Luqman Raji, Amir Abdul Razak, Ahmad Fadzil Sharol
Accurate simulation of photovoltaic (PV) systems is essential for optimal design, performance assessment, and investment decisions, particularly in tropical climates with variable irradiation and temperature profiles. This study aims to conduct a comparative performance analysis of 4 leading PV simulation tools, PVsyst, PVWatts, PVGIS, and RETScreen, for a 3.58 MWp grid-connected rooftop solar system installed in a tropical South Asian country. Site-specific meteorological data and detailed system configurations were input into each simulation tool to model the energy yields, performance ratios, and system losses. The simulations were validated against 12 months of measured operational data using statistical indicators such as the normalised mean bias error (NMBE), coefficient of variation of the root mean square error (CV-RMSE), and coefficient of determination (R²) to assess prediction accuracy. The results showed that PVsyst predicted an annual energy yield of 5123.99 MWh with a performance ratio of 80.46%, whereas PVGIS and RETScreen estimated yields of 5316.43 MWh and 5368.48 MWh, respectively. PVsyst showed better accuracy in modelling shading, temperature effects, and losses, making it suitable for tropical analysis, while PVGIS and RETScreen remained effective for early stage feasibility evaluations. These findings highlight the need for climate-sensitive simulation tools to improve PV planning, optimisation, and modelling accuracy for researchers and policymakers in Southeast Asia and other tropical regions of the world. Future work will focus on integrating machine learning-based correction models into existing simulation frameworks to further enhance the prediction accuracy of large-scale PV systems under diverse climatic conditions.