Group-contribution+ (GC+) based estimation of properties of pure components: Improved property estimation and uncertainty analysis

Amol Hukkerikar, Bent Sarup, Antoon Ten Kate, Jens Abildskov, Gürkan Sin, Rafiqul Gani

Research output: Contribution to journalJournal articleResearchpeer-review

Abstract

The aim of this work is to present revised and improved model parameters for group-contribution+ (GC+) models (combined group-contribution (GC) method and atom connectivity index (CI) method) employed for the estimation of pure component properties, together with covariance matrices to quantify uncertainties in the estimated property values. For this purpose, a systematic methodology for property modeling and uncertainty analysis of GC models and CI models using maximum-likelihood estimation theory is developed. For parameter estimation, large data-sets of experimentally measured property values of pure components of various classes (hydrocarbons, oxygenated components, nitrogenated components, poly-functional components, etc.) taken from the CAPEC database are used. In total 18 pure component properties are analyzed, namely normal boiling point, critical temperature, critical pressure, critical volume, normal melting point, standard Gibbs energy of formation, standard enthalpy of formation, normal enthalpy of fusion, enthalpy of vaporization at 298 K, enthalpy of vaporization at the normal boiling point, entropy of vaporization at the normal boiling point, flash point, auto ignition temperature, Hansen solubility parameters, Hildebrand solubility parameter, octanol/water partition coefficient, acentric factor, and liquid molar volume at 298 K. Important issues related to property modeling such as reliability and predictive capability of the property prediction models, and thermodynamic consistency of the predicted properties (such as, relation of normal boiling point versus critical temperature) are also analyzed and discussed. The developed methodology is simple, yet sound and effective and provides not only the estimated pure component property values but also the uncertainties (e.g. prediction errors in terms of 95% confidence intervals) in the estimated property values. This feature allows one to evaluate the effects of these uncertainties on product-process design, simulation and optimization calculations, contributing to better-informed and more reliable engineering solutions. (C) 2012 Elsevier B.V. All rights reserved.

Original languageEnglish
JournalFluid Phase Equilibria
Volume321
Pages (from-to)25-43
ISSN0378-3812
DOIs
Publication statusPublished - 2012

Keywords

  • Pure component properties
  • Group-contribution method
  • Atom connectivity index method
  • Uncertainty analysis
  • Maximum-likelihood estimation theory

Cite this

@article{f3e9ebb843ba4c5fbcec8f7c509209ed,
title = "Group-contribution+ (GC+) based estimation of properties of pure components: Improved property estimation and uncertainty analysis",
abstract = "The aim of this work is to present revised and improved model parameters for group-contribution+ (GC+) models (combined group-contribution (GC) method and atom connectivity index (CI) method) employed for the estimation of pure component properties, together with covariance matrices to quantify uncertainties in the estimated property values. For this purpose, a systematic methodology for property modeling and uncertainty analysis of GC models and CI models using maximum-likelihood estimation theory is developed. For parameter estimation, large data-sets of experimentally measured property values of pure components of various classes (hydrocarbons, oxygenated components, nitrogenated components, poly-functional components, etc.) taken from the CAPEC database are used. In total 18 pure component properties are analyzed, namely normal boiling point, critical temperature, critical pressure, critical volume, normal melting point, standard Gibbs energy of formation, standard enthalpy of formation, normal enthalpy of fusion, enthalpy of vaporization at 298 K, enthalpy of vaporization at the normal boiling point, entropy of vaporization at the normal boiling point, flash point, auto ignition temperature, Hansen solubility parameters, Hildebrand solubility parameter, octanol/water partition coefficient, acentric factor, and liquid molar volume at 298 K. Important issues related to property modeling such as reliability and predictive capability of the property prediction models, and thermodynamic consistency of the predicted properties (such as, relation of normal boiling point versus critical temperature) are also analyzed and discussed. The developed methodology is simple, yet sound and effective and provides not only the estimated pure component property values but also the uncertainties (e.g. prediction errors in terms of 95{\%} confidence intervals) in the estimated property values. This feature allows one to evaluate the effects of these uncertainties on product-process design, simulation and optimization calculations, contributing to better-informed and more reliable engineering solutions. (C) 2012 Elsevier B.V. All rights reserved.",
keywords = "Pure component properties, Group-contribution method, Atom connectivity index method, Uncertainty analysis, Maximum-likelihood estimation theory",
author = "Amol Hukkerikar and Bent Sarup and {Ten Kate}, Antoon and Jens Abildskov and G{\"u}rkan Sin and Rafiqul Gani",
year = "2012",
doi = "10.1016/j.fluid.2012.02.010",
language = "English",
volume = "321",
pages = "25--43",
journal = "Fluid Phase Equilibria",
issn = "0378-3812",
publisher = "Elsevier",

}

Group-contribution+ (GC+) based estimation of properties of pure components: Improved property estimation and uncertainty analysis. / Hukkerikar, Amol; Sarup, Bent; Ten Kate, Antoon; Abildskov, Jens; Sin, Gürkan; Gani, Rafiqul.

In: Fluid Phase Equilibria, Vol. 321, 2012, p. 25-43.

Research output: Contribution to journalJournal articleResearchpeer-review

TY - JOUR

T1 - Group-contribution+ (GC+) based estimation of properties of pure components: Improved property estimation and uncertainty analysis

AU - Hukkerikar, Amol

AU - Sarup, Bent

AU - Ten Kate, Antoon

AU - Abildskov, Jens

AU - Sin, Gürkan

AU - Gani, Rafiqul

PY - 2012

Y1 - 2012

N2 - The aim of this work is to present revised and improved model parameters for group-contribution+ (GC+) models (combined group-contribution (GC) method and atom connectivity index (CI) method) employed for the estimation of pure component properties, together with covariance matrices to quantify uncertainties in the estimated property values. For this purpose, a systematic methodology for property modeling and uncertainty analysis of GC models and CI models using maximum-likelihood estimation theory is developed. For parameter estimation, large data-sets of experimentally measured property values of pure components of various classes (hydrocarbons, oxygenated components, nitrogenated components, poly-functional components, etc.) taken from the CAPEC database are used. In total 18 pure component properties are analyzed, namely normal boiling point, critical temperature, critical pressure, critical volume, normal melting point, standard Gibbs energy of formation, standard enthalpy of formation, normal enthalpy of fusion, enthalpy of vaporization at 298 K, enthalpy of vaporization at the normal boiling point, entropy of vaporization at the normal boiling point, flash point, auto ignition temperature, Hansen solubility parameters, Hildebrand solubility parameter, octanol/water partition coefficient, acentric factor, and liquid molar volume at 298 K. Important issues related to property modeling such as reliability and predictive capability of the property prediction models, and thermodynamic consistency of the predicted properties (such as, relation of normal boiling point versus critical temperature) are also analyzed and discussed. The developed methodology is simple, yet sound and effective and provides not only the estimated pure component property values but also the uncertainties (e.g. prediction errors in terms of 95% confidence intervals) in the estimated property values. This feature allows one to evaluate the effects of these uncertainties on product-process design, simulation and optimization calculations, contributing to better-informed and more reliable engineering solutions. (C) 2012 Elsevier B.V. All rights reserved.

AB - The aim of this work is to present revised and improved model parameters for group-contribution+ (GC+) models (combined group-contribution (GC) method and atom connectivity index (CI) method) employed for the estimation of pure component properties, together with covariance matrices to quantify uncertainties in the estimated property values. For this purpose, a systematic methodology for property modeling and uncertainty analysis of GC models and CI models using maximum-likelihood estimation theory is developed. For parameter estimation, large data-sets of experimentally measured property values of pure components of various classes (hydrocarbons, oxygenated components, nitrogenated components, poly-functional components, etc.) taken from the CAPEC database are used. In total 18 pure component properties are analyzed, namely normal boiling point, critical temperature, critical pressure, critical volume, normal melting point, standard Gibbs energy of formation, standard enthalpy of formation, normal enthalpy of fusion, enthalpy of vaporization at 298 K, enthalpy of vaporization at the normal boiling point, entropy of vaporization at the normal boiling point, flash point, auto ignition temperature, Hansen solubility parameters, Hildebrand solubility parameter, octanol/water partition coefficient, acentric factor, and liquid molar volume at 298 K. Important issues related to property modeling such as reliability and predictive capability of the property prediction models, and thermodynamic consistency of the predicted properties (such as, relation of normal boiling point versus critical temperature) are also analyzed and discussed. The developed methodology is simple, yet sound and effective and provides not only the estimated pure component property values but also the uncertainties (e.g. prediction errors in terms of 95% confidence intervals) in the estimated property values. This feature allows one to evaluate the effects of these uncertainties on product-process design, simulation and optimization calculations, contributing to better-informed and more reliable engineering solutions. (C) 2012 Elsevier B.V. All rights reserved.

KW - Pure component properties

KW - Group-contribution method

KW - Atom connectivity index method

KW - Uncertainty analysis

KW - Maximum-likelihood estimation theory

U2 - 10.1016/j.fluid.2012.02.010

DO - 10.1016/j.fluid.2012.02.010

M3 - Journal article

VL - 321

SP - 25

EP - 43

JO - Fluid Phase Equilibria

JF - Fluid Phase Equilibria

SN - 0378-3812

ER -