Power System Dynamic Modeling and Synchrophasor Measurements

摘要

Electric power is fully interwoven into the fabric of American life. Its loss for extended periods has profound impacts upon public safety, health and welfare. The power system has been termed the most complex machine built by man. Not surprisingly, the measures to address the range of power system downtime causes are as diverse as the causes themselves. One important arc of effort is providing power system operators with full knowledge of the system's operating state, timely warning when changing conditions threaten system stability, and tools guiding control actions to maintain stable operations.;This research is motivated, in part, by the need to explore opportunities for leveraging nascent synchrophasor data streams against known power system stability challenges. Over the past half-decade, power system operators have aggressively installed large networks of phasor measurement units (PMUs) and phasor data concentrators (PDCs) across the United States and Canada. Today, the synchrophasor data network has reached a state of maturity that opens the door to useful application.;This dissertation investigates power system stability along three lines of effort. The first two efforts address steady-state power system stability---specifically methods for assessing system vulnerabilities arising from the phase angle difference between two buses connected by a transmission line. The third effort investigates the information that can be gleaned from synchrophasor measurements during a system's dynamic system response to changing system conditions.;The first line of investigation extends steady-state distribution factor theory. Distribution factors are computed from a known non-linear power system load flow solution. They provide a computationally light method for estimating new system conditions under different operating circumstances. Distribution factors are extremely useful for very rapidly screening the impact of unexpected changes in power system configuration---e.g. a transmission line dropping out of service due to environmental conditions. The Line Outage Angle Factor (LOAF) developed herein provides a method for fast computation of bus voltage angle changes after a line outage. The Line Outage Generator Factor (LOGF) modifies the simulated circuit topology to include synchronous machine transient reactances, enabling rapid screening of operating states in which line opening (or re-closure) risks damaging equipment. The LOAF and LOGF provide promising results in MATLAB simulation of the Western System Coordinating Council 3-Machine, 9-Bus System.;The second investigative line seeks to develop a Thevenin equivalent model to be used in tandem with synchrophasor data streams to provide real-time bus angle difference information for buses connected by a transmission line. The appeal is that real-time bus angle difference information could be computed on-site and very rapidly---and significantly, independent of other network bus measurements. The results show that developing a Thevenin equivalent that provides a useful angle difference measure often works well on paper, but is challenging using actual synchrophasor data. Efforts to develop a Thevenin equivalent using Monte Carlo methods show promise, but require further investigation.;The third line of effort shifts to investigate the useful information that a PMU can produce during a power system disturbance event. A synchrophasor is defined at a specific frequency, i.e. the system steady-state operating frequency. Thus a PMU produces a data stream recording power system changes progressing slower than the nominal system frequency; consequently, this is an "off-label" synchrophasor data application. The test system is a generator with electrical and mechanical controls connected by a pair of identical transmission lines to an infinite bus. The synchronous generator is modeled as a three-damper-winding synchronous machine. A MATLAB simulation was written to simulate both the full 14 dynamic state and the reduced order 11 dynamic state system models. A Real-Time Digital Simulator (RTDS) simulation emulating the test system provides the capability to produce real-time analog generator terminal waveforms to be sampled by a commercial off-the-shelf PMU to produce synchrophasor data. We find that the RTDS generated synchrophasor data stream is similar to the MATLAB reduced order model voltage and current generator terminal data in the dqo reference frame---reflecting parallel, but distinct, filtering processes.