Comments, with reply, on "Minimization of reactive power under nonsinusoidal conditions" by L.S. Czarnecki

The commenter argues that ignoring compensator rating in the optimal design of a shunt LC compensator for power engineering applications is neither realistic nor physically justified. Furthermore, minimization of source reactive power only with respect to either L or C results in a not necessarily global minimum. Computation and var specification errors were also found and corrected. For comparison purposes, a realistic optimal design procedure is presented. In replying, Czarnecki states that compensators of reactive power are built in order to minimize an additional cost caused by the presence of reactive power in the circuit, and at their optimization the capital cost of the compensator must be taken into account. When the load parameters do not change, the compensator designer does not need any algorithm for reactive power minimization, but only a method to minimize the overall cost, since just this minimum is the criterion for the compensator optimization. However, when the reactive power of the load varies over a broad range, the designer may conclude that the cost minimization requires an adaptable compensator. He discusses two problems that have to be solved when its parameters are being adapted to the load.<>