Huet – Modeling the Effect of C-S-H Incongruent Dissolution on Cement Reactivity in CO2 Rich Environment.
Bruno Huet (bhuet@slb.com)
Slumberger Cargon Services
1, rue Henri Becquerel, BP202
92142 Clamart Cedex, France
Modeling cement reactivity in aggressive environment with reactive transport codes requires 3 main types of physical and chemical parameters: i) the total composition of your system, ii) equilibrium constant for all species and iii) effective transport properties. The purpose of this work is to evaluate what level of detail should be used to describe C-S-H incongruent dissolution in order to quantitatively simulate cement evolution in CO2 rich environment. C-S-H equilibrium is described by a discrete solid solution model: a limited number of CS-H (two to eight) with fixed Ca to Si ratio is introduced. Two methods are used: i) reaction path simulation, for which the system composition is changed by species addition for example an addition of silica or CO2, and ii) reactive transport simulations, for which the time and spatial evolution of the initial system composition is driven by diffusion or advection. Different properties could be influenced by the number of C-S-H: solid phase composition, aqueous phase composition, stability domain of secondary phases (AFm, AFt or hydrogarnets), porosity and transport properties. Reaction path simulations indicate that a continuous change in system total composition, leads to a continuous decrease of Ca to Si ratio in C-S-H but a discontinuous evolution of aqueous phase composition. These discontinuities directly relate to the number of C-S-H of fixed composition. Changing the number of C-S-H also modifies the composition range for which secondary phases are stable. However, reactive transport simulations indicate that the overall reactivity of cement, i.e. the dynamics of the zone chemically modified, is not meaningfully influenced by the thermodynamic properties of the C-S-H solid solution. Indeed aqueous phase composition profile and transport properties profile remain the same, which means that the rate limiting step remains the same (leaching of calcium or ingress of CO2 through reacted layer). On the opposite, mineral zoning, for example the localization of AFm and AFt precipitation/dissolution front, and the profile of Ca to Si ratio in C-S-H is modified.
Based on these specific results, it seems that thermodynamic properties of C-S-H (solubility, density) are well constrained and are not key parameters for modeling cement reactivity. On the opposite, effective diffusivities of C-S-H and of decalcified C-S-H pastes, which directly depend on the microstructure of C-S-H phase, are key properties for evaluating cement reactivity. The latter properties are also the less well constrained by experiments.