Local voltage control in a low voltage grid with high photovoltaic penetration

Abstract: This thesis examines how active power control (Volt-Watt control) and reactive power compensation (Volt-VAr control) can help limit voltage increases on the low voltage grid due to high photovoltaic (PV), installations. Due to the large increase in interest to install small scale PV-systems within residential areas the Distribution System Operator E.ON, has noticed problems concerning high voltage levels on the customer side as the amount of PV installations increase. To limit the voltage increases the Distribution System Operator can upgrade the grid with newer and thicker cables and a new transformer, which takes time and costs money. In some cases it might also lead to relatively new cables being excavated. Another possible solution would be to use the existing grid in an optimal way. To test how the different control methods behave and to quantify how the overall impact would be for the PV-system owner, a model of an existing substation in the south of Sweden was built in PowerFactory. Once the model was built, local Volt-Watt and Volt-VAr controls were implemented on each PV-system. The system was then simulated for a full year using real load data from the customers and real solar irradiation values from SMHI where focus was on the aggregated behaviour over time. The simulations using Volt-Watt control found that, depending on where within the grid they were located, there was a large disparity between the different customers in terms of how much they would have to be curtailed. The figures ranged from 0 percent up to almost 20 percent of all yearly production in a scenario where 70 percent of all household had a PV-system installed. Volt-VAr was not as effective in decreasing the voltage magnitude as Volt-Watt due to the high R/X ratio but showed promise in cases when the voltage levels were not that prominent. Volt-VAr could help lower the voltage magnitude by a few volts if needed, and serve as way to delay the eventual need to curtail active power. This thesis also showed that there are many grid characteristics that influence how effective the voltage control can be. If situated in a weaker part of the overall grid, the voltage magnitudes in the local grid will become higher, leading to more active power curtailment if Volt-Watt is used, and limited chances that Volt-VAr could work. Another finding was that the tap position of the transformer makes a large difference in terms of curtailment of the PV-systems.