PERFORMANCE AND MODELING OF SOLAR VAPOR COMPRESSION REFRIGERATION SYSTEMS FOR COOLING FRUITS, VEGETABLES, AND BEVERAGES UNDER A THAI ENVIRONMENT

Abstract

In this research work, performances of an Auto-Regressive with eXogeneous variables (ARX) and Artificial Neural Network (ANN) were compared. It was found that the ARX outperformed the ANN in predicting load temperature of the solar vapor compression refrigeration system. In the second part, a commercial solar cooling system was experimented. The system consists of a vapor compression refrigeration unit with the capacity of 169 liters, two 300 W solar modules, two 12 V batteries with the capacity of 200 Ah (each), and a charge controller. It was found that the system technically performed well but it is too small for applications in Thailand. Finally, an existing solar refrigeration system with capacity of 789 liters, 550 Watt from electricity gird in Thailand was modified to be a 12 V PV solar refrigeration system. The modified system was experimented and it was found that the load temperatures were reduced to 10 °C – 12 °C within 12 hours for most cases. The experimental results were also employed to model the modified system using the ARX approach. It was also found that the model predict well the load temperature. Additionally, the economic evaluation of the modified system was conducted. Based on the evaluate, the payback period of the modified system was 11.21 years.

-