Estimation of Vapor Pressures and Enthalpies of Vaporization of Biodiesel-Related Fatty Acid Alkyl Esters. Part 2. New Parameters for Classic Vapor Pressure Correlations

Abstract

In Part I of this series of articles, predictive models capable of estimating vapor pressures and enthalpies of vaporization of biodiesel-related esters have been reviewed. In this study, we propose component-specific parameters for classic vapor pressure correlations. First, the previously collected database was evaluated and used in the regression of Antoine coefficients for all compounds. Then, a constrained regression methodology was employed to generate extrapolatable Wagner parameters for the esters whose experimental critical properties were available. Finally, expressions derived from the Clausius–Clapeyron equation have been investigated for the estimation of enthalpies of vaporization. A comparative analysis of the results produced by our parameters and by others available in the literature encourages the use of our values in future applications.

Wagner Equation Predicting Entire Curve for Pure Fluids from Limited VLE Data: Error Dependency Upon Data Interval & Fully-Determined Case

“…Antoine analytics have been used for the parameterization of Wagner constants previously, but the verification of the predictive capability was relative to the data used for the original Antoine parameterization [10], or the verification utilized data only slightly below that of the parameterization interval -Ref. [20], for example, verified down to the minimum reduced temperature of 0.5 in some cases. As can be seen from Tables 7 and 10 here, extension upward to the critical point (Tr > 0.6) has significantly less error because it involves pseudo extrapolationalthough prediction is extended beyond the data interval's upper limit, there is an element of interpolation because the critical point is used in the reduced formulation, and thus, is an upper bound anchor point.…”

“…The authors of Ref. [20] pruned their data set of 839 VLE points down to 348, a reduction of 58.5%, when performing their validation of experimental data for use in their Antoine parameterizations, consistent with the general data deviations we saw [21]. Consequently, one can use limited data to parameterize Wagner constants valid for the entiretwo phase curve provided one is confident that the limited data represents the species' best values, the reduced temperature width of the interval is ≥ 0.1, and the lower bound is ≤ 0.6.…”

Criação De Banco De Dados Para Simulação Da Produção De Biodiesel. Parte 1: Estimativa De Propriedades Termofísicas De Óleos Vegetais

“…A literatura revela que há um grande volume de trabalhos realizados no projeto de processos e fabricação de biocombustíveis a partir de óleos vegetais, mas os dados experimentais disponíveis para as propriedades termodinâmicas de óleos vegetais e/ou modelos de previsão são limitados (EVANGELISTA et al, 2017a;EVANGELISTA et al, 2017b;GARCIA et al, 2013;NIKITIN e POPOV, 2014;LI et al, 2016;EVANGELISTA et al, 2018) A escassez de dados das propriedades termofísicas dos componentes do biodiesel e óleos vegetais pode estar relacionada à dificuldade em obtê-los de forma pura e na realização dos experimentos, pois essas substâncias podem se degradar em condições severas (WALLEK et al, 2013;POLING et al, 2001). Sabe-se, no entanto, que os óleos vegetais não são compostos apenas por um triacilglicerol.…”

Estimation of Physical Constants of Biodiesel-Related Fatty Acid Alkyl Esters: Normal Boiling Point, Critical Temperature, Critical Pressure, and Acentric Factor

“…Despite the efforts put in that work, especially to ensure its completeness in terms of database and investigated models, we have recently realized that some of its conclusions must be revised for the following reasons: (1) A few T b values should not have been used, either because they were not obtained at pressures close to 760 mmHg or because they are not sufficiently reliable. (2) After the recent publications by Nikitin and Popov − and by us, enough P c , T c , and ω data are available to perform a direct evaluation of the available models.…”