Active power-controlled photovoltaic (PV) energy sources are equipped with control schemes to regulate the active power fed into the grid (real power) even below the maximum power point. In [1], two methods — a model-free and a model-based algorithm — were presented to investigate how quickly PV systems can adjust their power to provide negative and positive inertia. Furthermore, for the orchestrated frequency and voltage stabilization of distributed energy resources (DERs), a general concept of dynamic virtual power plants (DVPP) was presented in [2]. DVPPs explicitly incorporate dynamic aspects at all levels for controller design and grid-integrated operation. The basic idea is that the DVPP, as a whole, participates in the ancillary network services and interacts dynamically with neighboring network elements. For this purpose, the desired dynamic global behavior of the DVPP in a network section is described by an aggregated transfer function. To distribute the tasks to the individual generator units of a virtual power plant, these are divided (disaggregated) into local reference models, which are used for a model reference controller design [3].

So far, the concept has only been investigated for large distribution grids with few concentrated and symmetrical loads. In this work, a DVPP for active distribution grids with asymmetric loads and distributed PV generation units without storage is investigated using the 13-node test feeder grid topology, which provides a suitable benchmark for the DVPP system in active distribution grids for unbalanced systems. The 13-node test feeder is characterized by being short, relatively highly loaded, and single voltage regulator at the substation, overhead and underground lines, shunt capacitors, an in-line transformer, and unbalanced loading [4].

For the investigations presented here, the original 13-node test feeder system is expanded to include active power-regulated photovoltaic energy sources and an external grid, with power balance-dependent network frequency calculated using an aggregated swing equation that takes into account power imbalance at the point of coupling between the distribution and external grids. The design process of the global DVPP, the local control of the PV systems, and the results for various load scenarios are presented and discussed.

[1] Schulte, Horst: Advanced Control of Grid-Integrated Renewable Energy Power Plants, John Wiley \& Sons 2024 (IEEE Press Series), doi: 10.1002/9781119701415

[2] Marinescu, Bogdan.; Gomis-Bellmunt, Oriol.; Dörfler, Florian.; Schulte, Horst; Sigrist, Lukas: Dynamic Virtual Power Plant: A New Concept for Grid Integration of Renewable Energy Sources. In: IEEE Access, 2022, doi: 10.1109/ACCESS.2022.3205731

[3] Schulte, Horst: Takagi-Sugeno Model Reference Control of Photovoltaic Power Plants to Provide Instantaneous Reserve. In: Proceedings of IEEE International Conference on Fuzzy Systems 2025 (FUZZ-IEEE). IEEE. Reims, France, June 06-09.07.2025.

[4] 13-Node test feeder was created in 1992 and approved by the DSA Subcommittee during the 2000 PES Summer Meeting.

https://cmte.ieee.org/pes-testfeeders/resources/