexity:

1. Stage 0 (code development): analysis of data, conceptual model and process model development.
2. Stage 1: 1D gas flow

* 1D gas flow test on saturated bentonite under constant volume (Mx80-D).

3. Stage 2: spherical gas flow

   * A: spherical flow through saturated bentonite under a constant volume boundary condition (Mx80-10).

   * B (optional): spherical flow through saturated bentonite under a constant volume boundary condition (Mx80-A).

4. Stage 3: application to previous models to natural clay-based systems

   * A (optional): triaxial test on Callovo-Oxfordian claystone.

   * B (optional): gas flow in hydrated pellets under constant volume conditions.

This report summarises the outcomes of stages 0 and 1 with work conducted from May 2016 to March 2018 by the participating modelling teams

1. BGR/UFZ (Germany): Federal Institute for Geosciences and Natural Resources and the Helmholtz Centre for Environmental Research.
2. CNSC (Canada): Canadian Nuclear Safety Commission.
3. KAERI (Korea): Korea Atomic Energy Research Institute.
4. LBNL (United States of America): Lawrence Berkeley National Laboratory.
5. NCU/TPC (Taiwan): National Central University and Taiwan Power Company (Taipower).
6. Quintessa/RWM (United Kingdom): Quintessa Ltd on behalf of Radioactive Waste Management.
7. SNL (United States of America): Sandia National Laboratories.
8. UPC/Andra (Spain/France): Universitat Politècnica de Catalunya, funded by l’Agence nationale pour la gestion des des déchets radioactifs.

and presented at the following DECOVALEX workshops:

1. 1st DECOVALEX-2019 Workshop (Berkeley, USA, May 18–20, 2016): a general presentation of the task was given (available information and description of the gas flow tests) and a work plan amongst the teams was agreed. Teams were asked to develop one or more modelling approaches to address Stages 1 through 3. A range of approaches were encouraged from highly mechanistic models which may attempt to replicate nearly all aspects of experimental behaviour, to highly simplified homogenised approaches which aim to capture key features of the data.
2. 2nd DECOVALEX-2019 Workshop (Taipei, Taiwan, November 29–December 1, 2016): first analysis of the 1D gas flow together with any prototype numerical models were presented by the teams. A wide range of modelling approaches were shown.
3. 3rd DECOVALEX-2019 Workshop (Stockholm, Sweden, April 25–28, 2017): team approaches were collated and the first key comparison exercise was done. A first attempt to model the evolution in gas pressure and stress response matching breakthrough times and flux in/out of the core was made.
4. 4th DECOVALEX-2019 Workshop (Kingston, Canada, October 10–13, 2017): the different modelling approaches were compared and judged (based on timing, magnitude and shape). Results from that meeting (including work up to March 2018) are reported in this paper.

It is not the intention of this report to provide an exhaustive description of the individual contributions from each team, but rather give a technical overview and synthesis of key conclusions and results.