Simulationswerkzeuge

SmartGrid - the network of intelligent producers and consumers opens up fundamentally new perspectives for the more conscious use of increasingly scarce resources. The inherent complexity of a SmartGrid makes its design a challenging task that can be solved with the help of extensive simulations and complex models. Commercial tools often only allow manual consideration of exemplary cases. Automated optimization of holistic simulations is typically not part of the functional and performance scope of available software.

At e²n, accelerated methods for parallel computers and graphics card accelerators are developed. The main research topics are:

• Efficient computation of the electrical load flow on graphics accelerators
• Cluster-based fast annual simulation
• Distributed plant models
• Efficient interfaces for distributed SmartGrid simulations

For extensive simulations of the future distributed power supply system with a high share of renewable energies, the department has its own computing cluster for GPU-based calculations. Furthermore, a second computing cluster is available in collaboration with Fraunhofer-IEE. At the end of 2016, this was expanded by the department as part of the "OpSimEval" project(www.opsim.net). A total of more than 24 compute nodes with a total of 440 CPU cores, several NVIDIA GPUs and more than 3.5 terabytes of RAM are available.

Possibility to study the behavior of microgrids with a high share of renewable generation.

A laboratory facility consisting of two converter cabinets, a control cabinet and a set of machines is available to the e²n department. The laboratory is located at the Design Center Modular Supply Technology (DeMoTec), which is jointly operated by Fraunhofer IEE and the University of Kassel. The inverters are freely programmable via a Matlab/Simulink interface. One inverter is used to control the machine set. The other inverter can be configured as a voltage or current source. This allows the behavior of a photovoltaic system to be simulated, for example. The power class of the inverters and the machine set is 15 kVA. The system is designed for a nominal voltage of 400 V. All components of the system can be connected in a practical way via the control cabinet. The addition of further elements, such as a three-phase load, is also possible.

The experimental setup is used to investigate the dynamic behavior of microgrids. The focus is particularly on the interaction between the control of synchronous generators and converter-based systems, such as photovoltaic systems.

Professional control center software for the operational management of power grids

The e²n department has a test control center from the industry. Control stations are used to monitor and manage real networks. They receive measurements from the network, which are processed and visualized. Commands, e.g. for switching, are sent back to devices in the network. A test control station was provided to the e²n department by a leading control system manufacturer; the control station was connected to the OpSim simulation tool via an IEC 60870-5-104 proxy and can send/receive data in both directions (see figure below).

The control station provides management of network data both in databases and visually. Using created SCADA network images, networks can be monitored and controlled. Another mode, the so-called HEO mode, is used to perform network calculations and simulations. These include power flow calculation or network state estimation, but also functions for calculating failure variants. This functionality, in combination with other OpSim components such as the real-time network simulator, enables the simulation of real network operation for testing or training purposes. Use cases are tests of new network management strategies or components that cannot be directly connected to the control room. Due to the large number of interfaces, communication between the control center and external components can be established via OpSim. The OpSim system can also be used flexibly for comparison (benchmarking) of novel algorithms with the industry standard of the control room.