Discussion of "advanced SVC models for newton-raphson load flow and newton optimal power flow studies"

Abstract

SVC’s normally are connected to the power system through a medium or low voltage transformer [3]. This transformer is modeled together with the equivalent reactance of the SVC regulation characteristic [4] [1]. This implies that an exact model should involve a reactance equivalent of the transformer and the high voltage node should be regulated remotely. However, in your paper this situation is not explained, even though in Section IV-E a control strategy with several devices is explained based on remote voltage control in PQ nodes. Could the authors explain how the Newton‐Raphson method would be modified under this situation? The comparisons between the proposed SVC models in this paper and the generator model are based on the quality of the power flow solution. In this sense, there is not any comparison with the generator/fixed-compensator model. Have the authors some results about this comparison? Also, from the robustness and computational effort points of view, would the proposed models in this paper represent any advantage compared with the generator/fixed-compensator model? On the conclusion section it is pointed out that the proposed model using the firing angle as state variable gives a more complete information. Would you like to explain the content of this information? Even though here it is discussed the Newton‐Raphson method, the generator model of SVC’s does not present any problem when the Fast Decoupled Load Flow (FDLF) method, while their limits are not violated. When the equivalent susceptance and firing angle models are used updating of nodal admittance and B matrices becomes necessary, which will result in some complications to implement these models on the FDLF method. Would the authors like to make some comments about this problem?