Comparison of a CO2 refrigeration and heat pump test system with and without ejector.

This experimental study investigates the performance and some questions related to controllability of a smallscale CO2 ejector test system. The test system is operated in both refrigeration and heat pump operation with and without the ejector. Different inlet temperatures on the secondary side of the gas cooler are investigated for both operations as well as a variation of compressor speed. Additionally, the possibility to flood the evaporator in ejector mode is studied including the use of a prototype “quality sensorTM” (capacitive void fraction sensor). The results showed that the ejector performed well in refrigeration operation, considering the pressure drop caused by the undersized piping from the diffuser to the suction of the compressor. The COP improved by around 0.2 (12 % to 17 %) and if the superheat (10 K) was eliminated by flooding the evaporator, a further 0.3 improvement was possible (33 % compared with standard cycle mode). In heat pump operation, the COP became similar or lower because the ejector was slightly oversized for that operation (gas cooler pressure became lower than optimal). The quality sensor showed a significant change in the electrical capacitance as the evaporator was increasingly flooded. Furthermore, the measured capacitances were converted into reasonable void fractions by using both a linear calibration function and an annular flow void fraction calibration function from the literature. The gas cooler pressure was possible to control with the speed of the compressor within a range of around 10 bar in refrigeration mode, but had insignificant effect on the gas cooler pressure in heat pump mode.