以致動盤模型預測風場功率

Abstract

本研究以台灣三處陸域風場的長期風況量測值為基礎，統計風向、風速和機率的關係，藉由統計結果定義特徵風況，並使用風機致動盤模型程式WIFA3D求解三維穩態連續方程式、動量方程式及k-ε 紊流模型，獲得風場中特徵風況下的功率輸出。接著，使用特徵風況計算三處風場在長期風況下的功率輸出與容量因子，並與實測值比較驗證。由於風場實測資料中缺乏獨立測風塔的資料，僅有各風機的風速與風向實測值，因此在每段時間間隔中以各風機的實測風速與風向值決定實際入流風況進行功率預測。計算結果顯示，本研究預測方法的風場容量因子計算值在前兩座風場中與實測值大致相符，唯第三座風場因實測功率曲線明顯偏離設計功率曲線，因此風場容量因子計算值與實測值差異較大。為確認本研究預測方法的合理性，使用第三座風場的實測功率曲線作為修正風機功率曲線的依據，發現修正後的容量因子預測值與實測值相近，說明本研究提出的方法在預測正常操作狀態下的風場功率具有一定的精確度。

This study predicts the power characteristics of three onshore Taiwanese wind farms based on the wind direction distribution and wind data from long-term meteorological measurements in those wind farms, where typical wind condition is then defined from the statistical result of measurement. To calculate the power output under typical wind condition in the wind farm, a three-dimensional flow field is described by the steady continuity and momentum equation coupled with k-ε turbulence model, which is solved by an actuator disk model code, WIFA3D. The power output of long-term SCADA data in the wind farm is then predicted with a method proposed in this study and compares with the measured wind farm power output. In the proposed approach, inflow condition for a specific time span is determined by the measured data of wind turbines according to the incoming wind direction. The predicted result shows that the averaged capacity factor of the first two wind farms via the proposed method is close to the measurement while substantial error occurs in the third wind farm since the measured wind turbine power deviates considerably from the design power curve. For confirming the accuracy of the prediction method, the wind turbine power curve in the third wind farm is then corrected according to the measured power characteristics. The corrected capacity factor is in good agreement with the corresponding measurement, which suggests that the proposed approach has a satisfactory accuracy in predicting wind farm power behaviors.