Accurate data on the thermal properties such as heat capacity and enthalpy are of great importance in process calculations and design. Such data are also useful in checking and improving the correlation procedures for the thermodynamic properties.

The various experimental methods for determining the thermal properties of gases were examined and the heat exchanger method of determining the heat capacity ratios was selected. Equipment was designed, fabricated, and assembled, in order to carry out the necessary experimental measurements.

The data on the ratios of heat capacity at a given pressure to the heat capacity at a low pressure (about 27 psia) were obtained for nitrogen at pressures up to 2250 psi and in a temperature range from 60° to 150°c. The results obtained agreed with the data available in the literature within the expected accuracies of about 0.5 percent for this work and the literature results. Data were also collected for two binary mixtures of carbon dioхide and methane containing 14.5 and 42.3 mole percent methane. The data were obtained in the temperature range from about 40° to 150°C and at pressures up to 2250 psi. The expected accuracies of the heat capacity ratios thus determined were better than ± 0.5 and ± 1 percent in the regions removed from and close to the maxima respectively.

Various mathematical techniques for estimating the co-efficients of the Benedict-Webb-Rubin equation of state from only volumetric information were studied. These techniques were then extended for estimating the parameters by utilizing volumetric and heat capacity information, and the required computer programmes were developed.

An extensive evaluation of the various parameter sets obtained by different mathematical techniques was made, and their ability to predict the volumetric and heat capacity data of carbon dioxide, methane, and carbon dioxide-methane mixtures was compared. These comparisons were made in the region of pressures up to. 5000 psi and temperatures between 40° and 200°c. The parameters estimated by the non-linear least squares method, choosing volume as a dependent variable, gave the best predictions of the compressibility factors and the heat capacities of the pure components. This was also true for the mixtures when the BWR combining rules were used. For correlating more than one property of a pure component, the non-linear least squares method of minimizing the sum of the weighted sum of relative error squares in volume and heat capacity was found to be superior to the use of the linear programming technique:​ For a given set of pure component cоefficients, the P-V-T and heat capacities of mixtures were correlated better by using a mixing rule that provided an interaction term rather than the BWR mixing rules.