Validation, testing and implementation of the linear state estimator in a real power system

Thesis (Ph.D.), School of Electrical Engineering and Computer Science, Washington State UniversityWith the advent of phasor measurement units (PMUs), the synchro-phasor data related applications at control centers have been researched in recent decades. Time-stamped phasor data can benefit the applications in modern power system, such as state estimation, voltage stability, oscillation monitoring, transient stability and so forth. As one of the key applications in Energy Management System (EMS), state estimation utilizing synchro-phasor data is the major objective of this dissertation. As more Phasor Measurement Units are installed, portions of the power grid become observable with just phasor measurements making feasible the estimation of the state of these observable portions. This linear state estimator (LSE) can have the same periodicity as the PMU rate which is much faster than the traditional state estimator using SCADA measurements. As part of the Western Electric Coordinating Council (WECC) smart grid project design and implementation, validation of such data as well as preparing end users for various scenarios (e.g. loss of PMU signal and its impact) is a necessity. One of the WECC companies has chosen to integrate the Linear State Estimator (LSE) into their EMS. One of the business use cases of LSE is the validation of PMU data. It is important that system dispatchers trust LSE from day one, as much as they do their SCADA-based state estimator. In this paper, we present the design, development and implementation of an LSE that can estimate the state of a portion of the Extra High Voltage (EHV) network of the Western Interconnection. The PMU data from all of the substations that are observable in this portion of the network, are received at the phasor data concentrator (PDC) at the control center and the LSE utilizes the power system data from the Energy Management System (EMS) data base to solve 30 times per second. This integration of the LSE to the existing EMS environment is presented in the paper, and so are some of the off-line and on-line test results. The testing of the LSE application has been carried out with different data sets: simulated steady state data, simulated RTDS data and field PMU data. Various conditions of testing of the LSE successfully demonstrate the correctness of the LSE algorithm and provide the useful practical information of integrating LSE application. All the implementation experience is also helpful on how to integrate "smart grid" applications for the future power grid.School of Electrical Engineering and Computer Science, Washington State Universit