| 摘要: | 本文在戶外實測中壢及屏東地區的兩種雙軸追日太陽光電系統（PV和CPV）的電力性能，並使用PVsyst軟體模擬與實驗做比較。在中壢地區，本文比較了二種追日系統（太陽軌跡計算與光感測器）的PV性能，雖然太陽軌跡計算之追日系統的平均追日偏差角度為0.89o，相較於光感測器追日系統的平均偏差角0.3o大，但前者方式可穩定追日而不會受到天候不佳影響。反之，在天氣不佳時，光感測器易受影響而未能正常追日。太陽軌跡計算之追日系統的發電量仍略優於光感測器的追日系統，以十二天的實驗記錄期間，太陽軌跡計算之追日系統的每日平均最終產能為3.92 kWh/kW，而光感測器每日平均最終產能為3.83 kWh/kW。對於PV系統而言，發電量才是決定系統性能的最終指標，故太陽軌跡計算之追日控制仍為一實用且有效的追日方法。
本文藉由量測PV模組的電流電壓特性探討模組於戶外運作的實際性能，顯示在相同日照條件下，模組溫度上升會導致短路電流些微上升，開路電壓下降，導致PV模組整體發電量及效率都下降。
本文利用CPV模組搭配光感測器進行追日，當天氣條件不佳，太陽受到雲層遮蔽或是散射光較強的情況下，使光感測器無法正確感測到太陽位置，就會導致CPV系統輸出功率相當低或為零值。由於CPV模組接受角度僅有0.5o，且只能接受直射日照，因此使用CPV模組需搭配追蹤精度高的追蹤器，才能使其發揮最大效益。
由於日照量為直接影響系統發電量的主要因素，台灣南部日照和PV發電量都優於北部。本文比較相同追日控制（太陽軌跡計算追日）的同型雙軸PV追蹤器分別安裝在中壢及屏東地區的PV系統性能。在實驗期間（2012年12月至2013年5月），中壢及屏東地區的平均最終產能Yf分別為76 kWh/kW、95 kWh/kW，平均PR分別為0.85、0.9。屏東的追日PV性能高於中壢的追日PV。本文受限PV量測時間不長，因此目前的實測數據不足以論斷中壢及屏東地區的PV系統何者為佳。要評估任何一地的PV表現，需分析長時間（一年到數年）的Yf，才能評估一地是否適合安裝PV系統的參考。
由於氣象資料是最直接影響模擬PV發電量的因素，本文利用兩種不同的氣象年資料模擬中壢及屏東PV系統電力性能，發現利用真實年氣象資料模擬屏東雙軸PV系統的模擬誤差較小，而利用平均年概念模擬中壢PV系統電力性能，誤差較大。因此必須考慮適當的氣象資料形式對PV性能模擬影響。
This study performed outdoor test of performance of the PV/CPV system in two areas (Jhongli and Pingtung), and PVsyst software was used to compare the measurements. In Jhongli, comparison of PV performance between two dual-axis sun-tracking systems (ephemeris calculation and light sensor) were made. The average offset angle of PV system based on ephemeris calculation was 0.89o, which is larger than that of light sensor (average offset angle was 0.3o). Although the former approach is less accurate than the approach of light sensor, but it can stably track the sun and does not affected by the weather. Meanwhile, the light sensor is easy affected by bad weather which prevent it function normally. Yet, the electricity output of the tracking PV system with ephemeris calculation is still better than the PV system using light sensor. For any PV system, electricity is the ultimate factor which decides the performance of PV. Thus, the approach of sun-tracking based on ephemeris calculation is a practical and effective sun-tracking method.
Results of I-V characteristic measurement of PV modules showed realistic performance of modules in outdoor tests. Under the same insolation, the increasing of module temperature causes rise of the short-circuit current and decrease of the open-circuit voltage, these decreased the overall electricity output and efficiency.
The CPV module uses the light sensor to track the sun. In bad weather condition, either clouds block the sun or strong diffuse irradiation will prevent the sensor to detect the sun, which degrade the power output of CPV module. Since the acceptance angle of CPV module is only 0.5o, and it only accept direct irradiation, therefore the CPV module have to use high accuracy tracker and order to maximize its effectiveness.
Since the irradiation directly affects the electricity production, the southern Taiwan has higher irradiation and PV electricity than that of the northern Taiwan. This study uses the same tracking PV with ephemeris calculation to install two system in Jhongli (northern Taiwan) and Pingtung (southern Taiwan), respectively. In the period of experiment (December 2012 to May 2013), the average final yield (Yf) for Jhongli and Pingtung are 76 kWh/kW and 95 kWh/kW, respectively. The average performance ratio (PR) for Jhongli and Pingtung are 0.85 and 0.9, respectively. The performance of tracking PV installed in Pingtung was better than that installed in Jhongli. This study was limited by short period of outdoor measurement of PV, and present experiment data was insufficient to decide the performance of both areas. To properly assess the behavior of PV in any location, Yf of longer time frame (one to several years) should be analyzed.
Due to the meteorological data directly affected the PV electricity production, this study uses two kinds of the meteorological data to simulate the electrical performance of the PV system in Jhongli and Pingtung. It was found that less error was obtained when using the real year meteorological data to simulate the PV system in Pingtung. As for the case of Jhongli, the error is smaller when using the average year meteorological data. Therefore, considering the appropriate type of meteorological data in the simulation of PV performance is important. |
