Analysis and forecasting of photovoltaic energy capacity factors in Greece

Solar energy is one of the most important sources of renewable energy for Greece, due to the abundance of solar radiation the country receives. However, energy production from photovoltaic (PV) systems is directly linked to meteorological processes such as solar radiation and temperature, introducing uncertainty regarding their performance. Accurate forecasting of the capacity factor (CF) is essential for optimizing energy production management and ensuring its smooth integration into the power grid. This thesis develops and analyzes four CF analysismodels, based on real meteorological data from various regions of Greece, aiming to improve data accuracy and, consequently, estimation reliability, as well as to enhance the efficient utilization of PV systems. Specifically, hourly energy production data from PV systems in Greece for the years 2015-2023 were collected via the ENTSO-E platform, while installed PV capacity data for the same years were retrieved from the DAPEEP reports. These data were analyzed to generate an hourly capacity factor time series for Greece over the specified years. Subsequently, 40 representative locations across the country were selected, and hourly solar radiation and temperature data were obtained through the PVGIS platform, forming corresponding time series that were validated using ground-based stations. These datasets were then correlated with the capacity factor, leading to the development of four models (forecasting and simulation, linear and power-type) to determine the most efficient approach for CF estimation and forecasting.