Decentralized Voltage Optimization based on Network Partitioning in Distribution Systems with DGs

As well known, the large-scale spreading of Distributed Generators (DGs) in distribution systems can cause a significant impact on the voltage profiles along the feeders. On the other hand, the reactive power injections by DGs can be used to reduce such an impact and regulate the nodal voltages. In this paper, the basic idea of a decentralized voltage optimization problem is proposed: it minimizes the deviations of the nodal voltages from their reference values so as to determine the optimal set-points of the reactive powers injected by DGs. To reduce the number of measures from the grid and to avoid a large number of non linear equality constraints, the partition of the network in two voltage control zones (VCZs) with pilot nodes (PNs) and a linear method for the steady-state operation of the distribution system are used. The decomposition of the problem into two problems is based on the theory of Lagrange multipliers and it results into a quadratic programming problem of limited dimension to be solved by each VCZ. The overall problem solution is obtained by iterate the VCZ solutions and exchanging a limited number of scalar quantities between PNs. With respect to a centralized approach, the proposed decentralized voltage optimization provides the same solution with a reduction of the requirements for the communication infrastructure. Numerical results in a case study are presented to give evidence of the improvement of the voltage profiles along the feeders of a 24-nodes LV distribution grid with DGs.