The extended UNIQUAC model is a thermodynamic model able to estimate thermodynamic properties of aqueous electrolyte solutions under a wide range of temperature, pressure, and composition conditions. Thermodynamic properties include species activity coefficients, excess molar Gibbs energy, excess molar enthalpy, excess molar heat capacity. These properties are important for aqueous speciation calculations, mineral and gas solubility computations, chemical kinetic modeling of mineral dissolution and precipitation, and in reactive transport simulations considering chemically complex aqueous electrolyte solutions. In this paper we present a brief literature review on the extended UNIQUAC model, we report on its implementation in C++ in the Reaktoro framework for modeling chemically reactive systems, and we show its use from Python for computing mineral and gas solubilities in aqueous solutions at a wide range of temperature, pressure, and salinity conditions. We validated the calculations against experimental data and results obtained through the software ScaleCERE implementing the extended UNIQUAC model. Our conclusion is that the extended UNIQUAC model has been successfully implemented into the Reaktoro framework, thereby providing a suitable activity model for geochemical and reactive transport modeling. • Accurate estimation of mineral and gas solubilities using the extended UNIQUAC model in Reaktoro. • Validation against experimental data and ScaleCERE ensures reliability and accuracy. • Provides detailed formulation and necessary model parameters for the extended UNIQUAC model. • Demonstrates mineral solubility calculations in Reaktoro using straightforward Python scripts.

Deep geological disposal is a widely proposed approach for safe storage of low- and intermediate-level radioactive waste (L/ILW) for tens to hundreds of thousands of years. Although the use of cement materials will maintain a high pH environment that is favorable for radionuclide retention, slows down metal corrosion, and suppresses microbial activity, different gases may still be produced by chemical reactions, such as pH-dependent anoxic corrosion of metals, and the degradation of organic matter as well. Both reactions consume water and may lead to the formation of a free gas phase within and around the repository.In order to investigate the controlling factors of this gas production processes, a coupled reactive transport model of component-based two-phase flow in the OpenGeoSys framework is adopted here. Building on the previous work by Huang et al. (2021), this work extends the geometric configuration of the model from a single drum to the scale of a waste container (12 drums) in order to more realistically model the water and gas fluxes. The geochemical evolution of a container filled with cemented steel drums and surrounded by mortar and host rock is simulated in a two-dimensional setup over 500 years. The relative humidity in the pore space as measure of water availability determines the chemical reactivity.We have conducted a sensitivity study for the influence of mortar capillarity and permeability on gas production over time in the waste drums. As expected, the amount of gas produced within the first several years depends primarily on the initial water content within the waste drum and the amount of fast degradable waste stored in a drum. Later, this feedback system is forced into a new equilibrium, where the water transport has to match the water consumption rate at the waste package. Both, the mortar capillarity and permeability control the point in time when the equilibrium is reached and the subsequent rate of gas production.This study shows how reactive transport models are capable to help with the understanding of couplings between chemical and transport processes that control gas generation in L/ILW waste repositories, especially in barrier systems with different material properties. Future studies could benefit from more accurate parameterization of the chemical reactions, depending on the availability of new experimental data and/or more realistic process-based model approaches.