This paper presents detail models of a multi-split VRF system from the component to the whole system. Two generic algorithms noted as AGM-I and AGM-II are proposed ultimately with an insight into the complex characteristics of the VRF systems. The critical operation rules of the algorithms are addressed.
The simulation models are validated using the experimental data reported in open literatures. The simulation model accuracy to predict system cooling capacity, energy consumption and COP are about 4.69%, 4.64%, 1.44% respectively.
Results of single step simulation by using the AMG-I are depicted at first. The results reveal that the developed simulation model(s) is feasible and effective to simulate VRF systems with different number of evaporators. Relatively small time is cost for each simulation case. The simulation time does not increases significantly as the number of evaporators increasing dramatically, which indicates the developed system model is evaporator-number independent and shows potential to be applied for control analysis. Although the results obtained by using AGM-II were almost the same as those by the AGM-I, computation speed varies. For the VRF system with only one evaporator, the much larger time consuming of the AGM-I weakened its potential to be used in such systems, but for the multi-evaporator VRF system, the AGM-I has advantages over the AGM-II.
Tests also show the system model good ability responding to varying conditions, including the inlet air temperature of the evaporator, the outdoor air temperature, the opening of the EEVs and the compressor speed, which are all very important variables for the control analysis. It is concluded that the system model can not only predict the performances but catch the dynamic trends of the VRF systems with high accuracy.
All the results lay the solid foundation for the further energy performance, basic and optimal control analysis, etc.