(666e) Experimental Measurements and Predictions of the Properties of Refrigerant Fluids Using Ab-Initio, SAFT-VR Mie and COSMO Calculations
AIChE Annual Meeting
2016
2016 AIChE Annual Meeting
Engineering Sciences and Fundamentals
Thermophysical Properties and Phase Behavior III
Thursday, November 17, 2016 - 9:42am to 10:00am
First, we performed ab initio calculations on isolated molecules to determine the ideal gas properties of refrigerant fluids. Ideal enthalpies of formation and heat capacities can be accurately predicted by using either DFT [1] or some specific semi-empirical methods. We then used several activity coefficient models based on the COSMO approach, to predict the VLE of refrigerant mixtures. VLE and azeotropes can be well predicted by the COSMO-RS model [2]. While the first versions of COSMO-SAC are very inaccurate for such systems, it is possible to obtain good predictions with the COSMO-SAC-dsp model [3] by readjusting some universal parameters. New experimental measurements of vapor-liquid equilibria on mixtures involving HFO have been performed with a static analytic equipment, to complete the literature data for mixtures of HFO with other refrigerants.
We then propose a modified version of the SAFT-VR-Mie [4] equation of state (EoS) that includes a polar contribution to explicitly take electrostatic interaction into account. The polar SAFT-VR-Mie EoS is used to describe the VLE and some derivative thermodynamic properties of refrigerant mixtures. A new association model that takes the formation of cyclic clusters is proposed for hydrogen fluoride (HF), which is one major compound of the refrigerant production process. The new models can be used in process simulation software, thanks to the cape-open and Simulis Thermodynamics® interfaces.
References
[1] A. Osmont et al., Combustion and Flame, 151, 262 (2007)
[2] F. Eckert, A. Klamt, Fluid Phase Equilib., 172, 73 (2010)
[3] C.-H. Hsieh et al., Fluid Phase Equilib., 367, 709 (2014)
[4] T. Lafitte et al., J. Chem. Phys., 139,154504 (2013).
Acknowledgements
The results presented in this work are part of the ongoing project "PREDIREF". We are grateful to the National French Research Agency (ANR-13-CDII-0008) for providing financial support to this project.