Island partitioning is an interesting solution for service restoration in the emergency operation of the smart distribution system. Since rapid decisionmaking is required in an emergency condition, the island partitioning without regard to the time allowed for the implementation of control actions may lead to the violation of the allowable frequency limits and the activation of frequency relays. Therefore, this paper presents a two-stage solution procedure to restore the smart distribution system in emergency condition considering the frequency stability, operational constraints, the position of switches, the load priority and controllability, the controllable and uncontrollable distributed generations, and the time delay in implementation of control actions. In the first stage, considering the frequency stability, the time interval for the implementation of control actions is determined. Then, considering this time interval, the load/generation changes and the boundaries of the partitions are determined, using the partitioning algorithm based on black widow optimization, in order to maximize the value of the restored loads/ minimize the operational cost of DGs. In the second stage, the formed partitions are adjusted to satisfy the operating constraints of the system. Applying the proposed method to the PG&E 69-bus test system, considering different scenarios, and comparing the proposed method with the previous methods show the efficiency of the proposed method.

Island partitioning is a solution to restore service in smart distribution systems after the disconnection of the main grid due to faults. A two-stage and multi-objective solution is presented to restore the services of the island system. At first stage, the proposed partitioning scheme is presented with regards to the available time to decide on control measures, considering the voltage stability and the priority and controllability of the loads in the load shedding, and considering the frequency response and the voltage stability margin in finding the borders of the partitions. In the second stage, optimal power flow is done to set the partitions and satisfy the system operational constraints. The results show the proposed method not only leads to the creation of partitions with the allowable bus voltage and frequency but also leads to the lowest losses, the largest amount of restored load, and supplying loads with the highest values. Moreover, proposed method is more comprehensive due to considering voltage stability, frequency stability, occurrence of simultaneous faults, existence of controllable and uncontrollable distributed generation, power reserve, controllability of loads and DGs, position of switches, maximum time available to make decisions about control measures and the time delay between control measures