Numerical Transient Exergy Analysis of a Battery-Free Solar-Powered Phase Change Preservation System: A tomato Case Study under Tropical Conditions
Description
The proposed battery-free solar-powered refrigeration system comprises four integrated subsystems: (i) a photovoltaic (PV) power supply, (ii) a DC vapor-compression refrigeration cycle, (iii) a phase-change-material (PCM) thermal-storage unit, and (iv) a tomato-storage chamber. The system is designed to use solar energy directly during the charging period while storing cooling energy in the PCM for operation during periods of insufficient or unavailable solar radiation. The PV array supplies the DC compressor through an MPPT/DC-DC controller. A supercapacitor is incorporated between the controller and compressor to provide short-term electrical buffering, smooth transient PV fluctuations and supply the compressor starting current. The refrigeration cycle employs R-290 (propane) and consists of a hermetic DC compressor, condenser, thermostatic expansion valve (TXV) and evaporator. During charging, the evaporator extracts heat from the PCM, causing it to cool and progressively undergo phase change. The PCM chamber surrounds the tomato compartment, providing a thermal barrier and enabling passive cooling during the discharge period. The storage chamber contains 100 kg of tomatoes, while the PCM provides the stored cooling capacity required to maintain the product within the prescribed temperature range. The overall energy and exergy pathway is therefore: Solar radiation → PV array → MPPT/DC-DC controller →Supercapacitor → compressor → refrigeration cycle → PCM → tomato chamber.
Files
Steps to reproduce
Reproducibility Steps 1. Install software: Install GNU Octave and COCO Simulator (Cape-Open to Cape-Open). 2. Download and extract dataset: Unzip the Mendeley Data folder. Read README.txt for variable names, units, and file mapping. 3. Run steady-state COCO model: Open the /coco/ files. Run the R-290 vapour-compression cycle using the Peng–Robinson property package. Export refrigerant state points and cycle inputs to /outputs/coco_statepoints.csv. 4. Check Octave paths: Open GNU Octave. Set the working directory to /octave/ and add /inputs/ and /outputs/ to the path. 5. Run transient scripts in order: Execute the scripts in this sequence (or their equivalents): solar irradiance → PV exergy → controller/supercapacitor → compressor → condenser → TXV → evaporator → PCM charging → PCM discharging → tomato chamber → integrated exergy balance → plotting. 6. Confirm simulation settings: Use a 60-second time step, 6-hour charging period, and 48-hour discharge period. The solar irradiance script uses a fixed random seed for reproducibility. 7. Generate outputs: Results are written to /outputs/ as CSV/MAT files. Figures are generated in /figures/. 8. Verify balances: Check that energy and exergy balances close with residuals below 0.5% and 0.1%, respectively. Compare key values against those listed in the Description. 9. Reuse or modify: Edit parameters in /inputs/ to test new climates, loads, PCM masses, or component efficiencies. Cite the dataset and the related manuscript when reusing.
Institutions
- Federal University of Petroleum Resource EffurunDelta, Warri