An analytical method for solving reactive optimization in power systems

Based on the decoupling principle, a quadratic programming model with "nonincrement" variables for evaluating an optimal scheme for reactive power compensation in power systems is presented in this paper. In the model, the reactive power flows on all lines are regarded as the variables to be evaluated and the objective is to maximize the present worth of revenue savings associated with reactive power compensation and energy loss reductions resulting from the application of VAr compensation. In order to completely resolve the convergence problem, the Lagrange multipliers method and the method of "setting the violating quantity equal to its bound" are employed. The proposed model is suitable for resolving the reactive optimization in 35/spl sim/110 kV power networks. This paper concludes with the results of some numerical studies performed on the IEEE 6-bus sample system and a practical 35 kV power network with 21 buses. The method is easy to use and the results are satisfactory.<>