The effect of mineral precipitation on the diffusivity of porous media: microfluidic experiments and modeling

The precipitation of minerals and its impact on solute transport through porous media play a pivotal role in the long-term performance of a multitude of engineered subsurface systems, including carbon dioxide and hydrogen storage, geothermal reservoirs, and the disposal of high-level radioactive waste. In the context of nuclear waste management, mineral precipitation reactions are promoted at geochemically perturbed barrier interfaces, reducing the porosity, altering the pore space geometry and potentially clogging main transport pathways of solutes. Consequently, the macroscopic (continuum scale) transport properties of porous media can be significantly altered, particularly in the scenario of pore clogging. While precipitation-induced clogging may appear to be beneficial in terms of potentially retarding and/or immobilizing radionuclides through the formation of solid solutions, it can also be detrimental in the event of a gas pressure build-up due to corrosion of steel canisters. Reactive transport modeling (RTM) is indispensable for predicting the long-term behavior of subsurface systems. It integrates geochemical reactions with transport processes across various time and length scales. However, recent continuum-scale models parameterizing the effective diffusivity (De) as function of porosity (ϕ) based on classical Archie’s law appear to be inadequate for describing the porosity-diffusivity relationship in response to mineral precipitation reactions. For the specific case of clogging, significant discrepancies between the modelled results and experimental observations exist, suggesting non-negligible effects of sub-continuum processes, including nucleation mechanisms, crystal morphologies, microporosity of precipitates or yet-to-be-identified processes. However, the implementation of precipitation induced clogging challenges numerical modeling approaches. Therefore, a comprehensive understanding of mineral precipitation within porous media under varying reactive transport conditions, along with its influence on continuum-scale parameters, is critically necessary. The present thesis investigated the effect of mineral precipitation reactions in porous media and its impact on the porosity-diffusivity relationship, using microfluidics in combination with 2D pore-scale modeling. The microfluidic setups provided real-time and space-resolved (operando) observations of celestine (SrSO4) precipitation over a wide range of Damköhler (Da) numbers, capturing the pore space geometry evolution by time-lapse optical imaging. The initial flow fields and concentration distributions were modeled for each microfluidic reactor design to assess the initial transport conditions at steady state, utilizing the COMSOL Multiphysics software. An image segmentation workflow was developed to provide 2D data sets for pore-scale modeling. For the computationally efficient derivation of the effective diffusivity (De), two modeling approaches (TransLBM and TauFactor) were tested. The first case study was dedicated to the effect of the Da number on the porosity-diffusivity relationship, assessing characteristic precipitation pattern formation in a realistic pore geometry by a parallel injection of two reacting fluids, initiating the precipitation by diffusive mixing of reactants across the interface. Two Da regimes were identified based on the formation of precipitation patterns, the applicability of Archie’s law, and the percentage reduction in the effective diffusivity. In the high Da regime (> 10), mineral precipitation occurred non-uniform with localized pore clogging, which reduced the effective diffusivity across the precipitation zone by 53% and resulted in the inapplicability of the continuum-scale description by Archie’s law. In the system with the highest degree of reactivity (Da), the precipitation of a fibrous phase, interpreted as an intermediate metastable precursor phase of celestine (SrSO4 · 1/2H2O), was observed which affects the porosity-diffusivity relationship and thus, confirming previously suggested effects of crystal morphology on the continuum-scale equations. In the low Da regime (< 1), the precipitation of celestine exhibited a more uniform pattern without clogging pore throats. The decrease in the Da number by one order of magnitude resulted in an effective diffusivity reduction of 25 %. In the case of Da = 0.8, the porosity-diffusivity evolution exhibited only minor deviations from Archie’s law whereas for a Da of 0.45, Archie’s law became applicable, indicating a maximum threshold in the reactivity of the system and the applicability of Archie’s law. The experimental identification of two Da regimes was found to align with the applicability ranges of continuum-scale equations defined based on reactive transport modelling studies. Moreover, the pore-scale experiments confirmed the characteristic pattern formation across different Da regimes and, additionally proposed a potential influence of intermediate nucleation mechanisms. The second case study focused on porosity clogging, employing a micronized counter diffusion experimental setup to provide a purely diffusive transport regime (Pe < 1). In combination with confocal Raman spectroscopy, isotopic tracer experiments visualized the diffusive transport of molecular water through an evolving microporosity of the clogging mineral phases operando, allowing an estimation of the critical effective diffusivity (Dceff). The estimated values predict a significant decrease in the effective diffusivity by two to three orders of magnitude by a total porosity reduction of 25%. However, post-clogging dissolution-precipitation reactions demonstrated the non-final state of clogging, confirming the potential evolution of a microporosity of precipitates and might explain previously observed non-zero fluxes and their increase over time. The evolution of the porosity-diffusivity relationship in response to precipitation reactions showed a behavior deviating from Archie's law in all cases, except for transverse mixing induced precipitation reactions in the low Da regime. Considering an additional pre-exponent improved the description of the porosity-diffusivity relationship. In the purely diffusive – low Da regime, the introduction of a critical effective diffusivity (Dceff) and porosity (ϕc) taking into account the clogging state refined the description but still neglected post-clogging dissolution-precipitation reactions and potential further changes in the critical effective diffusivity. The employment of innovative microfluidic experiments, in conjunction with optical and spectroscopic analysis methods, facilitates the identification and quantification of coupled reaction and transport processes in reactive porous media at the microscopic level. This approach also enables the evaluation of so-called critical parameters. The identification of key processes and the quantification of individual parameter values permit the validation of empirical equations and their applicability for reactive transport modelling on the continuum scale. Moreover, the experimentally obtained data can be utilized for future sensitivity analysis, and to evaluate uncertainties in model predictions.