Use of accurate forecasting of direct normal irradiance (DNI) would allow for establishing optimized operational strategies in concentrating solar power plants that lead to increase plant profitability. Our research work* analyses this statement and presents a methodology based on two multivariate regression (MR) models and weather forecast covering a 24-h period. This one is provided by the Integrated Forecasting System (IFS), the global model from the European Centre for Medium-Range Weather Forecasts (ECMWF). Considering two consecutive years of local measured and forecasted data, the models were trained using the first year and tested using the following one, to enhance the original forecasts.
The first MR model deals with the daily energy collected by DNI (called DNI availability) and employs previous-day measured DNI availability as well as meteorological variables to enhance next-day DNI availability forecasts, which shows an increase of roughly 6% towards the original predictions, using the skill score metric. A more accurate model uses measured DNI from previous hours and predicted meteorological variables, such as cloud cover, to improve next-hour predictions. As a nowcasting tool, this multivariate regression model significantly improves hourly predictions, with an increase of around 27.29% towards the original hourly forecasts. An economic analysis shows that this methodology lead to a power plant’s profit in 0.44 M€/year, approximately, as compared with the original forecasts. Operational strategies are proposed accordingly.
(*) Improved ECMWF forecasts of direct normal irradiance: A tool for better operational strategies in concentrating solar power plants, Francisco Lopes, Ricardo Conceição, Hugo Silva, Rui Salgado, Manuel Collares-Pereira. Renewable Energy Vol. 163, January 2021, Pag. 755-771. https://www.sciencedirect.com/science/article/pii/S0960148120313859
More information: Ricardo Conceição, Postdoctoral researcher, High Temperature Processes Unit, email@example.com
Hourly clearness indices for DNI (kb) between measurements (OBS) and improved DNI predictions with (a) SRM, (b) MRM1 and (c) MRM2, where the identity line y=x is represented as a red-dashed line. Analysis is performed over one year of data in Évora (southern Portugal), from July 1st 2018 to June 30th 2019. The respective hourly probability density functions (PDF’s) are also depicted (d, e and f).