In previous studies we investigated how dissolved gases were transported through hardened cement paste and how this was influenced by degradation processes.
The current project focus is on the investigation how free gas becomes transported in a cementitious material starting from HCP but extending to mortar (more relevant for real used cementitious materials) where mesoscale heterogeneities such as aggregates and their corresponding interfacial transition zone are present (another PhD proposal focusses on the effect of combined leaching and carbonation on the physical and transport properties of mortars).
In order to support argumentation for safe geological disposal of radioactive waste, a mechanistic understanding of the behaviour of gas in cementitious materials as part of the EBS (e.g. monoliths, plugs, seals) or auxiliary repository components (e.g. liner, backfill) is needed.
This is translated into following questions: 1) can we visualise gas flow through cement paste/mortar and can we derive consistent two-phase flow parameters for predictive modelling? 2) what is the role of heterogeneities and ITZ on the gas transport mechanism?
In order to reach the objectives we put forward following strategy:
Sample preparation and characterisation
- Production of HCP and mortar relevant for backfill related formulation (OPC based, high porosity). (Know-how available @SCK)
- Classical characterization of the samples in terms of microstructure and composition: N2 adsorption, MIP, XRD, SEM-EDX (Know-how & techniques available @SCK - except MIP)
- Determination of the water retention curve to extract 2-phase flow parameters, gas permeability (intrinsic permeability). (Techniques to be developed @SCK, based on well-established methodologies)
- Determination of the water saturated transport parameters: hydraulic conductivity, (dissolved) gas diffusivity. (Know-how & techniques available @SCK)
- Availability of gas pressure systems (Hg based pressure systems, syringe pumps) for performing well controlled gas injection experiments. Possibilities to perform tests under isostatic conditions.
Investigation of gas flow in HCP&Mortar
A large focus in the project will be on developing/optimising methodologies to visualise the desaturation process of cementitious materials. Although different well established techniques exist for visualisation of pores, flow&transport, this will remain a challenge as the porous structure of cementitious materials extends from the micropore region (<2 nm) up to the macropore region (>50 nm).
- µCT based techniques under guidance of promotor Prof. V. Cnudde (UGCT, Centre for X-ray tomography of Ghent University). µCT is a powerful technique to visualise 3D pore structure down to (sub)µm level (Cnudde et al., 2015). It is used to study 2-phase flow in heterogeneous rocks (Bultreys et al., 2015). 3D images form the basis for modelling transport at the pore level (pore network models, Lattice-Boltzmann (LB) models – see further). These models are then used to simulate capillary pressure curves, (relative) permeabilities. These outcomes can then be tested against experimental data (water retention curve, permeability & diffusivity data). The main difficulty is related to the sample size/resolution trade off in the imaging. Pores can only be resolved to a certain threshold, (sub)µm, hence the small meso- and micropores are not accounted for
- PET based techniques. HZDR (Helmholtz-Zentrum Dresden-Rosendorf) developed the so-called GEOPET (Lippmann-pipke et al., 2011) technique in which a tomographic technique is used to do spatio-temporal visualisation of transport processes of suitable radiotracers in porous media. This technique is often used in conjunction with CT where the pore structure is visualised. Combination of both provides images of where tracers are moving. Cementitious materials may be first saturated with (non-sorbing) radiotracers that are dissolved in the water. Upon applying gas pressure, the movement of the water containing the radiotracer can be visualised). Here we may benefit from a first collaboration within the EC CEBAMA project.
- NMR techniques (Leo Pel and Henk Huinink, 2009) may also be of interest. It allows to distinguish between free & bound water and have the potential to visualise flow and moisture transport. NMR has been used for these type of studies on cementitious materials and other porous materials.
- SEM based techniques. µCT is limited in resolving the pore structure at subµm level. Hereto, SEM based techniques, notably FIB-SEM, may complement in the 3D characterisation of the pore structure (Holzer et al., 2006). Also within the EC CEBAMA project, there will be an intense collaboration with BRGM for microstructural characterisation using FIB-nanotomography of cementitious materials as basis for LB based µstructural modelling.
Combinations of different techniques is needed as they provide complementary information.
Integration through modelling
The experimental data and observations will be used to build, parametrise and test 2-phase flow models at multiple scales.
Here, different types of modelling approaches may be explored:
- Pore-scale models based on lattice-Boltzmann's formalism (Patel et al., 2014). With this approach, a detailed morphology of hardened cement paste such as portlandite, hydrates, capillary pores, etc. can be explicitly included. Thus the gas transport pathways can be explicitly captured for given cement paste microstructure. This step provides fundamental answers such as whether desaturation takes place mainly through the capillary pores or also through gel pores for various gas pressure conditions. To achieve this, the existing pore scale for single phase transport will be extended to include two-phase (gas and moisture) transport formulation. In this step, effective gas diffusivity as well as effective unsaturated hydraulic properties such as moisture retention and hydraulic conductivity for the moisture phase will be derived. The same two-phase model will then be used to predict the effective properties of mortar. The mortar scale is composed of hardened cement paste, fine aggregates and interfacial transition zone (ITZ). At this scale, the hardened cement paste is treated as a homogeneous medium with the effective properties determined from the microscale modelling. The same approach holds with ITZ as this is also hardened cement paste but of relatively higher porosity compared to the bulk cement paste. Aggregates can be treated as non-reactive. Thus the effective properties of both hardened cement paste and mortar can be compared with experimentally determined properties.
- Pore network models (under guidande of co-promotor A. Raoof): In a pore network modelling (Raoof et al., 2012, 2013), 3D images are used to extract a simplified network of pores and throats. The simplification of pore space geometry and fluid mechanics is compensated by the method's high computational efficiency. However, when pore sizes span several orders of magnitude this poses problems for computation. Bultreys et al. (Team of V. Cnudde Ugent), treats the sub-resolution porosity as a continuous porous medium. A 2-scale network is built which incorporates microporosity information without taking every individual micropore into account. Starting from µCT scan segmented into 3 regions (solid, pore, micropore-region) a network is extracted. Pores that touch the same microporous cluster are called micro-connections. Continuum petrophysical properties are assigned to these clusters and connections. These properties are obtained from N2-adsorption, MIP and FIB-SEM.
Continuum based models: With this approach, gas/moisture transport behaviour in a homogeneous mortar can be predicted based on effective properties obtained from the pore scale or pore network models. Note that pore-scale, pore-network, finite element or any other numerical approaches can be used to solve continuum scale problems. This step also provides an alternative approach to compare the two-phase flow behaviour observed in experiments by directly using experimentally derived properties as opposed to numerically estimated effective properties.
- Phung Q. (2015) Effects of Carbonation and Calcium Leaching on Microstructure and Transport Properties of Cement Pastes.- Ghent, Belgium: Ghent University, 2015.- 261 p.- PhD thesis.- ISBN 978-90-8578-783-9
- Patel R., Perko J., Jacques D., De Schutter G., Van Breugel K., Ye G. (2014) A versatile pore-scale multicomponent reactive transport approach based on lattice Boltzmann method: Application to portlandite dissolution: 3rd International Workshop Mechanisms and modelling of waste/cement interactions, Ghent, Belgium, 6-8 May 2013.- In: Physics and Chemistry of the Earth, 70-71, p. 127-137.- ISSN 1474-7065
- Patel R., Perko J., Jacques D., De Schutter G., Ye G., Van Breugel K. (2013) Lattice Boltzmann based multicomponent reactive transsport model coupled with geochemical solver for pore scale simulations.- In: Computational Methods for Coupled Problems in Science and Engineering V, Santa Eulària, Ibiza, Spain, 17 June - 19 January 2013 / CIMNE, Barcelona, Spain, CIMNE, 2013, p. 806-817.
- Cnudde V. et al. (2015) Looking at pore scale processes in geomaterials using time-resolved 3D imaging and multi-scale imaging. 2nd UGCT seminar 2015. https://biblio.ugent.be/publication/6929562/file/6929592.pdf
- Bultreys et al. (2015) Multi-scale, micro-computed tomography-based pore network models to simulate drainage in heterogeneous rocks. Adv Water Resour 78, 36– 49
- Bultreys et al. (2015) A multi-scale, image-based pore network modelling approach to simulate 2-phase flow in heterogeneous rocks, SCA2015-027 https://biblio.ugent.be/publication/6923699/file/6923722.pdf
- Bultreys et al. (2015) Multi-scale, image-based pore network models to simulate two-phase flow in heterogeneous rocks. Porous Media, 7th International conference, Abstracts&Poster https://biblio.ugent.be/publication/5972195
- Leo Pel and Henk Huinink (2009) NMR imaging of moisture and ion transport in building materials. Magnetic Resonance Microscopy, edited by Sarah. L. Codd and Joseph D. Seymour, Wiley-VCH, Weinheim, Germany, 451-463.
- Lippmann-Pipke et al. (2011) Matching 4D porous media fluid flow GeoPET data with COMSOL multiphysics simulation results. https://www.comsol.co.in/paper/matching-4d-porous-media-fluid-flow-geopet-data-with-comsol-multiphysics-simulat-11127
- Raoof, A., and S. M. Hassanizadeh (2012), A new formulation for pore-network modeling of two-phase flow, Water Resour.Res., 48, W01514, doi:10.1029/2010WR010180. http://www.geo.uu.nl/hydrogeology/raoof/Papers_Amir_web/4_Raoof2012TwoPhase.pdf
- Raoof A., Nick H.M., Hassanizadeh S.M., Spiers C.J., PoreFlow (2013) A complex pore-network model for simulation of reactive transport in variably saturated porous media, Computers & Geosciences 61, 160-174 http://www.geo.uu.nl/hydrogeology/raoof/Papers_Amir_web/10_Raoof2013_PoreFlow.pdf
- Holzer L., Gasser P., Muench B. (2006) Quantification of capillary pores and hadley grains in cement paste using FIB nanotomography. In: Measuring, monitoring and modeling concrete properties, ed. M.S. Konsta-Gdoutos, Springer, pp509-516