New voltage sensitivity analysis for smart distribution grids using analytical derivation: ABCD model

Abstract Sensitivity Analysis plays a significant role in voltage prediction and control of power networks. However, the classical sensitivity methods require significant computation time. As active distribution networks require real-time implementation for voltage control, reducing the computation time becomes a necessary task for network operators, especially in the context of optimization techniques. This work develops a new analytical and fast voltage sensitivity analysis method via the derivative of the nodal quantities (power, current and voltage) with respect to power injections. The proposed method mainly depends on the construction of the ABCD matrix. The values of the matrix elements remain the same regardless of the bus on which the power is injected. Thus, it has a high potential to be implemented in online applications. To make a complete separation between the sensitivities to active and the sensitivities to reactive power injections, the analytical formulations are expressed in Cartesian coordinates. A radial distribution network including several DG units is used to verify and assess the proposed sensitivity method under different scenarios.