Verification of the General Applicability of a Predictive Model for the Physical Properties of Molten Salts in Heat Energy Storage Applications

To meet the growing demand for molten salt thermal storage technology, this study aims to enhance the accuracy of predicting molten salt properties. This is achieved through the systematic compilation of literature data on individual molten salts (including chlorides, fluorides, hydroxides, sulfates, carbonates, and nitrates) and their binary and ternary mixtures. To this end, we evaluated two existing theoretical models based on molar and mass fractions (Models I and II). We assessed their predictions for key physical parameters, including density, melting point, enthalpy of fusion, specific heat capacity, and surface tension. Quantitative assessment employed mean absolute relative deviation and correlation coefficients. Results demonstrate that Model I exhibits significant advantages in predicting melting points, enthalpy of fusion, specific heat capacity, and surface tension. Through in-depth analysis at the micromechanism level, we elucidate the fundamental causes of performance differences and confirm that Model I possesses outstanding comprehensive predictive capability. This provides a reliable modeling approach for the precise prediction of molten salt physical properties.