Benders' decomposition for voltage profile management of distribution network with distributed generators

This paper deals with Benders' Decomposition techniques applied to Volt Var Control algorithms for MV distribution networks. This partitioning technique helps to solve certain classes of Mixed Integer Nonlinear Programming problems with decomposable structure. Therefore, this approach can be applied to the standard case of Volt Var Control algorithms in distribution systems which use continuous and discrete control variables. Considering the mathematical properties of prior algorithms, such as mixed continuous and discrete variables, Bender's Decomposition techniques allow an effective global optimum research. The effectiveness of this method is tested on the academic 30 bus distribution system and on a typical French MV network with 150 nodes/ 149 lines/ 5 onload tap changers transformers/ 3 DGs/ 1 capacitor bank.

1. Introduction

Distribution Networks (DN) were subjected to profound mutations during the last decade. One of the major modifications was the growing connection of Distributed Generators (DG) that leads to voltage profile modifications, among other impacts. Indeed, DN have been built initially to operate passively, with power flowing down from high voltage to low voltage levels. In order to mitigate impacts of DG, Distribution Network Operators are investigating the development of Advanced Distribution Automation functions (ADA) [1] such as Volt Var Control (VVC) [2]–[4] to support optimal operation of their DN. The objective of ADA functions is to use various flexibilities in the grid components associated with DGs coordinated support in order to make networks more active and to increase their operational efficiency.