This paper presents a longitudinal performance evaluation of a rooftop grid-connected solar photovoltaic (GSPV) plant, emphasizing alterations in energy production and performance ratio (PR) over time. The assessment involves year-wise data analysis for 6 years of a 1.39 MW GSPV plant in north India. The purpose is to assess performance, degradation trends in energy production, system losses, and capability under varying climatic conditions. The key indicators critically examined to quantify plant performance are energy production, PR, capacity factor, yield factors, losses, and rate of degradation of the plant. The total energy production received in 6 years is 8887.91 MWh. The average annual reference yield, array yield, and final yield are 4.47, 3.09, and 2.29, respectively. In order to create a technical benchmark, the plant is modeled in PVsyst software and analyzed annually using site-specific meteorological and system configuration data. The simulated results are compared with the real-time measured data of the plant obtained from the metering and recording unit to confirm the reliability and accuracy of the simulation-based performance prediction. At the end, the percentage error (% ξ) and mean average percentage errors (MAPE) between real-time and simulated results are evaluated. The MAPE is small, less than 5% for the initial year, which increased in subsequent years due to inadequate assumptions in factors like soiling, thermal losses, component ageing, etc. This study highlights the factors that essentially affect the accuracy of the static baseline model for long-term performance prediction.
Tiwari et al. (Fri,) studied this question.