PREVENTION OF CASCADING OUTAGES USING SPARSE WIDE AREA SYNCHROPHASOR MEASUREMENTS

Abstract

Power System cascading outages can result in major system loss and the other indirect losses. These outages are unpredictable in nature and hence their modeling is a fairly complicated task. The modeling of these outages is based on statistical inferences. These outages can start from very simple outages and can grow into a big outage if some chaotic parameter is touched. They can also start from inter tie oscillations and can grow into big outages. The health of the power system is represented by frequency and whenever there are some oscillations which are small in frequency then that means some dynamic is there. If these don’t get damped out, then there is a chance that they will grow and that will definitely cause system wide cascade if some critical equipment failure occurs. In this report a small part of a big problem of preventing cascading outage is discussed. Under this, the fault location needs to be found for real time topology update of the system network. This ensures that if some critical state is present, it can be sensed with the help of generator coherency information. The method uses the electromechanical wave velocity concept and the concept of multilateration to determine the fault location. The system on which this approach is tested is IEEE- 9 bus system. The method to detect the fault location does require the use of PMU and IED which can be used for measurement of signals. With very less amount of data and less number of PMUs, but sufficient enough to make the overall power system observable, fault location can be calculated. The report only focuses on the fault locations at the buses, but in actual system fault could also occur between the buses and the method can be modified to include line faults. Also after determining the location of fault, system data can be modified with new calculations. This modified data can be used further to estimate the state of the system. The problem of state estimation has not been dealt with in this work. It is assumed already that the system is in vulnerable state or in emergency state and different areas can lose synchronism with each other which will result in different generators falling apart. In order to save the system, different buses and generators which are coherent need to be grouped. This is done by studying the coherency information of the system and finally the system is studied for optimum isolation under such condition (optimum in terms of minimum power flow disruptions).