File(s) under embargo
until file(s) become available
Experimental and Numerical Analysis of Performance Enhancements to a Multi-Stage Two-Evaporator Transcritical Carbon Dioxide Refrigeration Cycle
Due to increasing environmental concerns and stringent regulations, the Heating, Ventilation, Air Conditioning, and Refrigeration (HVAC&R) industry is working to develop technologies that utilize low-global warming potential (GWP) refrigerants and remain within competitive coefficient of performance (COP) values and capacities of current hydrofluorocarbon (HFC) systems. Carbon Dioxide (CO2) has been investigated extensively over the past 25 years as a potential substitute for HFCs in refrigeration applications in mild ambient climates. Through efforts to increase the efficiency of CO2 systems, researchers and industry have identified cycle modifications that are particularly beneficial in transcritical CO2 cycle applications, such as expansion work recovery and economization. Yet, systematic experimental comparisons between these cycle enhancements are still lacking in the open literature. This thesis presents the design, assembly, and operation of a multi-stage, two-evaporator transcritical CO2 cycle with a combined capacity between the two independently-controlled evaporators of approximately 8 kW. The cycle utilizes three stages of compression, intercooling between the second and third stages, and flash tank economization at the medium-temperature evaporator. Furthermore, the test stand was designed to enable on-line transition between three methods of expansion without the need to stop the compressors. In particular, expansion through an electronic expansion valve (EXV), ejector, and through an ejector with a pump used to modulate the ejector inlet state are investigated. The purpose of this experimental work is to provide a comprehensive comparison between these cycle architectures and methods of expansion work recovery as well as to assess both the pump and variable motive nozzle diameter as means of ejector control and system performance enhancement. Experimental testing assessed the performance of cycles utilizing economization, an ejector, and an ejector with a pump over four ambient conditions from 14 °C to 28 °C. The evaporator source temperatures were fixed to simulate refrigeration and freezing conditions. The pump controlled the ejector effectively and increased the ejector efficiency by up to 41%, despite decreasing the cycle COP. COP improvements of 6% and 5% were achieved with the open economization and ejector cycles, respectively.
In addition to the experimental aspects, a numerical ejector design tool that can be applied to a vapor compression cycle was developed. The model solves each subcomponent of the ejector intensively, and receives inputs of mass flow rates, geometric angles, and ratios to output physical dimensions of an ejector. The design tool agreed with experimental dimensions with a mean absolute error (MAE) of 3% to 4%.
A second numerical aspect of this thesis consists of a dynamic model of the transcritical CO2 vapor compression cycle test stand used for experimental testing. The dynamic model can be used to predict both steady state performance and dynamic performance of the system, and employs component models for the heat exchangers, compressors, and expansion valves utilized in the experimental setup. Steady state validation resulted in maximum MAE of 18.7% and 11.9% for cooling capacity and power consumption, respectively, and dynamic validation resulted in similar thermal time scales between experimental data and the simulation. An evaporator pull down and an evaporator excitation simulation were successfully implemented to validate the ability of the model to develop control schemes.
Experimental future work consists of both cycle modification recommendations and control method implementation. Ejector design future work is the implementation of higher-fidelity sub-component models, and future work for the dynamic model is the thorough characterization of mass transfer and development of models of additional cycle architectures.