Binary refrigerant-oil phase equilibrium using the simplified SAFT equation

Document Type


Date of Original Version



Refrigerants such as tetrafluoroethane (R134a) with compatible oils such as polyalkaline glycols (PAGs) and neopentyl (polyol) esters are now being used as replacements for CFCs like freons in household and automobile compressors as well as a number of other applications. Thus an understanding and accurate quantitative description of the phase equilibrium behavior of binary refrigerant-oil mixtures is important. For example, seal failure in compressor operations can be adequately described through a combination of the associated refrigerant-oil phase equilibrium and a two-phase flow model. Data for modeling the behavior of liquid and/or vapor refrigerant is usually readily available. However, in many realistic situations the chemical structure (i.e. molecular weight, segment length, presence or absence of branching, etc.) and associated physical properties (vapor pressure, liquid density, etc.) of the oil are either unknown or unavailable for proprietary reasons and this makes modeling any phase equilibrium very difficult. Often times the required properties of the oil must be estimated. It is also important to keep in mind that accurate modeling must address the relative size of the molecules involved or that the refrigerant molecule has a low molecular weight and is largely spherical while the oil has a higher molecular weight and can be either branched or long chain. A unified framework for binary refrigerant-oil phase equilibrium consisting of chemical structure analyses, mathematical modeling, and experimental data is presented and applied to the phase equilibrium of R134a and neopentyl ester oils. In particular, gas chromatography (GC) and GC-mass selective detector (GC-MS) analyses are used to characterize the molecular weight and branching structure of a given polyol ester. From this, group contributions methods are used to estimate vapor pressures and liquid densities of the oil as a function of temperature and pressure. Group contribution-generated vapor pressure and liquid density data for the oil are then used to regress statistical associating fluid theory (SAFT) parameters for the simplified SAFT equation of state (EOS). Phase diagrams for R134a-polyol ester for a number of temperatures are generated using the simplified SAFT EOS and compared to the experimental data of Takaishi and Oguchi (1993). A number of equation-solving issues are also addressed. Comparison of the modeling results with experimental data clearly shows that the proposed framework of chemical analyses, group contribution methods and the simplified SAFT equation provides an excellent description of refrigerant-oil phase equilibrium. © 2002 Elsevier Science Ltd. All rights reserved.

Publication Title

Advances in Environmental Research