Probabilistic spatial load forecasting for assessing the impact of electric load growth in power distribution networks

Abstract

Recently, high rates of urbanization reveal the need for more granular approaches of long-term load forecasting. The development of modern spatial load forecasting (SLF) methods can contribute to more effective planning of power distribution networks, while serving as decision making tools in the hands of regional planners and distribution system operators (DSOs). Nonetheless, the ever-present error of spatial allocation of future electric loads arises the need for probabilistic assessment of the long-term point forecasts. This paper proposes a methodology of probabilistic SLF that can be employed to assess the impact of load growth in power distribution networks. Firstly, an hierarchical trend method forecasts the peak of the electric loads at each subarea of the study area. Subsequently, the spatial root-mean-squared error across the service territory is calculated to be used for the construction of prediction intervals. Finally, a probabilistic power flow study based on Monte Carlo simulation seeks potential technical issues on the distribution network. The proposed methodology is demonstrated on a real-world distribution network. The results prove that the proposed probabilistic SLF method can lead DSOs to more reliable decisions in comparison to business-as-usual approaches.