The rise of photovoltaic (PV) penetration is contributing to the increasing incidence of overvoltage detection in the electrical grid during times of high-power generation. Overvoltage can cause disturbances or (partial) failures in the electrical supply network, since the components used are designed for a certain voltage band. One option to counteract too high voltage levels and thus ensure power quality, grid stability and resilience is the absorption of active power by means of a battery energy storage system (BESS). In this paper, we first built a suitable simulation setup for a typical European network section, including a large-scale PV system connected to the 10 kV level and a BESS model. A suitable charging and discharging algorithm for the BESS with the aim to realize peak shaving for the grid voltage was developed and implemented. Simulations, performed in MATLAB/Simulink®, show the dependence of the battery capacity and power on the grid-serving effect of BESS. By determining appropriate values for these two factors a significant reduction of the voltage level could be achieved.

Photovoltaic (PV) power plants are currently built at a high rate, both small ones for single family houses and large-scale plants with a power output in the MW range. This development accelerated even more since the beginning of the energy crisis in 2021. Due to intermittent energy production, large voltage swings may result in the electrical grid, which are challenging for grid and operators. We are considering the problem of overvoltage in times of excessive PV production, which is well documented in the literature. To avoid overvoltage, without PV-curtailment or grid expansion, grid-connected battery systems are proposed. The research focus is on advanced control strategies for such systems which charge and discharge at appropriate times to reduce the overvoltage in the grid while simultaneously minimizing the required battery capacity. The work presented extends a previous contribution of the authors. The analysis was done in a simulation environment (MATLAB/Simulink).