Optimal short-term operation of a DisCo including voltage-sensitive loads

This paper presents a probabilistic approach for the short-term operation of a distribution company (DisCo) considering voltage-sensitive loads. The DisCo's operational decisions are optimized in a day-ahead operation planning to maximize the difference between the energy sold to consumers and the purchases from the wholesale electricity market and distributed generators. The network is modeled using ac power flow. The DisCo's model also includes technical and economic constraints. The technical constraints represent voltage-sensitive loads, stationary batteries, elastic demand, and physical limits. The economic constraints include the time-varying retail pricing constraints. The resulting model is a large-scale, nonlinear, mixed-integer program. Linearizations are not practical and, tractability and accuracy would be sacrificed. Instead, we propose a solution approach based on Tabu Search capable of handling nonlinearities associated with both the ac power flow model and voltage-sensitive loads. Numerical results from a 69-bus distribution system illustrate the effectiveness of this approach.