Implementation of a Probabilistic Power Flow System at Eversource Energy

Complexity on distribution systems is increasing with the rapid adoption of new technologies, uncertain forecasts of localized adoption and dispatch behavior, customer optimization through co-sited storage, and the introduction of resource aggregation for market participation through FERC Order No. 2222. With many degrees of uncertainty, developing stand-in model scenarios for future grid conditions is an increasingly inadequate method to address growing complexities and provide an accurate representation of future system conditions. Probabilistic power flow modeling improves distribution modeling capabilities and supports investments in enhancing necessary data analytics solutions. The need can be summarized in two key points: the ability to use probability distributions of uncertain input parameters to run a probabilistic supported simulation model in a probabilistic power flow model, and the technical capability to evaluate the vast number of results using advanced data science, machine learning, deep learning, and decision-making processes (through visualization). This contribution will introduce a comprehensive environment based in MATLAB^® that successfully supports large-scale processing, visualization, and advanced analytics for massive amounts of power load flows.