A decision support framework for electric demand planning in distribution systems during extreme minimum temperatures

Records of underestimated electric demands during winter storms in the southeastern United States (US), like Uri and Elliot, contributed to national disasters, outages, and loss of lives and property during those temperatures when electricity was most needed. A regional utility organization in the Southeastern US reported that demand-related issues caused 97% of distribution system outages during Winter Storm Elliot in December 2022. If decision-makers and operators possessed timely knowledge of winter-peaking distribution networks in advance, they could have applied proactive mitigation measures to avoid several outages. Existing research emphasizes electric demand forecasts at system levels like cities and overall utilities. However, there is a notable gap in analytical frameworks that apply machine learning techniques for predictive modeling and proactive planning within distribution systems of US power grids targeted at minimum temperature planning. We propose a novel analytical and empirically validated decision-making framework that utilizes machine learning and statistical techniques for effective demand modeling and planning within distribution systems of applicable power grids.