System Performance And Performance Prediction For High Temperature Heat Pump

With the increasingly serious problems of energy shortage and environmentalpollution, high temperature heat pump technology has enormous potential market aswell as broad industrial application prospect because of its high heating capacityunder current global environment. A new non-azeotropic mixture refrigerant namedBY-4with an ODP (Ozone Depletion Potential) of0and low GWP (Global WarmingPotential) was proposed considering that high temperature was the active demand forheat pump technology. This investigation studied and analyzed its cycle performanceby theoretical study and experimental analysis. This study also conducted thefeasibility and effectiveness analysis on performance prediction of high temperatureheat pump applying support vector machine (SVM). In order to promote thepopularization and development of high temperature heat pump technology, anautomatic test system was built and practical engineering application of hightemperature heat pump was also analyzed and discussed.Theoretical cycle performances of BY-4and other high temperature refrigerants(HFC236ea、HFC236fa、HFC245ca、HFC245fa、HCFC123、HCFC124、CFC114)were analyzed. The result showed that BY-4had good environmental performanceand proper critical temperature (150.2℃). The comprehensive performance of BY-4was optimal under two typical high temperature conditions including a variablecondensing temperature working condition (the range of condensing temperature was90-110℃and evaporating temperature was60℃) and a variable evaporatingtemperature working condition (the range of evaporating temperature was50-70℃and condensing temperature was100℃). The condensing pressure was lower than2.0MPa when the highest heat output temperature was up to110℃. Thecompression ratio and coefficient of performance (COP) were about3and6.1,respectively. Therefore, BY-4with excellent desired cycle performance was worth tobe explored in depth.A life and electrical performance test of scroll compressor SY300A4has beendone in the test bench including three models of M1(with discharge pressure of28bar, suction pressure of1.95bar, discharge temperature of155℃), M2(withdischarge pressure of35.2bar, suction pressure of7.1bar, discharge temperature of 130℃), M3(with discharge pressure of13.6bar, suction pressure of0.78bar,discharge temperature of90℃), to ensure the reliable operation of heat pump systemunder high temperature condition. Compressor ran continuously for168h in everymodel condition. The results showed that the compressor performed well even underhigh temperature and high pressure and electrical performance was also good.The excellent thermodynamic cycle performance of BY-4was obtained underhigh temperature condition. Under the experimental conditions of the inlet watertemperature of evaporator at50-70℃, the outlet water temperature of condenser at75-110℃, The evaporator and condenser water temperature difference between inletand outlet of5℃, The experimental results showed that the COP of heat pump washigher than3.5when the temperature difference between the condenser outlet waterand the evaporator inlet water was less than30℃. The COP was about3when thetemperature difference was less than40℃. The highest output temperature rose up to110℃and COP was up to3.61when the inlet water temperature of evaporator was70℃. However, the maximum pressure and discharge temperature of the systemwere only1.73MPa and111.80℃. The compression ratio was also appropriate. Allfactors were beneficial to safe and stable operation of heat pump system.A support vector machine (SVM) approach was developed to predictperformance of high temperature heat pump system because of single obtainingmethod. Moreover, the performance of the proposed approach was compared withback propagation (BP) neural network. The study conducted performance predictionfor high temperature heat pump system under two conditions including a hightemperature refrigerant (MF1) and two high temperature refrigerants (MF1and MF2).The simulation results showed that the performance of SVM model was superior tothat of BP neural network model under two conditions. The simple linear SVM wasthe optimal choice for the prediction of heating capacity and input power. Moreover,the non-linear SVM with radial basis function (RBF) kernel was the best model forCOP prediction. The results of SVM model were very satisfied. The achievement willprovide guidance for the choice of high temperature refrigerant and optimization ofhigh temperature heat pump.Based on the research results of high temperature heat pump and low-temperaturepower generation technologies, a test platform for the high temperature heat pump andlow-temperature power generation was designed and built. The achievement will provide good experiment platform for two technologies and technical support for theirterminal product and commercialization.