Requirements for Thermodynamic Property Formulations

Abstract

The importance of accurate experimental data in the development of a thermodynamic property correlation is well established. Systematic errors in experimental data sets affect the quality of the correlation that is based on such data. To minimize the effects of such errors, care is taken to analyze each data set used in a correlation for consistency with independent measurements and with data forms in the same and adjacent regions of the thermodynamic surface. It is best to avoid using experimental data that contain large systematic errors in the fitting process. Large random errors may be tolerated when no alternative data sets are available, but the use of such data increases the uncertainty of the formulation. If data are not available for a particular region of the thermodynamic surface, it is often helpful to use predicted values in the fitting process. Such predictions can be verified later or discarded when experimental data become available.