Wide-Area Monitoring of Large Power Systems Based on Simultaneous Processing of Spatio-Temporal Data

Accurate identification of electromechanical oscillations on power systems and determination of its stability condition is a fundamental process in order to carry out an appropriate control action to prevent the partial loss or complete blackout of the system. However, the non-linear characteristics of measured variables often lead to incorrect information about the development of the electromechanical oscillations, making wide-area monitoring a challenging task. In addition, significant amount of information in extra large power systems is produced, which has to be stored on local servers requiring large amounts of central processing unit (CPU) storage. For these reasons, algorithms for Big Data problems in power systems are required and the methods presented on this chapter introduce some potential solutions. In this context, different data-driving methods based on spectral analysis of linear operator are presented for the analysis of electromechanical oscillations from a spatio-temporal perspective. These algorithms have the ability to process spatio-temporal data simultaneously, making possible to characterize inter-area and global oscillations (from 0.1 Hz to 1.0 Hz). To validate the effectiveness of the proposed approaches, two test systems with different structural and generation capacities are analysed: the Mexican Interconnected (MI) system and the initial dynamic model of Continental Europe from ENTSO-E. First, data collected from a transient stability study on the MI system are used to illustrate the ability of data-driving methods to characterize modal oscillations on longitudinal systems; where several inter-area modes produce interactions of different electrical areas. Then, simulation results from the initial dynamic model of ENTSO-E are analysed to characterize the propagation of its global electromechanical modes across Europe, which have been denominated as the North-South and East-West modes with frequencies of approximately 0.15 Hz and 0.25 Hz, respectively. The second analysis include the interconnection of Turkey (TR) to Continental Europe in December 2010, which derived on the grow of size and complexity of the original system having as result a decrease in the frequency value for the East-West mode and the introduction of a third inter-area mode on the system. The chapter concludes comparing the results of the proposed approaches against conventional methods available in the literature.