Control methods against voltage collapse based on real-time calculation of voltage stability indicators

Abstract

Power systems always face disturbances, which could lead to voltage collapse or blackout, and weaken with the increase of power demand or load, which is an everyday challenge to Transmission System Operators (TSOs). Nowadays, there aren’t specific methods to prevent voltage instability or collapse if a disturbance occurs in a system. The thesis presents a method to prevent voltage collapse occurrence by investigating voltage instability using indicators such as Impedance Stability Index (ISI), Voltage Stability Index based on Short Circuit Capacity (VSCIscc or VCSS or VSCC) and Voltage Collapse Proximity Index (VCPI), i.e. an optimal coordinated control actions for various controllable devices in real-time when disturbance occurs in a system is proposed, and the motivations are as follows : • obtain good information on the power system conditions, • act as fast as possible to increase the chances to save the system, • decrease operation and maintenance costs, • shed less loads The indicators were first evaluated in a 10-bus system from [1], in which both ISI and VSCIscc gave good indications of voltage instability and collapse, as opposed to VCPI, which proved to be inefficient. The ISI was selected for the next evaluation only due to the fact that the VSCIscc acted the same way and to limit computational burden. Afterwards, the ISI was tested in the Nordic32 power system with moderate loading [2], in which two different cases were tested to observe voltage collapse, consisting of applying a three-phase balanced fault in the first case, and tripping of a generator in the second one. High oscillations and transients were observed in the ISI indicator, which led to adding simple filter, assumed to be acceptable only for this thesis. Using the ISI, the model, which is built using a software called Power System Simulator for Engineering (PSS/E) and tested on the Nordic32, consists of control methods of available synchronous generators automatic voltage regulators (AVRs) in the system, taking into account the signals from the Over-excitation limiters (OELs). Consequently, the model identifies the weakest buses, and then increases the AVR set-points of the generators closest to these buses if ISI is greater than 0.5, and then by activating load shedding on loads at or closest to the weakest buses if AVR set-points of all generators were increased or if ISI is greater than 0.6. The voltages, angles, ISIs and OELs are considered to be input signals. The model was successfully verified in both cases, where voltage collapse was prevented. In the first scenario, only the AVR set-points generators closest to the weakest buses were increased. In the second case, with a more severe event compared to the first one, load-shedding was conducted on loads located at the buses of interest, along with the increase of AVR set-points, due to the fact that the ISI was greater than 0.6. Loads were shedded twice at buses 1044 and 1045 by 35% and a high voltage overshoot was observed at these buses.