Advanced control methodology for improving small-signal stability of a large electricity network

Low frequency electromechanical oscillations, ranging from 0.1 to 2 Hz are inherent phenomena of interconnected power systems. Problems due to inadequate damping of such oscillations have been encountered throughout the history of power systems.     Power system stabilizer (PSS) alone, sometimes may not be sufficient to enhance smallsignal stability of the electricity network. To alleviate this problem, Flexible AC Transmission System (FACTS) devices deployed in transmission networks can be used by adding a supplementary controller. Such supplementary controllers are often known as Power Oscillation Damping Controllers (PODs). The operating condition of a power system always changes with the variations of generation and load patterns. Therefore, it is essential that these PODs should provide necessary damping to the electromechanical modes over a wide range of operating conditions. To address this research problem, in this thesis a robust controller based on a H∞ loop-shaping technique has been designed. To check the robustness of the controller it has been tested on a wide variety of operating conditions on small and large scale electricity networks. While improving small-signal stability of low frequency inter-area modes, the locally available POD input signals do not always have optimal observability of the mode of interest. This degrades the performance of the damping controller. In this situation, phasor measurements accessible from Wide Area Measurement Systems (WAMS) provide new opportunities for designing effective damping control schemes. With WAMS, maximum observability for particular modes may be obtained from the remote signals or the combination of remote and local signals. In this thesis, a centralized damping controller based on wide area measurements has been designed to increase modal observability in the feedback signal. Also, the combination of decentralized and centralized controllers using local and remote signals has been investigated. To validate this approach a two level control design is presented to improve small-signal stability of the system. This two level control design approach is tested on a 14-generator South East Australian system and on the Queensland network. An essential component in designing a centralized damping controller is the proper selection of the feedback signal. In this thesis, three different methodologies for selecting this remote signal have been evaluated. A systematic comparison is also presented to identify the best method for the feedback signal selection. Apart from operating point uncertainties, electricity loads exhibit changes in their characteristics based on the time of day, week, season and weather. With the proliferation of digital technology, the electricity loads show more uncertainties in their characteristics. This requires the designed damping controller to improve small-signal stability of the system in the presence of various load characteristics. In this thesis, a wide variety of load models have been included while testing the effectiveness of the H∞ loop-shaping controller. It has been found that the proposed damping controller can provide effective damping to the mode of interest under a wide variety of load models at various operating points. In the deregulated electricity environment, explicit system models may not be available to a system operator. In this circumstance, a model-free controller design is required to improve smallsignal stability of the system. In this thesis, a model-free damping controller based on an extremum seeking control has also been investigated for a large scale electricity network and its performance has been tested against a decentralised damping controller. It has been found that the extremum seeking approach works better than H∞ loop-shaping for a test system. However, this controller is yet to be tested for a practical power system and in the presence of system load model uncertainties. Increasing environmental concerns and reducing carbon footprints are bringing more and more renewable energy sources into an existing grid. Rapidly evolving wind energy conversion (WEC) technologies and the widespread use of these technologies to generate a large amount of power poses a significant challenge to network operation and planning. It is believed that the WEC technology used today does not directly affect small-signal stability of the power system. However, it has a significant indirect impact on the overall oscillatory behaviour of the system. Therefore, it is vital to examine the impacts of wind energy sources on the small-signal stability of the network. To address this research problem, two case studies are presented in this thesis. In the first case study, impacts of large wind farms in South Australia have been studied on a 14-generator South East Australian network. A second case study has been carried out on the Queensland network considering prospective wind power penetration in North Queensland. It has been revealed that small-signal stability of the network is significantly influenced by large wind power penetration. This effect varies from one power system to another and also depends on the strategy used to accommodate wind power. The contribution made in this thesis can be extremely helpful to the electricity network operators for designing damping controllers to improve small-signal stability and also to understand the influence of wind power on small-signal stability of the network.