OR/13/004 Introduction

Rochelle, C A, Purser, G, Milodowski, A E, Noy, D J, Wagner, D, Butcher, A, and Harrington, J F. 2013. CO2 migration and reaction in cementitious repositories: A summary of work conducted as part of the FORGE project. British Geological Survey Internal Report, OR/13/004.

Some repository concepts envisage the use of large quantities of cementitious materials — both for repository construction and as a buffer/backfill. Key aspects of these materials are their good mechanical properties and their ability to buffer pH to alkaline conditions. Such high pH conditions are important as they greatly limit metallic corrosion and radionuclide solubility — and as a consequence, radionuclide migration. Some wastes placed within a subsurface repository will contain a significant amount of organic material (e.g. ion-exchange resins, contaminated clothing etc). Over time, these may degrade to produce carbon dioxide (CO₂), which will react rapidly with cement buffer/backfill to produce carbonate minerals such as calcite. This may be beneficial in terms of immobilisation of ¹⁴C as a carbonate if it were present. Carbonation reactions occurring on the inside of the repository cement will be quite separate to those occurring on the outside of the repository cement (through interaction with surrounding carbonate groundwaters), though they may share very similar reaction mechanisms.

The conversion of cement minerals to carbonates will reduce the ability of the buffer/backfill to maintain highly alkaline conditions and as a consequence its ability to limit the migration of certain radionuclides. However, the reaction may also alter the physical properties of the buffer/backfill, possibly changing its permeability and strength. Although carbonation reactions might improve some properties, it is currently unclear whether the overall changes due to carbonation will be beneficial to long-term radionuclide immobilisation, or deleterious.

This study investigated the effects of cement carbonation processes via laboratory experiments conducted at a range of likely future in-situ repository conditions, including those expected over glacial timescales, which might influence the form of the carbon dioxide. Thus the experiments were run under conditions representative of those that might be found currently within a deep repository or potentially in the future, with temperatures between 20–40°C and pressures between 40–80 bar (equivalent hydrostatic heads of approximately 400–800 m) (Figure 1). Within this range of conditions, the phase behaviour of CO₂ is such that it can exist in gaseous, liquid, or supercritical states, as well as dissolved in water in contact with these states. The results of the experiments will serve as examples with which to test predictive modelling codes that incorporate reaction kinetics, and coupling between geochemical reaction, porosity changes and fluid flow.

Figure 1    Experimental run conditions plotted on a phase diagram for CO₂. Conditions were 20° and 40°C, and 40 and 80 bar, though only 3 of the possible 4 combinations of these were studied.

Two types of experiments were conducted within the BGS Fluid Processes Laboratories:

1. Numerous batch experiments to provide information on changing mineralogy and porosity upon reaction with free phase and dissolved gaseous, liquid and supercritical CO₂. A detailed description of these experiments is given in Rochelle and Purser (2010), Rochelle et al. (2014) and Milodowski et al. (2013). These experiments involved samples that were unconfined, with reactions controlled by diffusion of free-phase or dissolved CO₂ into the cement. Three cement pore fluids were utilised with compositions based on previous calculations of those anticipated in the vicinity of a cementitious waste repository (Bond et al., 1995a, b):

• A young pore fluid of high pH and with a significant component of NaOH and KOH (‘young near field porewater’ — YNFP).
• A slightly lower pH pore fluid with a significant proportion of Na, K and Cl, and some Ca and OH (‘YNFP + Cl’).
• An evolved pore fluid representing equilibration of groundwater with cement, and having lower pH and a chemistry dominated by Ca(OH)₂ (‘evolved near field groundwater’ — YNFG).

2. Four flow experiments with gaseous, supercritical and dissolved CO₂ were undertaken, and were aimed at providing information on changing permeability upon reaction. A detailed description of these experiments, and the results from them, is given in Purser et al. (2013a, b). These experiments involved samples confined within Teflon™ jackets, and held within a uniform confining pressure. Free-phase or dissolved CO₂ was pumped through the samples, with input/output flow rates and pressures used to derive permeability data.

Within a repository setting buffer/backfill cement has at least two key notable functions; to maintain highly alkaline conditions and hence reduce radionuclide solubility, and to allow controlled escape of gas away from waste canisters. As a consequence of the need for the latter of these functions, the cement is relatively permeable compared to some other cements (e.g. such as those used for borehole sealing). Some gases generated within the repository (such as hydrogen) may undergo very limited chemical reactions once formed, and hence their migration will be largely controlled by transport processes. CO₂ on the other hand, is highly reactive towards cement minerals, and so its migration will be controlled by a complex interplay between transport processes and chemical reactions.

The cement composition used was the Nirex reference vault backfill (NRVB) cement (Francis et al., 1997). It consisted of approximately: 26% Portland cement, 28.6% limestone powder, 9.8% lime, and 35.6% water. The mixture was cast into cylindrical blocks of approximate size 12 cm diameter by 30 cm long, and was cured for at least 40 days prior to subsampling for use in the experiments — see Rochelle and Purser (2010). Towards D3.6: results of the tests on concrete (part 1). Laboratory experiments at the BGS. FORGE project technical report. 16pp.</ref> for more details. Use of this cement allows the present study to be comparable with measured parameters/datasets from earlier studies. It is acknowledged however, that the compositions of backfill cements used in any future repository may well be different to that used in this study, but it is considered that many of the reaction processes we observed are likely to be common to a wide-range of cements.