Enhanced Power System Monitoring and Analysis Using Synchrophasor Technology

Abstract

Electric power system is the main source of energy for all major industrial applications and residential appliances. The power system has been rapidly evolving due to transformation from regulated to deregulated structure, increasing domestic and industrial consumers. A secure and reliable supply of electric power from such a complex system is a major challenge for the power engineers while carrying out the planning and operation. This calls for continuous real-time monitoring and control of the power system. This research work aims on exploring this research area dealing with setting up measurement system, dynamic state estimation and event detection in the power system using the latest synchrophasor technology. Voltage stability analysis is considered important for secure power system operation. Many blackouts in India and abroad have been caused by voltage instability. So, the first step in the direction to solve this issue is voltage profile monitoring. A proper assessment is to be carried out so that the system operator can take preventive measures to keep the power system under stable conditions. Therefore, forecasting-aided state estimation (FASE) is performed for power systems having normal load variations. The estimation is carried out utilizing the measurements from the remote terminal units as well as the phasor measurement units. So far for the FASE studies, it is assumed that the changes in the system parameters, such as load variations are slow. Sudden changes in the power system states (namely voltage magnitude and angle) are therefore not considered. The effectiveness of the proposed algorithm Iterated Square-root Cubature Kalman filter (ISCKF) along with a state forecasting tool during normal load variations is evaluated with respect to already existing Kalman filter approaches. A state forecasting technique called Holt’s Double exponential smoothing is utilized to forecast the states during the interval between two time instants of receiving the measurement sets from the field. Research in the area of power system transient stability has recently shifted its focus on dynamic state estimation (DSE) involving PMU data with high reporting rate. Several mathematical models for induction motor and synchronous machine have been developed and various estimation approaches have been proposed in the literature for this purpose. In this work, the mathematical formulation of non-linear state space modeling and the principles of Kalman filter are utilized. Extended, Unscented and Cubature Kalman Fil- ters (EKF, UKF and CKF) are the three non-linear estimation methods explored for dynamic state estimation in an induction motor for our preliminary studies. In the next stage, after presenting a thorough explanation about modeling of the synchronous ma- chine, dynamic state estimation is applied on different power system case studies and the results of estimation methods are compared. Estimation studies are carried out in thescenario of three phase fault at one of the buses in the power system. The improvised CKF called Iterated Square-root Cubature Kalman filter (ISCKF) is proposed for its ap- plication in DSE and its performance is compared with that of the EKF, UKF and CKF algorithms. The simulation results obtained show the great potential of the proposed es- timation approach for accurately estimating the states of the machine as well as reducing the effect of noises. With the deployment of PMUs for wide area monitoring system (WAMS), it is feasible to have an insight into the real-time events occurring in power systems based on mea- sured data. Critical disturbances of the power system that may cause huge losses to the generation authority are noted as Events in this work. Initial studies are performed using analysis based on PMU data using various signal processing methods like Fast Fourier Transform, Yule-Walker Spectral Analysis, Matrix Pencil and Min-Max. That is, the events identifiable by two or more of these methods have been considered, to truly detect the occurrence of an Event based on the approach used by National Renewable Energy Laboratory. The results have been compiled on practical data from northern region of Indian power system during the Amphan cyclone. Then, a new major event detection method is proposed in this work, to detect the event efficiently using a data-driven ap- proach and a dynamic thresholding technique. For real time monitoring of the power system state variables, the measurement system needs to be properly set up. State of a system consists of a set of variables used to describe the behaviour of the system at a particular time. The optimal placement of measurement devices/sensors is essential for economic reasons and to maintain observability of the sys- tem. For measurement systems of electrical grid, supervisory control and data acquisition (SCADA) system was introduced around half a century ago to shoulder the operation and control of electric power system. As electric power system has been expanding in size due to advancements in the technologies and many new interconnections, the oper- ation and monitoring of the power system has also become complex. Consequently, the measurement systems have been upgraded continuously. From the past three decades, synchrophasor technology has become one of the cutting edge advancement in the field of electrical power systems monitoring. But the devices of this technology are installed judiciously for economical reasons. But, the extent of installations of phasor measure- ment units (PMUs) is not full fledged so as to replace the SCADA. This is due to the high reliability of the SCADA systems and PMU installations are a costly affair. Further, PMU installations are being carried out in phases. But, the days are not far where certain applications such as state estimation, voltage stability monitoring, etc. could be carried with PMU data alone. This is the direction in which this research work has been carried out.