Effect of the refrigerant charge, expansion restriction, and compressor speed interactions on the energy performance of household refrigerators

•The combined effect of the refrigerant charge and the expansion device restriction was investigated.•294 energy consumption tests were conducted at different surrounding temperatures and compressor speeds.•Minimal energy consumption figures were achieved for the highest charges and restrictions.•A single charge-capillary tube pair provided satisfactory performance levels regardless of the compressor speed.•It was found that only 9 datapoints are required to generate a full energy consumption map. This work is aimed at investigating the effects of the refrigerant charge, expansion device restriction, compressor speed, and their interactions on the performance of a household refrigerator. To this end, a micrometric valve was installed in series with the capillary tube of a frost-free refrigerator to vary the expansion restriction within a range equivalent to capillary tubes with inner diameters spanning from 0.55 to 0.85 mm. The refrigerant charge was manually changed from 35 to 65 g whereas the compressor speed was adjusted between 2500 and 4500 rpm. In total, 294 steady-state energy consumption tests were conducted, being 147 at a surrounding temperature of 32 °C, and the rest at 16 °C. It was noticed that low energy consumptions can be achieved for several combination of refrigerant charge and expansion restriction, albeit the minimal energy consumption figures were always achieved for the highest charges and restrictions. It was also verified that the refrigerator operates in a sub-optimum region so that increasing the charge from 42 g to 52 g while decreasing the capillary tube diameter from 0.64 mm to 0.60 mm led to a 3% decrease on the energy consumption. It was also found that a single combination of refrigerant charge and equivalent diameter provided satisfactory performance levels regardless of the compressor speed. Finally, a method to reduce the number of tests was proposed, in such a way that only 9 datapoints are required to generate a full energy consumption map for a given compressor speed and surrounding temperature.