Capacitor bank model validation with particle swarm optimization algorithm

In most substation protection studies, capacitor bank parameters are derived from manufacturer nameplate values. However, the actual parameters could be different from the nominal rating as a result of manufacturing variance, assembling, material aging, and temperature dependencies. Wide manufacturing tolerance could lead to non-negligible calculation inaccuracy that affecting successive studies. For impedance based protection, inaccurate parameters for relay settings can also cause mal-operation. In this paper, the field current and voltage measurements are used to validate the size of capacitor bank and the capacitor ratings. A 10–15% deviation between the instantaneous current peak measurements and simulation results from capacitor bank model is observed. This paper introduces, the use of particle swarm optimization algorithm to fine tune parameters for possible relay settings. MATLAB simulation shows that a set of capacitor bank parameters is yielded by PSO algorithm through reducing normalized deviation between field measurements and simulation to around 4% – 5%.