Long‐Term Comprehensive Risk Assessment of Distribution Transformers Based on Random Forests and Global Climate Models

With the increasing integration of distributed photovoltaics (PV), the risk profile of distribution transformers is becoming increasingly intertwined with meteorological factors. Meanwhile, global climate change has led to a rise in extreme weather events and greater interannual variability in meteorological conditions. As a result, traditional risk assessment methods, based on historical data, struggle to accurately predict future risks. To address this, this study proposes a long‐term comprehensive risk assessment method for distribution transformers based on forecasted meteorological data. Firstly, the insulation aging risk of transformers is introduced, with the insulation aging process quantified through the transformer's insulation lifetime. Secondly, a random forest model is employed to enhance the temporal resolution of global climate models predictions. Using these forecasted meteorological data, the load factor of distribution transformers is predicted, generating potential operational scenes for these transformers. On this basis, a comprehensive risk assessment of distribution transformers is carried out by considering insulation lifetime loss, load loss due to transformer outage, and photovoltaic curtailment loss, employing a typical day method for risk evaluation. Finally, a case study of a distribution transformer in Shanghai is conducted to validate the superiority of risk assessment using forecasted data. The study also analyses the impact of PV integration on the overall risk of distribution transformers.