Modelling natural refrigerant mixtures for residential heat pump applications in Python']Python

Natural refrigerants hold some drawbacks which can challenge their application in residential heat pumps, especially those located indoors. Natural mixtures may compensate for the unwanted characteristics of the pure fluids by extending the range of applicability of natural refrigerants, while possibly providing superior performance. A modelling framework was developed in Python and utilized to predict the steady-state operation characteristics of a single-stage heat pump. A variety of key performance indicators considering uncertainties were analysed. DME, R432A and R510A were the best alternatives regarding COP and required refrigerant charge for an application with small temperature glides. Propylene showed the highest volumetric heating capacity while Isobutane showed the lowest discharge temperature. Uncertainties and local sensitivity analysis were evaluated through error propagation. The most sensitive parameters differed significantly from one refrigerant to another. For mixtures containing CO2, mixture composition resulted in being a critical parameter.