Voltage Sensitivity Analysis of a Laboratory Distribution Grid With Incomplete Data

Abstract

New voltage control algorithms are necessary to cope with the increasing amount of distributed generation and electric vehicles in distribution networks. Many of the newly proposed voltage control algorithms are based on linearized dependencies between the voltage magnitude, and the active and reactive power consumption. These linearized dependencies are normally obtained by algorithms, which rely on accurate grid topology information. Due to the traditionally passive operation of low voltage (LV) distribution networks, this information is typically missing, incomplete, or inaccurate. Therefore, this paper introduces a method to extract these linear dependencies based on historical smart meter data only. No information about the grid topology is required. The model adapts to the changing load conditions in the network. The algorithm has a low complexity and is applied to an unbalanced LV distribution network. Data of a practical laboratory setup is used to validate the proposed method in real-life conditions. With the obtained voltage sensitivity factors a voltage management strategy was implemented for the laboratory grid.