Presentations

Abstract

Accurately estimating the annual energy production (AEP) of wind farms is critical for project planning and financing. However, discrepancies between pre-construction AEP estimates and actual post-construction performance are not uncommon, often leaving asset owners and potential buyers uncertain about the underlying causes. These discrepancies can stem from various sources, including misestimations of the wind resource and deviations in turbine performance. To address this challenge, we propose a systematic workflow for post-construction assessment of wind farms, aimed at dissociating and quantifying AEP deficits from different origins—namely, wind resource overestimation, turbine performance deviations, and other contributing factors. The workflow comprises two primary components: minimizing wind flow model uncertainty and assessing actual turbine power output performance.  The first component leverages advanced modeling approaches and calibration techniques to minimize uncertainty in wind resource estimation. Using a state-of-the-art mesoscale-microscale model chain, the workflow incorporates multi-source data, including wind turbine operational data (SCADA) and on-site wind measurements, for calibration. The specific tuning strategies applied are adaptable to the characteristics of the wind flow model and the available data in individual projects. By comparing AEP simulations generated from the original and tuned wind flow models, this approach quantifies the contribution of wind flow model uncertainty to the overall AEP discrepancy. The second component centers on accessing the actual turbine power output performance, with a primary focus on deriving the true power curve. This requires knowledge of the free-stream wind speed, corrected for turbine blade disturbances. On-site wind measurements—obtained via met masts, ground-based lidars, or nacelle-mounted lidars—serve as key inputs for nacelle transfer function (NTF) assessment as per IEC standards (IEC 61400-12-1, -12-2 and -50-3). In complex terrain, numerical site calibration (NSC) in accordance with IEC TR 61400-12-4 is employed to refine power curve measurements without dismantling wind turbines. After performing rigorous data quality checks and classifying turbine operation statuses, SCADA data is processed to extract valid operating conditions for NTF assessment. When turbine status codes are unavailable, we employ a robust algorithm that relies on 10-minute nacelle wind speed and output power readings as minimal inputs. The results of two AEP simulations, based on the theoretical and actual power curves, are used to quantify the contribution of turbine power curve performance to the AEP deficit. Finally, the difference between the modeled production value—calculated using both the tuned wind flow model and the actual power curve—and the actual production value reveals additional losses attributable to other factors. The proposed methodology was demonstrated on a specific wind farm experiencing significant underperformance. The analysis uncovered that the discrepancy was due to overestimated wind resource potential as well as suboptimal turbine power curve performance. This study presents a robust and efficient workflow for post-construction AEP assessment, highlighting data collection requirement, key processing steps, and practical implementation using available tools. By applying this methodology, wind project stakeholders can gain deeper insights into the factors contributing to AEP deficits, understanding project quality in terms of wind resource and turbine performance, and ultimately enabling more informed decision-making.