A new model for aqueous electrolyte solutions near the critical point of water incorporating aqueous reaction equilibria

Aqueous electrolyte solutions at temperature and pressure conditions near the critical point of water are difficult to describe using traditional equations of state based upon the excess Gibbs energy. Models based upon the residual Helmholtz energy have proven more effective. Anderko and Pitzer1 developed a residual Helmholtz energy model (AP) for aqueous electrolyte solutions in which the electrolyte is assumed to be fully associated. The model has been effectively used in describing densities and vapor-liquid equilibria for simple electrolyte systems. The model is less effective for describing enthalpic properties such as heats of dilution.;Oscarson and coworkers2,3 modified the AP model for NaC1 solutions by adding a term accounting for the change in Helmholtz energy as a result of aqueous dissociation reactions. This new model, called the RI model, is more accurate than the AP model at conditions where the NaC1 dissociates more fully into ions.;Liu et. al4,5 modified the RI model by adding a term to describe interactions between ions in solution and by regressing new model parameter values. This new model, called the RIII model, is more accurate than both the AP model and the RI model and may be used to predict species concentrations in solution as a result of aqueous phase reactions. The RIII model has substantial thermodynamic inconsistencies, however, and is poorly suited for describing mixed solute solutions.;This dissertation presents the RIV model which is an electrolyte solution model for solutions in the ranges of 350 °C to 400 °C and 18 MPa to 40 MPa. The RIV model has been applied to aqueous NaCl solutions and aqueous LiC1 solutions. The RIV model is a modification of the AP model and includes aqueous phase reactions implicitly through fundamental species interactions. The RIV model is thermodynamically consistent. It is capable of describing densities and heats of dilution. Density predictions from the RIV model are less accurate than the AP model predictions (6.66 % error vs. 3.51 % error) but are reasonable. The heats of dilution predictions from the RIV model are much more accurate than those from the AP model (25.16 % error vs. 78.78 % error). Predictions of the ionic species concentration from the RIV model are likely to be poor as indicated by the poor agreement between experimental values and calculated values of equilibrium constants valid at infinite dilution.;In order to provide the necessary data from which to regress the parameters of the RIV model, experimental heat of dilution values were determined using flow calorimetry techniques. These values are also reported in this dissertation.