再生能源應用在現今溫室效應造成環境氣候變化的嚴重衝擊之下,益發突顯其重要性,尤其能源多元化已成為世界各國能源政策努力的目標。近年來,小型分散式發電系統已廣泛設置利用,且數量逐年成長,例如,電力儲能設備、太陽能電力系統與風力發電系統均已有效與市電系統併聯運轉。 所謂孤島運轉現象係指分散式發電系統與市電系統並聯供電時,當市電系統發生故障或因維修與分散式發電系統解聯,但分散式發電系統未即時檢知並切離市電系統,而呈現獨立供電的現象。當發生孤島運轉現象時,孤島運轉監測系統應立即監測出此狀況並且立即停止供應電力予用戶端負載,以免造成危害。 孤島現象偵測技術主要可分為被動式與主動式偵測法兩大類,由文獻測試結果顯示,上述方法對於某些特定負載仍然存在「孤島現象不可偵測區域」。因此,本文提出電壓差分相關性法以克服上述孤島現象監測法之缺失。電壓差分相關性法係利用反流器輸出週期性擾動脈波電流,並且監測負載端電壓峰值波動與週期性脈波電流之相關性,本文藉由所提之方法建構一孤島效應監測指標,以利分散式發電系統孤島運轉診斷。 為驗證所提分散式風力發電系統孤島現象偵測法之可行性,本文利用MATLAB套裝軟體模擬系統架構,並進行不同干擾測試,其中包括併聯運轉與孤島運轉測試,此外,針對瞬間風速改變與電力系統各種電力品質干擾訊號情形,本文亦進行模擬測試。經由數值結果證實,本文所提之孤島現象檢測法能有效縮小孤島現象不可偵測區域並且能提高孤島現象診斷準確率。
Renewable energy has been drawing more attention since the advent of global warming effects. Multiple energy sources have currently become one of the goals for the energy policy of the developed countries around the world. Over the last decades, an increased percentage of the installation of dispersed mall-scaled generators, such as battery energy storage systems, photovoltaic systems, wind generation, and cogeneration, have been operated effectively in parallel with utility networks. Islanding operation is a situation in which the dispersed power system keeps supplying electric power to a distribution network even when the utility grid loses power due to fault or maintenance. When the islanding operation occurs, the islanding detection systems should detect it and disconnect from the grid immediately. The existing of the categories of islanding detection methods, are generally classified into passive and active techniques. Nevertheless, it has been demonstrated that “non-detection zones” still exist for some specific load impedances. Hence, a new algorithm using differential voltage correlation method is proposed in this paper to alleviate this problem. The DVC method injects a perturbing periodic signal into the inverter output current to the power system and monitors the correlation function between the periodic signal and the perturbed voltage at the connection point. Through the proposed method, an islanding detection index can be evaluated that serves as a useful reference to activate the protective relays. To verify the proposed approach for islanding detection of dispersed wind power generation systems, the method has been implemented in Matlab programming tools. This approach has been evaluated under different scenarios, which include both loss of grid interconnection and parallel operation. Impacts of wind speed abrupt change and power-quality disturbances on the detection performance are also examined. From the test results, the effectiveness of the approach to reduce non-detection zone and improve the diagnosis accuracy has been demonstrated.