Comparison of PV Simulation Tools with Real-World Data: Residential Rooftop Systems in Surabaya, Indonesia

Accurate prediction of photovoltaic (PV) energy output is essential for system design, performance assessment, and investment decisions. However, the accuracy of widely used PV simulation tools in tropical residential settings remains insufficiently validated. This study aims to evaluate the predictive performance of three commonly used simulation platforms—PVWatts®, PVGIS, and PVSOL Online—by comparing their monthly and annual energy output estimates with real-world data from a 4.5 kWp residential rooftop PV system in Surabaya, Indonesia. Using standardized input parameters and one year of measured operational data (2022), the results show that all three tools produce reasonably accurate estimates, with annual deviations below 3%. PVSOL Online yields the closest agreement (+0.56%), followed by PVGIS (+1.52%) and PVWatts® (+2.48%). Although all tools exhibit consistent monthly overestimation, especially during high-yield dry-season months, they successfully capture the seasonal performance pattern of the system. These findings highlight both the strengths and limitations of current simulation tools in tropical environments and underscore the importance of model calibration and field-data validation. The study provides practical guidance for selecting appropriate simulation tools and contributes to improving PV performance modeling for residential applications in humid tropical urban regions.