The short-term characteristics of utility-scale PV variability become increasingly important for power system operation as PV penetrations grow. However, understanding how these characteristics and their aggregated impacts will change with new PV deployments is challenging given the limited and highly site dependent availability of high-resolution PV output data. This study proposes a methodology to generate a synthetic 4 s utility-scale PV output time series for a PV plant anywhere in a power system using hourly satellite-derived irradiance data, which is globally available, to select a set of 4 s output from models created from four PV plants operating within the Australian national electricity market (NEM). The method resamples 4 s clear sky PV output index variability from statistical distributions which are binned according to hourly clear sky index pairs of global horizontal irradiance and direct normal irradiance. The method is cross-validated against the observations from the four utility PV plants located in different climate zones. The monthly Kolmogorov–Smirnov Integral (KSI) tests on the modeled variability distributions show that they are not statistically different from the observed time series with most KSI values remaining under 80%. Finally, the method is applied to all 20 utility PV plants that were registered in the NEM as of 2019. The modeled result shows good agreement with the measured aggregated 4 s variability. Hence, our method can be usefully applied for modeling the short-term variability of future power system scenarios with high PV penetrations if at least some existing utility PV plant generation data are available.

中文翻译：

---

根据每小时太阳辐照度数据生成合成的4 s实用规模的PV输出时间序列

随着PV渗透率的提高，公用事业规模PV变异性的短期特征对于电力系统的运行变得越来越重要。但是，由于高分辨率PV输出数据的可用性有限且高度依赖于现场，因此了解这些特性及其总体影响将随着新PV部署的变化而变化是具有挑战性的。这项研究提出了一种方法，该方法使用每小时可获得的全球卫星小时照度数据为电力系统中任何地方的光伏电站生成合成的4 s实用规模的PV输出时间序列，从模型中选择一组4 s的输出由在澳大利亚国家电力市场（NEM）内运营的四个光伏电站创建。该方法根据统计分布重新采样4 s晴空PV输出指数的变异性，这些统计分布是根据全球水平辐照度和直接法向辐照度的每小时晴空指数对进行分组的。该方法针对位于不同气候区的四个公用事业光伏电站的观测结果进行了交叉验证。对模型变异性分布进行的每月Kolmogorov–Smirnov积分（KSI）测试表明，它们与观察到的时间序列在统计学上没有差异，大多数KSI值保持在80％以下。最后，该方法适用于截至2019年在NEM中注册的所有20个公用事业光伏电站。建模结果与实测的4 s变异性显示出良好的一致性。因此，