Experimental measurements and equation of state modeling of liquid densities for long-chain n-alkanes at pressures to 265MPa and temperatures to 523K (original) (raw)
Related papers
Equation of state modeling of high-pressure, high-temperature hydrocarbon density data
The Journal of Supercritical Fluids, 2010
Experimental densities are reported for n-pentane, n-octane, cyclooctane, 2,2,4-trimethylpentane, ndecane, and toluene to ∼280 MPa and ∼250 • C. These densities are in good agreement with available literature data that typically cover lower pressure and temperature ranges than those reported here. The data are modeled with the Peng-Robinson (PR) and Soave-Redlich-Kwong (SRK) cubic equations, the temperature-dependent, volume-translated SRK equation, the temperature-and density-dependent SRK equation, and the SAFT and PC-SAFT equations. Mean absolute percentage deviation (MAPD) values between densities calculated with the PR equation and literature data are 4.55 for n-pentane, 2.91 for n-octane, 3.68 for cyclooctane, 3.98 for 2,2,4-trimethylpentane, 5.58 for n-decane, and 1.99 for toluene. With the exception of 2,2,4-trimethylpentane, these MAPD values are substantially better than those obtained with the SRK and modified SRK equations. Although both SAFT-based models have MAPD values significantly lower than those with the PR equation, the PC-SAFT equation provides the lowest MAPD values.
Fluid Phase Equilibria, 2007
High-pressure density data for cyclohexane + n-hexadecane mixtures at a wide temperature range was modeled with several classical equations of state (EOS) and correlative models. A modification for softening the co-volume and another for a volume scaling of the Peng-Robinson EOS (VS-PR) were proposed. The VS-PR model is able to correlate the pure component experimental data employing only five adjustable parameters, with root-mean-square deviation (RMSD) between calculated and experimental densities essentially within the experimental error. This result is superior to widely used approaches, i.e., a six parameter Tait model and six parameter volume translations (temperature and pressure dependent) for Peng-Robinson and Patel-Teja EOS. The VS-PR model also represents well the isobaric thermal expansion and the isothermal compressibility coefficients of the pure cyclohexane, a small naphthenic substance as well as a long chain n-alkane hydrocarbon, n-hexadecane. When modeling the mixture data, the use of VS-PR model of pure components along with the Redlich-Kister expansion, truncated at the first term, the density was correlated within a RMSD only 60% greater than the experimental error. The proposed model is able to accurately represent all the tested mixture data with a relatively small number of parameters.
Industrial & Engineering Chemistry Research, 2013
Experimental density data for o-xylene, m-xylene, p-xylene, and 2-methylnaphthalene, are reported at pressures (P) to 265 MPa and temperatures (T) to 525 K using a variable-volume, high-pressure cell. The reported data agree to within ±0.4% of available literature data. o-Xylene has the largest densities and p-xylene has the smallest densities in the P−T range investigated in this study although the 525 K isotherms for all three aromatics virtually superpose at high pressures. The aromatic densities are modeled using the Peng−Robinson (PR), high-temperature, high-pressure, volume-translated Peng−Robinson (HTHP VT-PR), and perturbed chain statistical associating fluid theory (PC-SAFT) equations of state (EoS). Generally, the PC-SAFT EoS gives the best predictions of the HTHP density data with mean absolute percent deviations (δ) within 1.0%, even though the pure-component parameters are fitted to low-pressure vapor pressure and saturated liquid density data. δ decreases to 0.4% for calculations with a new set of PC-SAFT parameters obtained from a fit of the HTHP experimental density data obtained in this study.
Fluid Phase Equilibria, 2012
The performance of the SRK and PR cubic equations of state (EoS) for predicting molar volumes at the extremely high temperature, high pressure (HTHP) conditions associated with ultra-deep petroleum formations, are improved with a temperature-dependent volume-translation (VT) term. Rather than correlating the volume-correction to saturated liquid densities, as is done in most prior volume translation methods, the volume-translation term in the HTHP VT-SRK EoS and HTHP VT-PR EoS is correlated to pure component, single-phase density literature data at pressures between 7 and 276 MPa and temperatures between 278 and 533 K. VT parameters are determined for 17 compounds, including short-and longchain alkanes ranging from CH 4 to n-C 40 H 82 , several cycloalkanes, and several aromatics. Our recent HTHP density data for several hydrocarbons have been included in this HTHP density data base to enhance the accuracy of these models. The volume correction parameters are correlated to the inverse of the product of the molecular weight and acentric factor, (Мω) −1 , allowing these models to be used for compounds not included in the data base. The mean absolute percentage deviation (MAPD) values of (1-2%) and (1-4%) obtained with the HTHP VT-SRK EoS and HTHP VT-PR, respectively, are substantially better than those obtained with other models. The proposed models are also successfully extended to mixtures. The new HTHP VT-EoSs do not exhibit any thermodynamic inconsistencies as illustrated by the determination of density, isothermal compressibility, and speed of sound calculations over a very broad range of temperature and pressure.
Journal of Molecular Liquids, 2011
Experimental densities of three groups of liquid organic substances (acids, esters, alcohols) have been correlated using Goharshadi-Morsali-Abbaspour (GMA) equation of state and then the values calculated from the equation of state have been compared with the experimental data. The paper reports new correlation for the density of 20 organic liquids (7 acids, 7 esters and 6 alcohols) at temperatures between 293.15 K and 393.15 K and pressures between 0.1 MPa and 35 MPa. A comparison with experimental data in the specified range of temperature from low to high pressures has been made. Some generalized correlations are also used for comparison with GMA equation of state and experimental data. The results show that the equation of state reproduces the experimental PρT data of liquid organic compounds with good accuracy. The excellent agreement with experimental data indicates that this equation of state can be used to calculate the density of liquid organic compounds with a high degree of certainty. The comparison with other correlations shows that the GMA equation of state is better to some extent and reliable in the given temperature and pressure range.
Experimental measurements and prediction of liquid densities for n-alkane mixtures
The Journal of Chemical Thermodynamics, 2006
We present experimental liquid densities for n-pentane, n-hexane and n-heptane and their binary mixtures from (273.15 to 363.15) K over the entire composition range (for the mixtures) at atmospheric pressure. A vibrating tube densimeter produces the experimental densities. Also, we present a generalized correlation to predict the liquid densities of n-alkanes and their mixtures. We have combined the principle of congruence with the Tait equation to obtain an equation that uses as variables: temperature, pressure and the equivalent carbon number of the mixture. Also, we present a generalized correlation for the atmospheric liquid densities of nalkanes. The average absolute percentage deviation of this equation from the literature experimental density values is 0.26%. The Tait equation has an average percentage deviation of 0.15% from experimental density measurements.
Fluid Phase Equilibria, 2012
At pressures below ∼55 MPa, the perturbed chain-statistically associated fluid theory (PC-SAFT) gives reliable density predictions within ±2% for n-alkanes and other hydrocarbons. However, PC-SAFT tends to over-predict density values by as much as 5% at higher pressures, particularly for normal and branched alkanes. For many compounds, literature values for the three pure-component PC-SAFT parameters m, , and ε/k B are typically obtained by fitting the equation to sub-critical P T data or occasionally both sub-critical and supercritical density data. A new set of pure-component PC-SAFT parameters for density prediction at extreme conditions is reported here by fitting the PC-SAFT equation to single-component density data collected at temperatures from ambient to 533 K and pressures from ∼6.9 to 276 MPa, rather than sub-critical density data since these high temperature, high pressure (HTHP) conditions are similar to conditions typically associated with petroleum recovery from ultra-deep formations. Density predictions made using the new, HTHP PC-SAFT pure-component parameters at HTHP conditions are clearly superior to those obtained using the original PC-SAFT parameters. Although a correction term can be applied to the ε/k B parameter to make HTHP PC-SAFT pure-component density predictions at pressures below 6.9 MPa only slightly inferior to predictions with the original PC-SAFT parameters, vapor-liquid equilibrium predictions with the original PC-SAFT parameters are clearly superior to predictions made with the HTHP parameters. Correlations are developed to accurately predict the HTHP PC-SAFT parameters for normal and branched alkanes for which there are either incomplete or nonexistent experimental density data sets.
Liquid Density of Pure Alkanes and Halogenated Alkanes in a Corresponding States Format
1998
The development of a three parameters Corresponding States (CS) model is here proposed aiming at the prediction of the saturated and compressed liquid density of pure fluids pertaining to the two conformal families of alkanes (A) and hydrofluorocarbons (HFC) which are widely used as refrigerants. Two fluids of the same family are chosen for both their acentric factor value and for the saturated and compressed liquid density dedicated equations availability and, on the basis of the Teja et al. three parameters CS model, the saturated and compressed liquid density of a fluid of interest is obtained in reduced variables. Assuming experimental data of saturated liquid density for several components of each of the two families of fluids an improvement is introduced substituting the acentric factor with a new constant scaling factor. As a final result both the saturated and the compressed liquid models are predictive methods. The reached prediction accuracy of the proposed method is similar to that of the dedicated equations for all the fluids of a family. The result is particularly satisfactory for the application requirements in refrigeration.
Journal of Chemical & Engineering Data, 2011
This paper presents experimental liquid densities for n-pentane, n-octane, and n-nonane and their binary mixtures from (273.15 to 363.15) K over the entire composition range (for the mixtures) at atmospheric pressure. The experimental apparatus is a vibrating-tube densimeter. It is possible to compare the results to a generalized correlation for liquid densities of n-alkanes and to molecular dynamics simulations. The average absolute percentage deviation is (0.06 and 0.8) % using the equation and the simulation results.
Liquid Density of Alkanes and Halogenated Alkanes Mixtures in a Corresponding States Format
1998
The Corresponding States (CS) density models for mixture proposed here, one for saturated and one for compressed liquid, are analytically similar to the pure fluid liquid density exposed in the former paper, but now with critical constants and 8,, replacing wm , as from the mixing rules. The mixing rules present two adjustable interaction coefficients for each binary pair, but they are set to unity making the two models completely predictive: both in fact do not preliminarily require any density data for the mixture of interest. To improve the prediction accuracy a correlative mode is here proposed in which the om parameter is substituted with a 8,, (x) function which parameters are regressed from saturated liquid data, when available, for the binary mixture of interest. The two models are validated with mixtures experimental data for the families of alkanes and hydrofluoroalkanes (HFC) and the prediction accuracy obtained is significatively better with respect to the existing predictive liquid density models for mixtures. The result is particularly useful for the studies about the new generation refrigerants applications.