A DEGRADATION ANALYSIS MODEL OF LONG-TERM OUTDOOR EXPOSURE FOR TEN CRYSTALLINE AND THIN-FILM MANUFACTURERS UNDER SIMILAR CONDITIONS FOR 1 YEAR AND COMPARISON TO MEASUREMENT DATA

摘要

An outdoor testing facility with ten crystalline and thin-film manufacturers was setup by SB Energy Corporation in Obihiro, Hokkaido, Japan in 2011. The resultant data and information about array construction,measurement and setup was made freely available and is streaming in real-time online. In 2012 for a period of one-year data analysis was evaluated in this paper. A mathematical model was made to predict the amount of energygeneration in kWh for one-year of data based on New Energy and Industrial Technology Development Organization (NEDO) public irradiance data. This model was checked with the actual output for each array in kWh inHokkaido, Japan and Sydney, Australia. The purpose was to increase understanding of module performance in harsh environments and to increase the accuracy of assumption methodology for predicting yields of eachtechnology. Daily, monthly and yearly yield data was collected and analyzed from all ten photovoltaic arrays representing all ten manufacturers. These arrays varied in size between 3-5kWp of crystalline silicon, amorphoussilicon tandem HIT (Hetro-junction with intrinsic Thin layer), CIS (Copper Indium Selenium) and a-Si/μc-Si (amorphous silicon/micro-crystalline silicon tandem). Parameters such as kWh/kWp (energy production/peak ratedpower), ambient temperature, and irradiance were reported. New information was gathered about seasonal fluctuation of module kWh/kWp output and degradation factor in low temperature, high snow load environments and thecomparison of real measurement data vs. the model were checked. This is directly relevant for system integrators, designers for increasing energy yield and economic payback for large-scale photovoltaic plants. It is importantfor investors to check if their financial and energy yield modeling can be improved by incorporating the Japanese industrial standard methodology and by using more accurate de-rating factors, particularly in harsh high- snowload environments.