A hierarchical decentralized coordinated voltage instability detection scheme for SVC

A number of techniques have been proposed to detect voltage instability in power systems. Most of these techniques are either applied to fast local controls using local measurements or to centralized control using wide-area measurements. While the local detection schemes do not have sufficient situational awareness to suitably respond during a cascading failure, wide-area detection schemes overcomes that weakness but at the expense of speed. In the present work, we propose a decentralized voltage instability detection method that presents a tradeoff between speed and situational awareness. It suits large-scale system having both discrete and continuous control devices, with the former managed centrally and the latter managed locally. The technique consists of finding weakly coupled coherent areas in the system and then applying robust methods to estimate a multiple linear regression model that relates the reactive power reserves to the voltage stability margins in each coherent area. These models are then used to perform voltage stability analysis and to carry out secondary voltage control (SVC) in real-time. Simulation results performed on IEEE test systems reveal that our method has a better situational awareness than simple local detection, especially in case of cascading failures, and requires lesser communication bandwidth than a completely centralized detection based on wide-area measurements. This hierarchical decentralized coordinated scheme will make the grid more robust and resilient.