The thermal conductivity of hydrocarbon gases at normal pressures

Abstract

Dimensional analysis has been applied for the correlation of the thermal conductivity of a gas to its temperature, molecular weight, heat capacity, and critical constants. This approach indicates that the group k*h/C, should be a function of zc and T,, where h = M1/aT,1/E/Poa/3. Experimental thermal conductivities of hydrocarbons a t normal pressures (approximately 0.2 to 5 atm.) have been used to develop two relationships. The first is applicable to all types of hydrocarbons for 0.6 < T R < 3.0 with the exception of methane ond the cyclic hydrocarbons below TR = 1.0, for which the other relationship is applicable. These two relationships have been used to calculate thermal conductivities for twenty-eight gaseous hydrocarbons for which experimental data are available. Calculated values for normal paraffins, isoparaffins, olefins, diolefins, acetylenes, naphthenes, and aromatics produce an average deviation of 2.4% from experimental values for 154 points considered.

Thermal Conductivity of Real Substances from Excess Entropy Scaling Using PCP-SAFT

“…Predicting the transport properties of polar or associating substances remains difficult. We can distinguish three methods for analytically estimating thermal conductivities: corresponding states approaches; , description of the reduced transport coefficients through another characteristic property, e.g., a characteristic volume; − and approaches that achieve a collapse of experimental values to universal lines for certain ranges. − …”

“…The collapse is achieved through an empirical scaling in the general form αλ = f (β), where α and β are functions of temperature, density, molecular mass, and critical point properties. Definitions of α and β vary between different authors. − The resulting quasi-universal relationships f can easily be described through polynomials and lead to good descriptions for a certain phase and for certain classes of substances, e.g., for nonpolar substances in gaseous phase at atmospheric pressure …”

Estimation of thermal conductivity of pure gases by using artificial neural networks

“…These correlations, in addition to temperature, critical temperature, critical pressure and molecular weight, require accurate values for the heat capacity at constant volume and also vapors viscosity which is usually not available for compounds at any temperature (e.g., carbon dioxide at 1500 K and carbon monoxide at 81.88 and 91.88 K in Table 5). Misic and Thodos [19,20] recommended two correlations for pure component, low pressure (<350 kPa) hydrocarbon gases. One of their correlations has been proposed for methane and cyclic compounds below reduced temperatures of 1.0, and the other one has been recommended for all hydrocarbons above reduced temperatures of 1.0.…”

“…The best architecture of the feedforward network, obtained by trial and 10.75 (1) Not used in the training of the proposed ANN. (2) The correlation of Misic and Thodos [19,20]. (3) The correlation of Bromley for pure non-hydrocarbon monatomic gases [21].…”

“…Misic and Thodos [19,20] For pure component, low pressure (<350 kPa) hydrocarbon gases, they recommend two correlations. One of them for methane and cyclic compounds below reduced temperatures of 1.0, and another for these hydrocarbons above reduced temperatures of 1.0 and for other hydrocarbons at any temperature.…”