Development of pure component property models for chemical product-process design and analysis

Amol Shivajirao Hukkerikar

    Research output: Book/ReportPh.D. thesis

    1600 Downloads (Pure)


    Property prediction models based on the group-contribution+ (GC+) approach have been developed to provide reliable predictions of pure component properties together with uncertainties of predicted property values which is much needed information in performing chemical product and process design and analysis of sustainable chemical processes. For developing property models, a systematic methodology for property modeling and uncertainty analysis is employed. The methodology includes a parameter estimation step to determine parameters of the property model and an uncertainty analysis step to establish
    statistical information about the quality of parameter estimation, such as the parameter covariance, the standard errors in predicted properties, and the confidence intervals. For parameter estimation, large data sets of experimentally measured property values of a wide range of pure components taken from the CAPEC database, the US Environmental Protection Agency (EPA) database, and the USEtox database are used. In total, 21 thermo-physical properties and 22 environmental-related properties of pure components which include normal boiling point, critical constants, standard enthalpy of formation, liquid viscosity, fathead minnow 96-h LC50, oral rat LD50, global warming potential, emission to urban air (carcinogenic and noncarcinogenic) among others are modeled and analyzed. For all the estimated pure component properties, the corresponding 95% confidence intervals are also reported thereby providing information on the degree of accuracy of the property estimates. In addition, a method based on the ‘molecular structural similarity criteria’ is developed so
    that efficient use of knowledge of properties could be made in the development/improvement of property models. This method, in principle, can be applied to a wide range of properties of pure components. pure components. In this work, however, the application of ‘molecular structural similarity criteria’ is illustrated by considering performance improvement of models for enthalpy of formation, enthalpy of fusion, and critical temperature. For all properties listed above, it has been possible to achieve significant improvements in the performance of their models. The improved GC model for enthalpy of formation yields an average absolute deviation of 1.75 kJ/mol, which is well within the required chemical accuracy. Important issues related to property modeling such as: (i) quantity of property data used for the parameter regression; (ii) selection of the most appropriate form of the property model function; and (iii) the accuracy
    and thermodynamic consistency of predicted property values are also discussed. The developed models have been implemented into ProPred®, a property estimation toolbox of Integrated Computer Aided System, ICAS®, developed at CAPEC, DTU. Finally, a methodology for performing sensitivity analysis of process design due to uncertainties of property estimates is presented. This methodology allows the user to evaluate the effects of uncertainties of property estimates on the final design; list and rank properties that are most important from process design point of view; and establish acceptable levels of accuracy for property models. The application of this methodology is highlighted through three case studies namely, design of an extractive distillation process, design of a short-path evaporator, and design of a de-acidification system of vegetable oil deodorization process.
    Original languageEnglish
    Place of PublicationKgs. Lyngby
    PublisherTechnical University of Denmark, Department of Chemical and Biochemical Engineering
    Number of pages289
    ISBN (Print)978-87-93054-11-0
    Publication statusPublished - 2013


    Dive into the research topics of 'Development of pure component property models for chemical product-process design and analysis'. Together they form a unique fingerprint.

    Cite this