Optimal VAr management of a large power network using most efficient method of weak node detection

A robust method for weak node detection for optimal VAr allocation has been proposed in this research. Modal or eigenvalue analysis, loss-sensitivity analysis, power flow analysis, L-index, and fast voltage stability index methods were employed to identify weak nodes in medium and large power networks using IEEE 57 and IEEE 118 bus systems. After identifying the locations for shunt VAr allocation, several standard optimization techniques i.e. Genetic Algorithm, Differential Evolution, Particle Swarm Optimization, Teaching-Learning-Based Optimization, Big Bang-Big Crunch, Krill Herd Algorithm, and Sine-Cosine Algorithm were then applied for optimal reactive power planning. These methods were aimed to coordinate VAr injections by shunt capacitors at weak nodes with existing VAr sources such as generators and On-Load Tap Changers. A comprehensive study of various standard detection techniques and their usefulness in power system planning is presented by the authors. Finally, authors deduced LSI (Line Stability Index) method of weak node detection was the most accurate method for weak node detection. Carrying out the proposed methodology, the total active power loss with GA & DE was found as 0.2348 p.u. & 0.2351 p.u. respectively for IEEE 57 bus test network and total operating cost was found as $1.2344 × 10 7 for GA & $1.2357 × 10 7 for DE. Similarly, for the IEEE 118-bus test network, the active power loss reported under GA and DE was 1.3295 p.u. each, with corresponding operating costs of $6.9885 × 10 7 and $6.9880 × 10 7 , respectively. This research offers a comprehensive framework for reducing operating cost and active power loss in reactive power planning. • Detection of weak nodes of a large power network. • Proposition of optimal VAr adjustments by the OLTC's, generators, and the shunt VAr injection at weak nodes. • Use of different optimization techniques to accomplish the task. • Comparative analysis of different techniques on the basis of the results obtained. • Deduction of loss and cost minimum configuration of the power network.