Open AccessFast network restoration by partitioning of parallel black start zones
Author(s) -
Shen Cong,
Kaufmann Paul,
Braun Martin
Publication year - 2017
Publication title -
the journal of engineering
Language(s) - English
Resource type - Journals
ISSN - 2051-3305
DOI - 10.1049/joe.2017.0032
Subject(s) - blackout , computer science , process (computing) , task (project management) , electric power system , transient (computer programming) , power (physics) , state (computer science) , voltage , sequence (biology) , reboot , real time computing , reliability engineering , algorithm , control theory (sociology) , engineering , electrical engineering , artificial intelligence , physics , systems engineering , control (management) , quantum mechanics , biology , genetics , operating system
The restoration of large electrical power systems after a blackout is often a challenging task. A typical restoration process includes the partitioning of the power systems into subsystems and the successive booting of large non‐black start (NBS) units by black start (BS) units. A proper partitioning of the starting zone can reduce the restoration time significantly. This paper investigates a novel network partitioning algorithm to improve the restoration time and ratio of generation and load in each subsystem. The proposed algorithm consists of three stages. In the first stage, the number of subsystems is determined by the number of available BS units and their electrical distance. In the second stage, NBS units are assigned to each subsystem in the way that the rebooting time difference among subsystems is minimised. In the third stage, the substations are assigned to one of the subsystems to achieve the optimal ratio of generation and load in each subsystem. With the proposed algorithm, the switching transient and steady‐state over‐voltages at the receiving end of unload lines are kept within acceptable ranges and the self‐excitation phenomenon does not occur in the subsystems. Furthermore, the start‐up sequence of NBS units in each subsystem is determined simultaneously. The proposed algorithm is flexible and can be adjusted very easily according to the real status of the power system. The validity and performance of the proposed approach is demonstrated through simulations using a New England 39 Nodes network and a real network from south China.