Open-Source Modeling of Extreme Weather Impact on Distribution Networks

When appropriately integrated into smart grid technology, artificial intelligence (AI) has the potential of significantly improving the grid's efficiency, reliability, and resiliency. The development of advanced AI algorithms relies heavily on access to comprehensive datasets, which represents as the major challenge due to the proprietary and sensitive nature of grid-related data. Utility data are always of restricted access, requiring data-sharing agreements for specific purposes. In light of these limits, the use of synthetic network and data becomes a necessity for research, algorithm development, and performance comparisons. In this paper, a systematic modeling approach is proposed to acquire extreme weather conditions, to extract weather-prone features of synthetic grid models, and to develop a stochastic impact model of extreme weather on distribution grids. It is shown that synthetic weather conditions can be generated using standard distributions and their impacts on outages in distribution networks can be systematically analyzed and determined. The proposed methodology uses a set of fragility curves to describe the extent of weather impacts on physical features, which can be applied and validated in a straightforward way to real-world networks and extreme weather conditions once data of the actual physical networks and their past impact outcomes becomes available. The proposed methodology is provided as open source, ensuring accessibility and transparency for further research and application.