Prediction of the physical properties of next-generation refrigerant fluids – PREDIREF

On account of the constraints imposed by the European and International legislations, the refrigerant industry must constantly find alternative refrigerant fluids that have lower impacts on the global warming of Earth and Ozone layer.

Besides the safety and environmental constraints, these alternative fluids must have proper thermodynamic properties in order to be considered as efficient refrigerants. Their energetic properties must be equivalent to the previous fluid in order to avoid important modifications of the system and heavy investment costs. Working with refrigerant blends is often preferable to pure component fluids for energy saving and flexibility of operation. In order to select the optimal mixture composition for the design and operation of a refrigeration process, it is necessary to know the phase diagram and thermodynamic properties of mixtures. Vapor-liquid equilibria (VLE) and the location of azeotropes must be accurately known.

HydroFluoro-olefins (HFO) such as HFO-1234yf and HFO-1234ze have recently been considered as promising replacements, as they have a particularly low Global Warming Potentials (GWP). However experimental data are scarce for these molecules, in particular for mixtures. Experimental measurements can be difficult, time consuming and expensive. Thus, predictive methods are of great interest for the refrigeration industry. Recently, Liu et al. 2 has tested several working fluids which can be used in organic Rankine Cycles. They concluded that predictive methods for pure components and mixtures must be developed in order to have a quick screening of potentially interesting fluids.

The main objective of the PREDIREF project is the development of novel approaches to predict the physical properties of the next generation of refrigerant fluids such as HFOs. The project aims at bringing together predictive methodologies that are often used separately, and proposing a general approach for molecular design. Such an approach will be new in the field of refrigerants. We will first developed transferable force fields for HFO compounds to predict the thermodynamic and transport properties of pure compounds and mixtures, using both Monte-Carlo and Molecular Dynamic simulations. Ab initio calculations will be used to develop such force fields. Second, we will use the COSMO-RS and COSMO-SAC models to predict the phase behavior of mixtures. Finally, the SAFT-Mie equation of state will be employed to predict both phase equilibria and thermodynamic properties such as heat capacities, densities and heat of vaporization. Experimental measurements of phase equilibria, densities and heat capacities will be carried out in parallel, to complete the database which will be obtained from the literature and validate the predictive methods.

The PREDIREF project involves two academic laboratories and two companies that will share their knowledge in different fields, and will be coordinated by Arkema research centre to ensure the success of the project.