PMU PLACEMENT CONSIDERING COMPONENT RELIABILITY OF POWER SYSTEM AND ITS APPLICATION FOR ONLINE VOLTAGE STABILITY MONITORING

Abstract

In a large inter-connected power network, reliable and secure operation require real time monitoring of wide area network and its stability. The requirement of wide area monitoring system (WAMS) payed path for the development of synchro phasor technology. The invention of Phasor Measurement Unit (PMU) brought a major shift in the field of Wide Area Measurement, Protection and Control (WAMPAC). PMU provide synchronised and highly accurate real time data measurement of voltage and current phasors from the geographically dispersed location. PMU communicate the measured analog data in digital format to Phasor Data Concentrator (PDC) as per IEEE C37.118.2-2014(a) protocol which are located in state, regional and national level control centres. PMU installation has improved the performance of various applications such as state estimation, protection and transient stability analysis. The work presented in this thesis primarily focuses on three important aspects of WAMS: • The study of component reliability of PMU data measurement system for enhancing the availability/reliability of data measurement through reliable network. • Application of PMU measurement to prevent voltage collapse due to weak load buses. • Application of PMU measurement in online real time voltage stability monitoring. In WAMS implementation, first and the primary objective is the optimal placement of PMU in the power network. The PMUs must be deployed in such that whole system become observable, either completely or incompletely. The synchro-phasors data of PMU bus helps in determining the data of other buses which is connected to PMU buses using Ohm’s law. Hence, in a well-connected power network, every bus is not needed to be installed with PMU. The requirement of PMUs goes down to almost one third of the total number of buses present in the power network. The financial constraint also does not allow to install PMU at every bus. This has encouraged the researchers to develop algorithms, such as depth first search, integer linear programming, spanning tree etc. for finding minimum number of PMUs. Evolutionary techniques such as, Particle Swarm Optimization (PSO), Genetic algorithm (GA), Tabu search, Simulated Annealing (SA) have also been proposed for finding minimum number of PMUs. The problem of finding necessary location of PMUs in widely dispersed power network is called the Optimal PMU Placement (OPP) problem. i This thesis utilizes reliability concept to presents a PMU placement methodology for enhancing the observability of interconnected power network with the incorporation of component reliability. At first, modified genetic algorithm is proposed for implementation OPP problem in mixed ILP (MILP) framework with binary-valued variables. Multiple optimal solutions for PMU placement considering complete observability have been determined with the use of modified genetic algorithm. An overall system reliability index has been proposed for the selection of the most suitable solution among the determined solutions. Then, the reliability parameters of components involved in PMU data measurement are utilized to find the most suitable optimal solution among determined multiple optimal solutions. This strengthens the observability of interconnected power network through most reliable buses. There are different types of contingency present in power network which causes loss of data measurement. The (N-1), (N-2) contingency present in the power network generates the requirement of redundancy in data measurement. To overcome this problem, each bus is made observable by at least two or more PMU. Redundant observability is defined as a type of observability in which data of each bus is measured by at least two or more PMUs. The researches have also utilized the concept of redundant observability in PMU placement for increasing the redundancy in power network. Redundant observability uses more number of PMUs than complete observability. This thesis proposed a hybrid observability to reduce the requirement of PMUs for complete observability under contingency with the use of reliability concept. The proposed methodology secures the data measurement of unreliable and weak load buses of power network considering hybrid observability for multi-phase PMUs installation. The (N-1), (N-2) contingency present in the power network requires redundancy for data measurement. The network components with high reliability decrease the contingency of power network. This reduces the requirement of data measurement redundancy for the network component having high reliability. The present paper utilizes this relationship between network component reliability and power network contingency to propose the concept of hybrid observability for PMU placement. It combines the concept of complete and redundant observability. Power network reliability is determined by incorporating component reliability of PMU data measurement system. This strengthens the data measurement of weak load buses and low reliable buses preventing voltage collapse due to data unavailability ii Today, power systems cover huge geographical area and involve thousands of buses for power flow. This require installation of large quantum of PMUs for maintaining full observability of power network. The financial and infrastructure constraints do not allow all PMU installation in one go. To overcome these constraints, state and central transmission utilities install PMU in different phases. Many researchers have considered various characteristics of power system for multi-phasing of PMU placement. Complete PMU installation requires long time, leading to unavailability of data from remaining buses till the completion of all phases. The data of PMU installed buses is used for estimating data of remaining buses. This increases the importance of PMU sites during the initial phases of PMU installation. PMUs must cover the observability of critical locations such as weak load buses (in respect to voltage collapse), reliable buses (in respect of data measurement) and Generator buses (for small signal analysis) during the initial phase of PMU installation. To address aforesaid issue, this thesis presents a methodology for multi-phasing of PMU location using power system characteristics associated with above mentioned critical locations. The indices viz. Weak Load Bus Closeness Criteria (WLCC), Reliability Observability Criteria (ROC) and Generator Observability Criteria (GOC) are utilized to identify and rank the PMU locations which are associated with above mentioned critical locations using Analytical Hierarchical Process (AHP). WLCC and GOC determines the closeness of PMU location bus from weak load buses and generator buses. ROC evaluates the effectiveness of PMU location in measuring the overall reliability network. Inclusion of WLCC, ROC and GOC strengthens the data measurement of weak load buses and generator buses through highly reliable network during the initial phases of PMU installation. This thesis also developed an approach for the phasing of PMU locations based on the concept of Superset theory. It ensures minimum distance between unobserved and observed buses during initial phases of PMU placement. In recent years, voltage collapse has been discovered to be the main cause for the power system blackout around the globe. Voltage stability analysis is performed using traditional methods such as conventional power flow using Newton Rapshon, Gauss Seidel etc. The literature revels that various indices have been proposed by researchers to determine the vulnerability of power system using conventional power flow method. However, these methods have disadvantage as Jacobian matrix becomes singular at the maximum loading point and causes divergence of solution at the maximum loading point. The problem of iii conventional power flow method is resolved by the introduction of continuation power flow (CPF). The computation required to execute aforesaid method is comparatively large and not efficient for online applications. The literature also reveals that the synchronized, accurate and fast processed data obtained from PMUs can also be used for online application related to Wide-area monitoring, protection, and control (WAMPAC). Artificial Neural Network (ANN) is an effective and efficient approach for solving problems, which require fast and complicated computation. This study proposed a Particle Swarm Optimization based Artificial Neural Network (PSO-ANN) hybrid model for evaluation of voltage stability margin. PSO optimizes the meta-parameter of ANN for improving its convergence rate. This model helps transmission utilities in real time monitoring of long-term voltage instability to avoid grid blackout. Power system security is defined based on the single contingency ‘N - 1’ criterion, meaning normal system minus one element. Separate ANNs are designed for worst-case ‘N - 1’ contingencies along with base-case configuration. PMU installed at distant locations provide real-time synchronized measurement of voltage phasor, which forms input feature for proposed model. Load P margin evaluated by continuation power flow method is considered as output feature of proposed ANN model. The proposed approach is effectively tested on IEEE 14 and 30-bus system.