Global sensitivity based estimability analysis for the parameter identification of Pitzer’s thermodynamic model

This paper deals with parameter estimability analysis and identification of the Pitzer model used in prediction thermodynamic properties phase equilibria electrolytic solutions. The method is based on orthogonalization a sensitivity matrix to rank unknown parameters from most estimable least estimable. Although obtained results are interesting, algorithm shows its limits since it local sensitivities outputs respect may significantly affect reliability model. In this work, improved by computing means global analysis. For demonstration purposes, different sets experimental measurements sulfuric acid solutions carried out at temperatures concentrations. They mainly consist pH, density, conductivity molality measurements. then applied each sets. determined identified using branch-and-reduce optimization their accuracy assessed confidence intervals. Finally, quality quantified Pearson product-moment coefficient, high values show very good agreement between predictions