Statistical Approach to Quality Control of Large Thermodynamic Databases

Henrik Nyman, Tarja Höglund, A Roine, Mikko Hupa, Jukka Corander

    Research output: Contribution to journalArticleScientificpeer-review

    8 Citations (Scopus)


    In chemistry and engineering, thermodynamic databases are widely used to obtain the basic properties of pure substances or mixtures. Large and reliable databases are the basis of all thermodynamic modeling of complex chemical processes or systems. However, the effort needed in the establishment, maintenance, and management of a database increases exponentially along with the size and scope of the database. Therefore, we developed a statistical modeling approach to assist an expert in the evaluation and management process, which can pinpoint various types of erroneous records in a database. We have applied this method to investigate the enthalpy, entropy, and heat capacity characteristics in a large commercial database for approximately 25,000 chemical species. Our highly successful results show that a statistical approach is a valuable tool (1) for the management of such databases and (2) to create enthalpy, entropy and heat capacity estimates for such species in which thermochemical data are not available.
    Original languageUndefined/Unknown
    Pages (from-to)1113–1118
    Number of pages6
    JournalMetallurgical and Materials Transactions B: Process Metallurgy and Materials Processing Science
    Issue number5
    Publication statusPublished - 2012
    MoE publication typeA1 Journal article-refereed


    • Thermodynamics
    • Quality control
    • methods: statistical

    Cite this