Optimisation strategy applied to DG reactive power for decentralised voltage control in smart grids

Development of Smart Grids could increase the efficiency of modern distribution systems, facilitating large scale penetration of Distributed Generation (DG) devices which exploit renewable energy sources, encouraging an efficient use of energy by smart loads, reducing system losses and improving power quality. In this context, a first step is to overcome the well-known difficulties related to increasing penetration of the DG devices into existing distribution systems by properly accounting for the interaction of DG devices with the Smart Grid operation and control. Special attention has to be paid on assuring an adequate voltage profile along the feeders. In fact, the presence of DG devices interferes with traditional voltage control structures. Usually, in MV distribution system voltage regulation is performed by changing the transformer ratio in the HV/MV substation, using a On Load Tap Changer (OLTC) control system. In some cases, DG devices may cause overvoltages because they inject large active power into the distribution network; in other cases DG devices may cause undervoltages because they reduce the current supplied by the HV/MV substation and, consequently, the OLTC compensation. To overcome such problems, more appropriate voltage and reactive power control structures have to be developed for Smart Grids. Referring specifically to DG devices, the paper proposes to enrich the classical reactive power control scheme so as to improve the voltage profile of the feeder which the DG device is connected to. A new Optimisation Strategy Task is introduced, composed of two subtasks. The first subtask estimates the operating conditions of the distribution system, using only local measurements of voltage and current at the Point of Common Coupling (PCC). Then, on the basis of the results obtained by the first subtask, the second subtask determines the optimal reference value of the DG reactive power regulator. In particular, the paper focuses attention on the first subtask and analyses in details the algorithms that are necessary to estimate the HV/MV substation operating conditions and the loads of the distribution system.