Analysis of the systematic errors of energy yield assessment in the context of wind farm repowering

Abstract The repowering of wind farms is a rapidly developing area research and expected to represent 40 GW projects by 2030 in the European Union. This has led emergence energy yield assessment methods that incorporate operational data from existing with aim surpassing traditional rely solely on physical modelling onsite measurements. current literature relies upon assumption learning farm applies future farm. Indeed, calibration adjustment assume physics-driven models (PDMs) have spatially temporally correlated errors. study investigates this analysing PDM errors for 25 pairs nearby projects. A statistically significant correlation observed. We discuss whether it reasonable utilise farms, which possess different characteristics, improve long-term production prediction repowered