Changes in power system operating conditions cause dynamic changes in angle and frequency. These disturbances propagate throughout the system area with finite speed. This propagation takes the form of a traveling wave whose arrival time at a particular point in the system can be observed using a wide-area measurement system (WAMS). Observations of these waves both through simulation and measurement data have demonstrated several factors that influence the speed at which a disturbance propagates through a system. Results of this testing are presented which demonstrate dependence on generator inertia, damping and line impedance. Considering a power system as an area with and uneven distribution of these parameters it is observed that a disturbance will propagate throughout a system at different rates in differing directions. This knowledge has applications in locating the originating point of a system disturbance, understanding the overall dynamic response of a power system, and determining the dependencies between various parts of that system.

A simplified power system simulator is developed using the swing equation and system power flow equations. This simplified modeling technique captures the phenomenon of traveling electromechanical waves and demonstrates the same dependencies as data derived from measurements and commercial power system simulation packages. The ultimate goal of this research is develop a methodology to approximate a real system with this simplified wave propagation model. In this architecture each measurement point would represent a pseudo-bus in the model. This procedure effectively lumps areas of the system into one equivalent bus with appropriately sized generators and loads. With the architecture of this reduced network determined its parameters maybe estimated so as to provide a best fit to the measurement data. Doing this effectively derives a data-driven equivalent system model. With an appropriate equivalent model for a given system determined, incoming measurement data can be processed in real time to provide an indication of the system operating point. Additionally as the system state is read in through measurement data future measurements values along the same trajectory can be estimated. These estimates of future system values can provide information for advanced control and protection schemes.

Finally a procedure for the identification and extraction of inter-area oscillations is developed. The dominant oscillatory frequency is identified from an event region then fit across the surrounding dataset. For each segment of this data set values of amplitude, phase and damping are derived for each measurement vector. Doing this builds up a picture of how the oscillation evolves over time and responds to system conditions. These results are presented in a graphical format as a movie tracking the modal phasors over time. Examples derived from real world measurement data are presented.