Please use this identifier to cite or link to this item:
Authors: Sajan, K.S.
Keywords: monitoring and control
implementation of WAMPAC,
Issue Date: Dec-2015
Abstract: Real time monitoring and control is the prime requirement for reliable and secure operation of large interconnected power system. The invention of the Phasor Measurement Unit (PMU) in mid 1980’s enabled the synchronized measurements of voltages and currents in real time that has become the foundation of today’s Wide Area Measurement, Protection And Control (WAMPAC) systems. A wide area monitoring system consists of PMUs, which are geographically dispersed in the system, and are time-synchronized through Global Positioning System (GPS) clock. The PMU measures instantaneous voltage, current, frequency and rate of change in frequency from dispersed locations of a wide network. These measured phasors with precise time stamp are collected at a centralized location, called as the Phasor Data Concentrator (PDC) that provides a real-time picture of the large interconnected power system. This synchronized data aids in all kinds of monitoring, control, and protection applications. This thesis primarily focuses on application of PMU measurements in wide area monitoring and control of real power system. In the implementation of WAMPAC, the first and the foremost step is the installation of optimally placed PMUs in the power system. The principal objective of PMUs installation is to make the system observable, either completely or incompletely. In practical scenario, PMU installation requires large expenditure. In addition to this, the proper site preparation, establishment of adequate communication and GPS are other essential requirements, which further increase these financial constraints. Many Optimal Placement of PMUs (OPP) techniques have been reported in the literature, such as integer programming, depth-first search, spanning tree method etc. Intelligent techniques like nondominated genetic algorithm, Tabu search, simulated annealing, particle swarm optimization techniques etc. have also been used for the solution of the problem. Different OPP methods, discussed so far, provide the optimal number and locations; the optimization of the number of branch current phasors of PMU has not been taken into account. A branch current phasor of PMU installed bus is used for calculating the voltage phasor of the other end bus by applying KVL. The current phasors measurement requires the number of current transformers and other auxiliary equipment equal to the number of lines connected to the bus, which increases the overall cost. In addition, measuring the current phasors of all the connected lines to a PMU installed bus becomes unnecessary if the bus connected to these lines are already under observation of neighboring PMU through its current phasor in the same power system. Hence, number of current phasors ii need to be measured by a PMU can be optimized, which will reduce the overall installation cost. In present work, a new approach for the optimal placement of PMUs with optimal number of current phasor measurements considering the different cases of observability has been proposed. In first case, the problem for OPP has been solved for the complete observability and then the optimal current phasors are determined. In second case, the OPP problems have been formulated with the optimal current phasors for incomplete observability with Depth-of-One unobservability and Depth-of-Two unobservability. In another case, the similar OPP problems have been formulated that ensures complete power system observability with N-1 contingency. The ultimate goal is to find the minimum number of PMUs and their corresponding locations with optimal current phasor measurements so that the state estimation could be performed with measured phasor data at the lowest possible cost. The problems are formulated as a generalized binary linear programming problem and genetic algorithm is used to search the solution for deriving the optimal configuration. In a large power system, a large quantum of PMUs is required for full observability that entails high capital cost, in the installation. These requirements pose physical and financial constraints on the PMUs installation process. It is imperative to formulate a staging program for PMUs installation in a phased manner and effectively utilize the PMUs even during the installation process in a power system network. Some of the authors have considered phased placement with additional benefits in terms of generator bus and tie line observability. This thesis addresses the problem of phased PMUs placement based on voltage stability monitoring and control. In this thesis, a new multi-phase PMUs placement approach based on voltage stability monitoring has been proposed. The approach uses Revised Analytical Hierarchy Process (RAHP). In this process, the first step is to determine the optimal number and the optimal locations of PMUs along with the channel limit that ensure complete power system observability even under a branch outage or a PMU failure. The second step is to identify the critical buses, which are more prone to voltage collapse using the Fast Voltage Stability Index (FVSI). The objective of the present work is to monitor the most critical buses in first phase either directly (PMU bus) or indirectly (neighboring bus). RAHP is used for decision-making in phasing of PMUs allocation. To get the maximum priority to the critical buses in multiphase installation of PMUs, Critical Load bus Observability iii Criteria (CLOC) has been introduced in the decision making process in addition to five other observability criteria (OC) namely, Noncritical Load bus Observability Criteria (NLOC), Generator Observability Criteria (GOC), PMUs Distribution Criteria (PDC), Tie Line Observability Criteria (TOC), and Bus Connectivity Criteria (BCC). As the synchrophasor measurements become available from the PMUs placed in phases, it is useful to integrate the conventional SCADA measurements and the phasor measurements to improve the accuracy of the State Estimation (SE). Therefore, developing a hybrid state estimator that includes both conventional and phasor measurements is required to get better results. Several hybrid SE approaches have been suggested in the literature, such as multi-area SE, distributed hybrid SE and quadratic hybrid SE. Most of the proposed hybrid SE methods result in an increased accuracy of the SE. This thesis provides a state estimation model, which can integrate both the SCADA measurements and the available phasor measurements obtained from multi-phasing installation of PMUs. The bad data detection and elimination in both the conventional measurement and the phasor measurements have been carried out using the normalized residual test method. In the recent years, voltage collapse is a major cause for many power system blackouts around the globe. The traditional method for voltage stability analysis relied on static analysis using the conventional power flow methods such as Gauss-Seidel or Newton-Raphson. In literature a number of voltage stability indexes, based upon conventional power flow have been proposed. The main drawback of these techniques is the singularity of the Jacobian matrix at the maximum loading point. To overcome this problem, Continuation Power Flow (CPF) method is used to compute voltage stability margin. The aforementioned techniques require comparatively large computations and are not time efficient for on-line applications. The ability of synchrophasor measurements to capture the fast power system dynamics makes the voltage stability monitoring of the power system more accurate and fast. A Genetic Algorithm based Support Vector Machine (GA-SVM) approach for online monitoring of long-term voltage instability has been proposed in this work. To improve the accuracy and minimize the training time of SVM, the optimal values of SVM parameters are obtained using genetic algorithm. The proposed approach uses the voltage magnitude and phase angle obtained from PMUs as the input vectors to SVM and the output vector is the Voltage Stability Margin Index (VSMI). Total Vector Error (TVE) in iv PMU measurement may be maximum of 1% which may include some uncertainty in phasor measurement. Thus, the impact of these uncertainties in synchrophasor measurements is also analyzed to detect the deviation of VSMI at the operating point. The results of the proposed GA-SVM approach for voltage stability monitoring are compared with Grid Search SVM (GS-SVM) and Artificial Neural Networks (ANN) based approach with same data set to prove its superiority. Flexible Alternating Current Transmission Systems (FACTS) devices are usually used for the dynamic control of voltage, impedance, and phase angle of high-voltage ac lines. FACTS devices provide strategic benefits for improved transmission system power flow management through better utilization of existing transmission assets, increased transmission system security and reliability. The STATCOM plays an important role in reactive power compensation and voltage control because of its operating characteristics, which have been well studied in the past years. In the literature, various control methods have been proposed for STATCOM control. In most of the work, the control logic is implemented with the Proportional Integral (PI) controllers. The PI controller gives only a control point not the range of control. This is due to the integer nature of PI controller. Therefore, PI controller required to be tuned time to time. For the dynamic operation of the STATCOM, Fractional order PI (FOPI) controller has been proposed in this work. FOPI controllers have been designed using the fractional calculus. These controllers have more degrees of freedom and provide a complete range of control. FOPI controllers have better capability of handling uncertainties and provide better stability. The voltage phasors and current phasors, which are assumed to be obtained from PMU have been transformed into their d-q components. Voltage and current components in d-q frame have been used as input to the FOPI controller. The simulation studies have been carried out in MATLAB version 2013a. The performance of FOPI controller has been compared with the Integer Order PI (IOPI) controller. Simulation results of FOPI controller obtained in MATLAB have also been validated in Real Time Digital Simulator (RTDS).
Appears in Collections:DOCTORAL THESES (Electrical Engg)

Files in This Item:
File Description SizeFormat 
Final_Thesis_to_print.pdf7.46 MBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.