Multiple wake interactions in large wind farms

目前在美国部署的风力发电场通常包含数百个风力涡轮机，装机容量超过100兆瓦。这些大型阵列中的风力涡轮机尾迹导致风力发电场的总输出减少了20%。这些尾迹，包括风力涡轮机后面的风速下降和湍流增强的区域，也减少了涡轮机的寿命，因为增加了疲劳负荷。π吗?先前的研究表明，当前的风力发电场模型低估了大型阵列风力涡轮机尾迹的大小。本项目的主要目标是：(1)提高对风速、湍流和大气稳定条件下单、双和多尾迹发展的物理认识和建模；(2)通过更好地对风力发电机尾迹进行量化和建模，评估是否可以显著降低功率预测的不确定性，改善阵列配置。通过对风力发电机尾迹之间以及整个风电场尾迹与上覆大气之间的相互作用进行显式建模，可以大大减少预测大型风电场输出功率的不确定性。这项研究将涉及决定风力发电厂效率的因素的先进测量和建模，适用于目前部署的大型风力涡轮机和风力发电厂。pi将集中于风力涡轮机尾迹造成的功率损失和下游涡轮机额外疲劳负荷的量化，包括三个部分：1)使用多普勒光探测和测距（激光雷达）对风力涡轮机尾迹和相关大气和涡轮机参数进行高分辨率测量。pi将在大型风电场中使用遥感系统进行测量，量化大气状态和连续波，以准确地量化风和自由流湍流及其在一系列大气情况下的单、双和多尾流情况下的尖端高度（150 -200米）以上的剖面，并提供不同涡轮负载条件下尾流行为的详细数据。2)大型风电场多尾流相互作用的数据分析与建模。pi与许多风电场运营商合作，从五个大型陆上风电场获取数据集，这些数据集结合了规则和不规则阵列，可用于评估大型陆上风力涡轮机阵列的尾流行为。结合收集到的数据，该分析将用于量化功能依赖性，并开发包含涡轮机和大气参数的多个尾迹的模型参数化。3)建立新的多尾流模型。pi将开发一个新的模型，该模型是基于为单个尾流开发的数值尾流模型的扩展，并从分析模型中提取，以包括多个尾流相互作用。π吗?活动旨在鼓励风电场建模领域的广泛参与和科学严谨性：1)扩大印第安纳大学虚拟尾流实验室，为建模者提供风电场案例研究数据和时间序列，用于模型开发和评估。虚拟尾流实验室是一个基于网络的工具，它提供了可用于量化风力涡轮机尾流和评估尾流模型的数据集。2)与国际组织合作开发尾流模型基准，为风电场模型评估提供更好的指标，并在提供风电场最佳功率预测的过程中增加学术界和工业界的参与。3)利用最先进的模型和基于风电场的测量，与工业界合作，培训学生学习风力气象学。

1067007 PI BarthelmieCurrent generation wind farms being deployed in the US often contain hundreds of wind turbines with installed capacities in excess of 100 MW. Wind turbine wakes in these large arrays are responsible for reduction of total wind-farm power output by up to 20%. These wakes, which encompass the region of decreased wind speeds and enhanced turbulence behind wind turbines, also reduce turbine lifetimes due to increased fatigue loading. The PIs? previous research has shown that current generation wind-farm models underestimate the magnitude of wind-turbine wakes in large arrays. The main objectives of this project are (1) to improve the physical understanding and modeling of the development of single, double and multiple wakes in a range of wind speed, turbulence, and atmospheric stability conditions, and (2) to assess whether uncertainty in power prediction can be significantly reduced, and array configuration improved, by better quantification and modeling of wind-turbine wakes. The uncertainty in predicting power output from large wind farms can be substantially reduced by explicit modeling of the interaction between wind-turbine wakes, and between whole wind-farm wakes and the overlying atmosphere. The research will involve advanced measurement and modeling of the factors that dictate wind-farm efficiency, appropriate to the large scales of wind turbines and wind farms currently being deployed. The PIs will focus on the quantification of power losses and additional fatigue loading on downstream turbines due to wind- turbine wakes and comprises three parts: 1) Highly resolved measurements of wind-turbine wakes and associated atmospheric and turbine parameters using Doppler light detection and ranging (lidar). The PIs will conduct measurements in large wind farms using a remote sensing systems to quantify the atmospheric state and continuous wave to accurately quantity both wind and freestream turbulence and their profiles well above tip-heights (150 -200 m) in single, double, and multiple wake situations under a range of atmospheric situations and to provide detailed data on wake behavior under different turbine loading conditions. 2) Data analysis and modeling for multiple wake interaction in large operational wind farms. The PIs have partnered with a number of wind-farm operators to obtain data sets from five large onshore wind farms with a combination of regular and irregular arrays that can be used to evaluate wake behavior in large onshore wind turbine arrays. In conjunction with data collected, this analysis will be used to quantify functional dependencies, and develop model parameterizations of multiple wakes incorporating turbine and atmospheric parameters. 3) Development of a new multiple-wake model. The PIs will develop a new model based on an extension of the numerical wake model developed for single wakes and drawing from the analytical models to include multiple wake interactions.The PIs? activities are designed to encourage broad participation and scientific rigor in the field of wind-farm modeling by: 1) Expanding the Indiana University virtual wake laboratory to supply wind-farm case study data and time series for modelers to use in model development and evaluation. The virtual wake laboratory is a web-based tool, which supplies data sets that may be used to quantify wind-turbine wakes and to evaluate wake models. 2) Development of wake-model benchmarking in collaboration with international groups to provide better metrics for wind-farm model evaluation and to increase involvement from academia and industry in the process of providing optimal power prediction from wind farms. 3) Train students in wind-power meteorology in collaboration with industry using state of the art models and wind-farm based measurements.