Geochemical and reactive transport modelling are essential tools in hydrogeology. They help to identify and assess geochemical processes occurring in applications such as groundwater contamination, water-rock interactions and geologic carbon sequestration. In this thesis we present methods for mixing and speciation calculations to be used for both interpretation of hydrochemical data and numerical modelling. The first method presented allows solving geochemical speciation using redundant information, while acknowledging errors in data. Traditional speciation algorithms use a fixed number of data and equilibrium assumptions to calculate the concentration of the species present in a chemical system. We demonstrate that using redundant data (i.e., data or assumptions that exceed the minimum required and therefore are not strictly necessary) can improve speciation results by reducing estimation errors. In fact, we show that speciation errors decrease when increasing the number of redundant data. The second method presented allows calculating mixing proportions of a number of end-members in a water sample from uncertain chemical data. Traditional methods for evaluating mixing ratios require the use of conservative tracers, which severely limits their applicability. The novelty of the method lies on the possibility of imposing equilibrium conditions on the mixture, while acknowledging kinetic reactions, which naturally leads to quantification of reactions. We applied the method to a freshwater-saltwater mixing problem in a set of samples collected by Sanz (2007), where we also characterized carbonate dissolution/precipitation and the production/consumption of CO2. These methods have been implemented in an Object-Oriented library called "CHEPROO++". This library can be used for hydrogeochemical calculations such as mixing waters linked to mass balance programs, which allows extending conservative transport simulators to solve reactive transport. One peculiarity of CHEPROO++ is the possibility of defining components decoupling constant activity species (CAS) such as, for example, pure equilibrium minerals or water (if the solution is sufficiently diluted). CHEPROO++ treats CAS as primary species. Decoupling CAS can be useful for speciation calculations because it allows reducing the system to be solved iteratively. To check if decoupling CAS is advantageous, we applied the speciation algorithm that decouples CAS to a reactive transport application. In particular we used this speciation algorithm for the chemical step of the Sequential Iteration Approach for reactive transport modelling. We compared the proposed algorithm with the traditional method, which does not decouple CAS, on a onedimensional domain where calcite is dissolving in equilibrium. Results show that decoupling CAS can decrease the number of iterations necessary for transport and chemistry calculations in case of equilibrium dissolution.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados