A novel optimisation method for voltage and reactive power control of electric power systems

Abstract

Reliable electrical power supply is one of the most important utilities for modern society. This can be seen by the fact that any prolonged interruption of electrical power supply usually leads to enormous disruption of essential services and normal daily activities, and can in fact threaten to cause a lot of damage or losses if not promptly remedied. Moreover, recent developments in the power system, such as the deregulation and restructuring of the electrical power supply industry, the introduction of competitive electricity and power markets, and the rapid growth and expansion of distributed and decentralized electrical power generation, have led to a significant increase in the complexity of modern power systems, adding to the challenge of operating them reliably and efficiently. Thus, the need for optimal strategies for the secure, economical and efficient operation of the power system is arguably even greater now than at any other time in the history of the power system. In line with this identified need, this thesis investigates the theoretical design, development, and practical implementation of efficient algorithms that contribute to the secure, economical and reliable operation of electric power transmission systems. The focus of the research presented in this thesis is on the development of methods and algorithms for the solution of the Volt/VAR optimization (VVO) problem, which is a very important sub-problem of the optimal power flow (OPF) problem that is primarily concerned with determination of the optimal coordinated dispatch of voltage-regulating devices and reactive power sources, with voltage profile improvement and system power loss minimization as the main objectives (among others). Volt/VAR optimization is one of the most actively researched areas of power system operation. While most researchers consider either classical or heuristic optimization methods in isolation, the research work presented in this thesis investigates the design of efficient Volt/VAR optimization strategies considering both classical and heuristic optimization techniques. Two main optimization algorithms are developed for the solution of the Volt/VAR optimization problem in this thesis. One is based on the primal-dual interior-point method (PDIPM), which is one of the most efficient classical methods for large-scale nonlinear optimization. The other is based on the particle swarm optimization algorithm, one of the most popular heuristic optimization techniques. To enhance the efficiency of the developed algorithms, the model development for the Volt/VAR optimization problem considers both the polar and rectangular coordinate representations of the system voltages. Although most researchers make use of the polar representation, analysis reveals that the rectangular representation has relatively more favourable mathematical properties from the computational efficiency perspective, particularly for the methods and algorithms developed in this thesis. The efficiency of the developed methods and algorithms is further enhanced by incorporating the Newton- Raphson load flow computation into the Volt/VAR optimization algorithm, which is moreover also developed using the rectangular model formulation. Five power system case studies, a 3-bus system, a 6-bus system, and the IEEE 14-bus, IEEE 30-bus, and IEEE 118-bus power systems, are used to analyse the performance of the developed algorithms. The results obtained from the performance analysis reveal that the developed algorithms exhibit high computational efficiency and superior convergence characteristics. Moreover, a comparative performance analysis is also conducted between the PDIPM-based VVO algorithm and the PSO-based VVO algorithm. The performance analysis reveals that the primal-dual interiorpoint method outperforms the particle swarm optimization algorithm in terms of computational efficiency, since on average it requires fewer iterations to converge, and has a shorter running time. The particle swarm optimization, on the other hand, generally achieves a higher percentage real power loss reduction than the primal-dual interior-point method. This suggests that the two classes of methods (i.e. classical and heuristic optimization methods) have complementary performance characteristics, something which could be exploited to devise optimization strategies that seek to combine their relative strengths, and thus have a better prospect of exhibiting performance that is superior to that of the individual algorithms. The methods, algorithms and software programs developed and presented in this thesis are of great relevance both to industry and to academia, and can serve as a good foundation for further research and development, as suggested in the concluding chapter of the thesis.