Long-Term Cement Studies (LCS) - Phase 1
Phase 1 of the LCS project was mainly focused on the design and installation of the 2 in-situ experiments. The two main tasks were: a) the definition of the source term (e.g. OPC or low-alkali cement), and b) the design of the actual experiments.
Definition of the source term
In Phase 1 (2006-2009), two in-situ experiments were defined according to the wishes of the partners. At the LCS partner meeting in May 2008 in Meiringen it was decided:
- to focus on hydrochemical aspects of cement application
- to use two different source terms, namely a liquid grout source and a pre-hardened cement source
- to apply OPC
- to use the cement components applied by Posiva (Ultrafin 16 and Mighty 150 as superplasticiser)
The main reasons for initially focusing on OPC were: a) the better integration of former results from HPF (using an OPC solution); and b) the lower risk to obtaining weak hydrochemical signals at the observation points. The main reason of choosing the Finnish material was mainly the high level of knowledge about the material (e.g. Vuorinen et al., 2005; Kronlöf, 2005; Sievänen et al., 2006; and Raivio & Hansen, 2007). In particular, the leaching tests performed at Posiva were used for the detailed design of the experiments.
During the planning of LCS Phase 2 it was decided to perform a low-pH cement laboratory experiment with the same components as the ones used by Posiva (i.e. GroutAid).
Design of a long-term in situ experiment
The main task of LCS Phase 1 was the design and setup of the two experimental sites. Some of the issues concerning a long-term, low-maintenance experiment include:
- Confined natural flow field in a shear zone towards a tunnel (hydraulic sink, confinement of radionuclides, avoiding clogging problems)
- Measurement and sampling in boreholes and potentially at the tunnel wall
- Well defined source term
- Advantageous design for long-term maintenance of stable conditions.
From a modelling point of view, a number of requirements can also be formulated.
- Good knowledge of the hydraulic, transport and geochemical initial conditions is important
- More non-sorbing and sorbing tracer experiments in the flow field have to be performed to characterise the local transport properties of a shear zone volume where the experiment will be performed
- Quantification of changes induced by the alteration process should be possible, eg changes of sorption behaviour before and after leachate interaction
- Constant head or natural flow conditions should be established to investigate long term processes instead of constant flow (which better fits to one-dimensional modelling codes).
This required a detailed study of two potential sites (BK cavern and VE tunnel) leading to the selection of the experimental area in the VE tunnel. The main reasons for selecting the VE tunnel was: a) that the chosen shear zone is the same as the one used for the HPF experiments, thus enabling more straight forward conclusions; and b) that the fractures are well defined and more or less parallel to each other.
In principle the two experiments are organised the same way. However, due to their different design and setup the detailed organisation is expected to differ slightly.
The two experiments were set up according to the conditions defined above. The detailed documentation of the experimental sites and the experimental setups are given in NAB 08-14. Figure 1 shows a 3D picture of the field setup.
Figure 1: experimental setup of 2 in-situ experiments.
In LCS Phase 2, the following sub-tasks are planned to be performed:
- Long-term monitoring of the experiments. This includes continuous measurement of the hydraulic and some hydrochemical (pH, Eh, EC, T) parameters, and regular sampling for laboratory analyses (e.g. major chemical composition, stable isotopes, etc.)
- Analysis of cement elements to determine the degree of cement leaching over the duration of the experiment
- Regular hydraulic and tracer tests to quantify alterations of the water flow paths
- Excavation of cement sources and altered rock sections
- Analysis of cement and rock samples
Laboratory experiments are important to bridge the field scale and modelling/interpretation/system understanding and three types of laboratory experiments can be important in supporting LCS
- Complex experiments (scaled, but with relevant complexity to a repository or field situation).
- Feasibility/performance of components used in field experiments (eg OPC/LACP).
- Solubility, speciation, and sorption experiments to constrain basic thermodynamic and kinetic data.
In any case, short term tasks to support the experimental design should have clear priority over those experiments which will support the long-term concept and mechanistic understanding.
The feasibility of using a solid source has not yet been assessed in detail. In this relatively simple demonstration experiment, solid sources of both ordinary Portland cement (OPC) and LACP could be used in a simple setup to record the release of cement leachates under various geometries and different cement permeabilities (as discussed under Task 1 above). This could be expanded later to include some radionuclide tracers in the solid source and observe their release to provide the source term for the modelling task.
The cementitious materials and the equipment were tested prior to starting the experiment and scoping calculations were performed to assess the strength of the expected high-pH plume (Rueedi et al., 2009). The details of the material testing will be documented in Nagra reports. The main objective of the experiment was, besides the testing of the experimental setup, the assessment of the signal strength in the column outflow under different flow rates (Fig. 2).
Figure 2: a) Schematic drawing of laboratory experiment using a pre-hardened cement source (orange cylinder) emplaced into an overcore from GTS. b) measured concentrations of anions in column outflow under different flow velocities.
In Phase 2 it is planned to set up two similar laboratory experiments, one with OPC and one with LACP, and monitor the column outflow for 2-3 years. At the end of the experiment the columns will be opened to analyse secondary phases that should have precipitated inside the fracture. The results will be of direct relevance for the in-situ field experiment because a) it enables testing the analytical methods to detect the expected alterations of both source and fracture; and b) it provides high resolution and high quality data for well controlled laboratory conditions.
Other laboratory experiments
If possible, other laboratory experiments could be performed in LCS Phase 2 to investigate specific issues of high-pH impacts in PA. Potentially, issues like:
- Investigation of CASH phases
- Radionuclide migration under high-pH conditions
- The role of colloids under high-pH conditions
- Cement-bentonite interaction
- could be investigated.
The large number of tasks under the modelling module, and the requirement to integrate modelling with other modules throughout the duration of the LCS, indicate that detailed co-ordination will be necessary.
Evaluation and compilation of a TDB for high-pH solids
Several organisations have developed their own – mostly project-specific – sets of thermodynamic and kinetic data for secondary alteration phases relevant under hyperalkaline and alkaline conditions. The objective here was to compare these existing efforts, so producing critically reviewed, state-of-the-art documents with relevance to PA.
In LCS Phase 1 the following the databases were reviewed and documented:
- TDB comparison for aqueous species (NAB 08-xx)
- TDB comparison for clays (in preparation)
- TDB comparison of CSH phases (Posiva working report 2007-88)
- TDB collection for Zeolithes (JAEA internal document)
During LCS Phase 1 the main objective was to compare existing databases
Predictive capability of codes
The task should depart from the already high level of system understanding achieved within HPF and integrate the results of the structural and geochemical analysis of the overcored material. This is seen as a demonstration of confidence into using a “realistic” modelling approach.
For this reason a modelling group was established in LCS Phase 1, which mainly involves groups from JAEA, Posiva (with CSIC, Spain) and NDA (with Quintessa UK). In order to establish the group and to define the capability of the different codes used by the three groups, two benchmark experiments were defined by the partners (see below).
The availability of appropriate (non-trivial, realistic) benchmarks for testing both numerical modelling codes and approaches is a recurring issue. For this reason, two benchmarks were defined to test the different models applied to provide a first comparison between the different models in use. One benchmark focuses on the hydration process (Lothenbach & Winnefeld, 2006), and the second one aims at reproducing the data set from the HPF core experiment (Soler & Mäder, 2007). The main reasons of choosing these two benchmarks were their high level of knowledge (i.e. parameters measured) and the fact that the experiments were already modelled by other groups (Fig. 3).
Figure 3: a) Measurements (dots) and numerical modelling (curves) of the OPC hydration experiment (Lothenbach & Winnefeld, 2006). b) Measurements (dots) and numerical modelling (curves) of the HPF core infiltration experiment (calculations by Benoît Paris, ITASCA Consulting, see Pfingsten et al. 2005).
Development of a geochemical model for low-alkali cementitious products (LACP) and alteration (short-term)
A model for low-alkaline cementitious products (e.g. cements ‘diluted’ with silica fume and fly ash) will be required if “low-pH” cements are investigated within the LCS project. The objective would be to develop a geochemical module that can be used by batch codes or reactive transport codes to treat cement alteration and cement-water-host rock interaction.
LACP are a type of blended cements whereby all or most of the portlandite component is substituted by other materials with latent hydraulic properties. The early high-alkaline phase (pH>13) is responsible for fast initial reaction rates for setting the cement (essentially due to a fast release from silica to form C-S-H and other hydration products). Substitute materials include blast furnace slags (BSF, a by-product of steel production) and fly ashes (FA, coal firing plants). Both BSF and FA are glassy materials with some additional mineral components. Additionally, it is sometimes necessary to add a silica source, such as silica fume (SF), because of the problem of early reactivity (to develop early strength). In contrast to OPC, the LACP system commonly develops C-A-S-H gels rather than C-S-H.
The two main issues for a geochemical model for LACP are therefore the treatment of a glass component, and the incorporation of a model for C-A-S-H (SF is a minor issue). This requires the expansion of existing databases and some estimate of the hydration equilibrium constant for glass. There are some generalised models for the hydration enthalpy that can be used as a first estimate (from radionuclide waste glass literature). A model for C-A-S-H requires more effort and fitting to some limited literature data on the hydration of LACP systems (e.g. C-A-S-H composition, porewater composition, etc.).
At the end of LCS Phase 2 all information from field tests, laboratory work and natural analogue studies will be amalgamated to close the circle of evidence (Figure 1) for predicting the impact of cementitious materials on the repository near field.
Figure 1. Closing the circle of evidence (after Alexander et al., 1998).
Long-Term Cement Studies (LCS) Experiment