Analysis and evaluation of grid reduction methods for distribution grid state estimation

Abstract

Consulta en la Biblioteca ETSI Industriales (Riunet)

[EN] The increasing integration of decentralized, renewable generation units into the Ger-man distribution grids has resulted in new challenges for grid operation. Knowledge of the grid state has become more and more important for decentralized Energy Man-agement Systems (EMS) to work. An important tool for EMS is State Estimation (SE). Based on measurements taken throughout the network, SE gives an estimation of the state variables of the power system while checking that these estimates are consistent with the measurement. This thesis deals with the application of a Grid Reduction (GR) method for the state estimation of distribution grids, particularly for low voltage grids. The goal of the GR approach is to find an estimation of grid conditions by specially using existing meas-urements as input variables. GR tries to more intensely exploit typical properties of the distribution grid topology for state estimation. By employing MATLAB, a suitable pro-gram based on GR approach is developed to determinate the state variables of the grid. The influence of relevant networks parameters and load assumptions on estimation quality is analyzed. The location and minimum number of measurement facilities re-quired to improve grid state estimation is discussed. The implemented program is vali-dated on exemplary low voltage grids (radial and ring structures) to evaluate the esti-mation quality of the technique, especially with respect to network topology.