Co-otimização de fluxo de potência ótimo dinâmico baseado em preço com agendamento de reparos

Abstract

The main objective of the electric distribution system operators is to supply energy to the loads at the lowest generation cost over a given time horizon, anticipating sudden changes in demand and generation, a problem known as Dynamic Economic Dispatch (DED). When coupling the DED objectives to the network operational constraints, such as voltage levels and load limits of the lines, we have the problem known as Dynamic Optimal Power Flow (DOPF). In an emergency scenario, with damaged lines or equipments, the healthy regions of the network with Distributed Energy Resources (DERs) can operate islanding from the main grid, using MicroGrid (MG) concepts. In this scenario, the demand constraint is flexible and the DED follows a price signal, using inter-temporal technologies to maximize the power utility’s profit; with the DOPF being known as Price-Based DOPF (PB-DOPF). Due to limited generation or high attendance cost, load shedding can be performed in specific moments. Simultaneously with the operation of MGs, repair crews must be dispatched to affected locations to repair and restore the system to its normal operating state, which is known as emergency order scheduling (EOS) problem. Although PB-DOPF and EOS problems are interdependent in real-world applications, due to the associated high computational complexity, they are traditionally addressed in decoupled manner by utilities. Thus, this work proposes a mathematical model for the co-optimization of these problems in order to minimize the cost of PB-DOPF in a horizon. The PB-DOPF with EOS is formulated as a Mixed-Integer Linear Programming (MILP) model, with a three-phase power flow. The linear modeling offers, among other advantages, a high-speed calculation and allows to know the quality of the solution, critical factors in emergency scenarios. An algorithm is also proposed to perform the PB-DOPF and the repair crew dispatches over a given horizon, in both normal and emergency scenarios, integrating a pre-scheduling step to deal with computational complexity. The route is systematically released at each iteration and the evolution of the solution dynamically monitored. The proposed model is tested on three-phase unbalanced IEEE 123-bus and IEEE 8500-node distribution systems, both modified with additional DER systems. The algorithm is also compared with two variations, with the results highlighting its effectiveness.