Thesis (M.Eng.) - Asian Institute of Technology, 1993

A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering.

One major cause for failures in photovoltaic (PV) medical refrigerators is the improper sizing of the PV array and battery capacities. The use of average insolation and estimated number of continuous "no sun" days for the sizing is inadequate due to the dynamic behavior of PV system and stochastic nature of solar radiation. A computer simulation is therefore required to analyze the feasibility of a PV refrigeration system in terms of the desired output and reliability with respect to the climatological data of the location to be installed. Existing PV system simulation models apply only to PV systems with a constant or uniformly varying energy demand throughout the day. In this study, a simulation model was developed for a stand-alone PV refrigeration system, which uses a refrigerator designed according to the WHO standard specifications (i.e. the refrigerator is capable to maintain vaccine temperature between 0°C and 8°C and produce at least 2kg of ice within 24 hours). Models were developed for the main system components, i.e. the PV array, controller, battery and refrigerator. The model parameters were found through experimental tests performed on each component and the validity of each model was tested. The system simulation model was developed by the establishment of links among these sub-system models. Also a provision was made to compute reliability index for the PV system under investigation. An experimental investigation was carried out in order to test the validity of the system simulation model. According to the experimental investigation, the predicted PV array performance well agreed with the actual performance. Battery model also could predict the battery voltage satisfactorily except at very low currents. The simplified refrigerator model developed in this study gave fairly good results in performance prediction of the selected refrigerator for validation test, during stable running period. But it could not accurately predict the performance during ice-pack freezing, due to the influence of skin condensers of two cooling circuits of the selected refrigerator. However, good results can be expected for refrigerators with normal condensers. The limited time and unavailability of test room, prevented any improvements on the refrigerator model. The drawbacks of the existing model in applying to refrigerators with skin condensers have been pointed out and suggestions for the improvements have been recommended for future development. Finally, the relationships between the system reliability and the battery capacity at different PV array sizes were established for the PV refrigeration system from the results of an year long simulation performed using simulated meteorological data, under an assumed pattern of ice-pack loading. It demonstrated the applicability of the simulation model for sizing system components.