The energy transition occurs at the distribution grid system level. Through technical progress and political will, more than one million photovoltaic plants have been connected to the German public power system since the beginning of this century. To a large extent, this was accomplished by implementing the PV systems at the lowest voltage level. In more and more regions, this has led to a paradigm shift in the distribution grid, i.e. the flow of electricity has become bi-directional, which results in unprecedented challenges for grid planning and operation. Solutions for these are known collectively under the term of “Smart Grid” and have been studied in numerous public-funded research projects.

The present work should be classified in this context. Key features are a systemic view of the network as a whole system rather than as a collection of individual components. Due to the focus on the whole system, the results have been obtained exclusively by means of software simulations and not in the laboratory or in the field. Real network data, generic consumption profiles and real feed-in processes are used in connection with the simulation tools for this purpose.

Real network data from the medium- and low-voltage grid levels, which are subordinated to a single substation, were provided by a distribution network operator. On this basis, the low-voltage networks are divided into six clusters, based on specific qualitative and quantitative criteria, and twelve representative low-voltage networks, as well as the two medium-voltage networks, are modelled in detail. In addition, a distribution of PV plant sizes to grid connection points at the low-voltage grid level is assumed, based on an existing public register of photovoltaic plants. The time curves for the electricity feed-in are obtained from real production data. The assumptions for the penetration of PV systems and the clusters are used to define the various configurations. As a starting point, the representative low-voltage networks are considered without PV systems installed. In a second step these networks are considered for the case with maximum decentralized, PV system feed-in. Finally, the entire medium-voltage grid level is investigated by aggregating the low-voltage networks through representatives from their clusters.

Load flow simulations are the main focus of this study. For this purpose, a network calculation program is employed, which is why the standard equipment modelling has been adopted. The network simulation only considers a single-phase, that is, only in a positive sequence system. The simulations are carried out quasi-statically by specifying profiles for active and reactive power at the grid connection points and considering a simulation period of two years. Node voltages, as well as the loading of the transformers and lines are results of the evaluation. The simulations are carried out considering different scenarios. These vary in the methods used for the elimination of limit violations. In the low-voltage network, the voltage range is almost exclusively affected. In this case, the use of a regulated distribution transformer has proved to be the most effective measure. Reactive power control at the PV-inverter is more suitable in some cases, but also leads to higher loading of the equipment. In the medium-voltage grid, individual lines are also overloaded under the selected assumptions.

The effects on service reliability are examined with the help of a probabilistic reliability calculation. The assessment is based on standardized parameters, such as the average unavailability for network customers. This value is mainly influenced by additional switching options for re-establishing service, although the overall effects are rather small.

A financial analysis of the examined aspects completes this investigation. In the current regulatory situation, grid upgrades or compensatory payments are chargeable costs in general, due to the settling charges. For this reason, there are currently, at best, indirect incentives to the grid operator to solve voltage problems in the most efficient way, or to increase service reliability through grid upgrades.