A predictive out-of-step protection scheme based on PMU enabled distributed dynamic state estimation

Recent widespread blackouts have indicated the need for more efficient and accurate power system monitoring, control and protection tools. Power system state estimation, which is the major tool that is used nowadays for providing the real-time model of the system, has significant biases resulting mainly from the complexity and geographic spread and separation of an electric power system. Synchrophasor technology is a promising technology that has numerous advantages compared to conventional metering devices. PMUs provide synchronized measurements, where synchronization is achieved via a GPS clock which provides the synchronizing signal with accuracy of 1 μsec. As a result, the computed phasors have a common reference (UTC time) and can be used in local computations, thus distributing the state estimation process. The first part of the work presents a PMU enabled dynamic state estimator (DSE) that can capture with high fidelity the dynamics of the system and extract in real time the dynamic model of the system. The described DSE is performed in a decentralized way, on the substation level based on local measurements which are globally valid. The substation based DSE uses data from relays, PMUs, meters, FDRs etc in the substation only, thus avoiding all issues associated with transmission of data and associated time latencies. This approach enables very fast DSE update rate which can go up to more than 60 executions per second. The distributed state estimation architecture that synchrophasor technology enables, along with the fast sampling rate and the accuracy of the measurements that PMUs provide, enable the computation of the real-time dynamic model of the system and the development of numerous power system applications for more efficient control and protection of the system. In the second part of the work, a transient stability monitoring scheme is presented that utilizes the information given by the dynamic state estimation and enables real-time monitoring of the transient swings of the system and characterizes the stability of the system in real time. In particular, the real-time dynamic model of the system, as given by the DSE, is utilized to evaluate the system's energy function based on Lyapunov's direct method and extract stability properties from the energy function. The two major components of the scheme are a) the calculation of the center of oscillations of the system and b) the derivation of an equivalent, reduced sized model which is used for the calculation of the potential and kinetic energy of the system based on which the stability of the system is determined. Finally, as an application of the transient stability monitoring scheme, an energy based out-of-step protection scheme is proposed. The energy of the generator is continuously monitored and if it exceeds a predefined threshold then instability is asserted and a trip signal can be sent to the generator. The major advantage of the scheme is that the out-of-step condition is predicted before its occurrence and therefore relays can act much faster than today's technology. The scheme is compared to presently available state of the art out-of-step protection schemes in order to verify its superiority.PhDCommittee Chair: Meliopoulos, A.P. Sakis; Committee Member: Deng, Shijie; Committee Member: Grijalva, Santiago; Committee Member: Harley, Ronald; Committee Member: Zhang, Fumi