Systematic measurement errors frequently affect the anemometers installed on wind turbines, reducing the generator conversion efficiency, and resulting in significant discrepancies between the expected and the observed annual energy production. Despite these critical effects, the literature appears to lack a comprehensive focus on this topic. To address this research gap, this study proposes a novel data-driven multi-step methodology aimed at detecting systematic measurement errors in wind turbine anemometers and assessing their impacts on the generated power profiles. The proposed approach leverages the analysis of operation data of wind turbines by detecting their performance deterioration and divides a wind farm in potentially anomalous and normal wind turbines. Then, a data-driven regression model predicts the wind speed measured by the anomalous wind turbines as a function of the measurements collected at the normal wind turbines and, through the analysis of the residuals between model estimates and measurements, the systematic anemometer error is identified and its onset is correlated with a performance drop. Experimental data from 11 industrial wind turbines during an observation period of 10 years are elaborated to assess the effectiveness of the proposed framework. For the selected test case, anemometer errors ranging between 0.36 and 0.43 m s-1 have been identified, to which a loss of at least 3% and up to 7% of the yearly producible energy is estimated to be associated. Such data analysis inspired the on-site inspection, which confirmed the results of this work.
Assessing the effects of anemometer systematic errors on wind generators performance by data-driven techniques
Astolfi D.;De Caro F.;Castellani F.;Vaccaro A.;Flammini A.
2024
Abstract
Systematic measurement errors frequently affect the anemometers installed on wind turbines, reducing the generator conversion efficiency, and resulting in significant discrepancies between the expected and the observed annual energy production. Despite these critical effects, the literature appears to lack a comprehensive focus on this topic. To address this research gap, this study proposes a novel data-driven multi-step methodology aimed at detecting systematic measurement errors in wind turbine anemometers and assessing their impacts on the generated power profiles. The proposed approach leverages the analysis of operation data of wind turbines by detecting their performance deterioration and divides a wind farm in potentially anomalous and normal wind turbines. Then, a data-driven regression model predicts the wind speed measured by the anomalous wind turbines as a function of the measurements collected at the normal wind turbines and, through the analysis of the residuals between model estimates and measurements, the systematic anemometer error is identified and its onset is correlated with a performance drop. Experimental data from 11 industrial wind turbines during an observation period of 10 years are elaborated to assess the effectiveness of the proposed framework. For the selected test case, anemometer errors ranging between 0.36 and 0.43 m s-1 have been identified, to which a loss of at least 3% and up to 7% of the yearly producible energy is estimated to be associated. Such data analysis inspired the on-site inspection, which confirmed the results of this work.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.