Assessing the impact of weather uncertainty on power-load uncertainty and balancing characteristic in buildings with semi-transparent photovoltaic windows

Semi-transparent photovoltaic (PV) windows can simultaneously generate electricity and significantly reshape building cooling and heating loads. Although their energy performance has been widely investigated, the influence of weather uncertainty on the coupled power–load behavior of PV-window-integrated buildings remains poorly understood, and dedicated quantitative indicators for characterizing such coupling are still lacking. Against the backdrop of climate warming and increasingly complex, non-stationary meteorological conditions, understanding weather-uncertainty-induced source–load interactions is essential for the robust design and operation of building energy systems. To bridge this gap, this study develops a weather-uncertainty modeling framework and proposes two novel indicators to quantitatively evaluate power–load uncertainty and balance in buildings equipped with semi-transparent PV windows. Furthermore, the proposed approach is applied to multiple representative climate zones in China to systematically analyze power–load uncertainty characteristics and matching performance under diverse climatic and solar conditions. Results show that the weather uncertainty method proposed in this study is effective. The semi-transparent PV windows can reduce cooling load uncertainty by 2.46% - 61.73% and the magnitude by 12.05%- 75.06%. However, heating load uncertainty and magnitude may increase by 3.25% - 21.09% and 9.04% - 13.72%, respectively. Additionally, the probability that semi-transparent windows have a positive impact on the power-load balance ranges from 25.0% to 62.5%, primarily by reducing total building loads and increasing PV output. These findings contribute quantitative evidence and methodological insights to the uncertainty-aware evaluation and application of semi-transparent photovoltaic windows.