Coordinated Restoration of Integrated Gas-Electricity Distribution System With Dynamic Islanding: A Multi-Stage Stochastic Model With Nonanticipativity

With the sharp growth of extreme events and the tight connection between the natural gas and electricity systems, it is imperative to co-optimize the two energy systems after natural disasters. This paper proposes a multi-stage stochastic restoration model of the integrated gas-electricity distribution system (IGEDS) considering the nonanticipativity requirements of uncertain extreme events. The proposed model minimizes the total operational cost during the restoration process while considering repair crew scheduling, reconfiguration of the power distribution system, primary frequency response (PFR) of distributed generators (DGs), and gas and electricity demand response (DR). In addition, dynamic islanding via topology adjustment is proposed in this paper, which could effectively utilize the PFR reserve of DGs to supply critical loads while guaranteeing frequency stability. Moreover, uncertainties of fault branches in different stages and energy generation/consumption are taken into account to ensure nonanticipativity and all scenario feasibility. To solve the proposed model, a customized progressive hedging (PH) algorithm with improved iteration criteria is presented. Numerical results show that the proposed model could effectively improve the restoration process against extreme events.