Estimating the value of ECMWF EPS for photovoltaic power forecasting

We conduct a comparative study of deterministic-to-probabilistic (D2P) and probabilistic-to-probabilistic (P2P) forecasting methods for photovoltaic (PV) power generation. In this analysis, we go beyond traditional statistical metrics to introduce a novel metric in the field of PV forecasting. This metric evaluates the economic value of PV production across all possible cost–loss ratios, offering a comprehensive view of the forecast’s utility at different probability thresholds. The study, based on real-world data from PV plant production, includes a comparative assessment of forecasting techniques using the ECMWF’s ensemble forecasting system (EPS) for the P2P approach, and contrasts this with methods based on ECMWF’s deterministic weather forecasts for the D2P approach. While the advantages of using EPS for forecasting might not be immediately apparent through conventional metrics, a detailed examination of the statistical significance of the results, with and without EPS, demonstrates distinct and significant advantages of the former, especially in terms of economic value. This innovative approach to estimating the value of the forecast could be used by energy resource managers to perform an effective a priori cost–benefit analysis and assess whether the additional investment required to implement an EPS-based forecast is cost-effective compared to the deterministic approach alone.

@Article{MM24,
  author       = {Marrocu, M. and Massidda, L.},
  title        = {Estimating the value of ECMWF EPS for photovoltaic power forecasting},
  journal      = {Solar Energy},
  volume       = {279},
  pages        = {112801},
  month        = {september},
  year         = {2024},
  publisher    = {Elsevier},
  keywords     = {visual and data-intensive computing, energy, meteo},
  doi          = {10.1016/j.solener.2024.112801},
  url          = {https://publications.crs4.it/pubdocs/2024/MM24},
}