Day-ahead dispatch of distribution feeders considering temporal uncertainties of PEVs

This paper presents an approach to dispatch taps of Load Tap Changing (LTC) transformers, switched capacitors and Plug-in Electric Vehicle (PEV) charging in distribution feeders to minimize feeder daily peak demand, using a nonparametric Bootstrap technique, an alternative to Monte Carlo Simulations (MCS), to account for the PEV charging temporal uncertainties. From an initial sample of independent observations generated using a deterministic Genetic Algorithm (GA)-based optimization framework, Bootstrap samples are generated, which yield an estimate of the mean daily system peak demand, and the hourly tap, capacitor and PEV charging schedules. The proposed technique is applied to a distribution feeder model of an actual primary feeder in Ontario, considering a significant PEV charging penetration level. The results for an actual distribution feeder show the feasibility of the proposed approach, with a significant reduction of computational burden with respect to an MCS approach while still using a global search technique, which yields adequate tap and capacitor daily schedules for a Local Distribution Company (LDC) that properly accounts for PEV charging uncertainties.