The FfE distribution grid simulation model “GridSim” is a simulation model for the detailed analysis of distribution grids based on load flow calculations. The effects of diverse, decentralized generation and consumption systems on the distribution grid can be determined thanks to comprehensive mapping of the electrical energy system at distribution grid level. The name of the simulation model is derived from the English word “grid” for electricity grid and “sim” short for simulation.
The simulation model was developed at FfE to analyze and evaluate the changes and challenges that the distribution grids will face in the course of the energy transition. Since its creation in 2012, the model has been continuously developed and applied in various projects (see Figure 1 for the central projects). GridSim served as a tool for answering research questions in these various projects, enabling 2 dissertations and over 20 theses to be written in the process.
The foundations of GridSim were laid in the eGAP projects. The aim of these projects was to evaluate the effects of a high penetration of electric vehicles (SmartGrid @GAP) as well as prosumers with PV systems, home storage systems, and electric vehicles (Sun2Car @GAP) on the distribution grids in Garmisch-Partenkirchen. Likewise, possibilities for grid optimization using controllable local grid transformers (rONT) and longitudinal voltage regulators were investigated (Smart Grid Controller @GAP).
In the project Merit Order Grid Expansion 2030 (MONA 2030), the model was expanded to include the heating sector and its associated components, heat pumps and electric storage heaters (often also called night storage heaters). These extensions in MONA also allowed numerous grid optimizing measures to be implemented in the model and systematically compared with each other using typical representative grids that were selected within the project by means of a cluster method.
In the subsequent project C/sells, the model was used to determine the flexibility requirements in the demonstration cell centered around the town Altdorf bei Landshut. For this purpose, the real grid layout of the medium and low-voltage level as well as the associated grid loads were integrated into GridSim, and the calculated flexibility demand was transferred to the simultaneusly developed „Alf“ Flex platform.
In the eXtremOS project, simulations are used to exemplarily determine the flexibility that can be provided by the distribution grid without violating its grid restrictions. This value is then transferred to the energy system model ISAaR in order to be able to call up flexibility from the distribution grid there.
In the project Munich Electrified (Me), GridSim is used to assess the grid load caused by electric mobility in Munich. For this purpose, different variants for the use of the charging infrastructure (e.g. at home, at the workplace or in public) were integrated into GridSim. The current and future grid load in the distribution grids of selected study areas and the effects on the residual load for the whole of Munich were analysed in case studies.
As part of the model comparison in the MODEX MEO project, GridSim is methodically compared with other models from other research institutes by calculating different future scenarios, such as the expansion of renewable energies or the electrification of heat and mobility. The comparison serves to identify suitable modeling structures for specific questions, works out optimisation potentials of the different models, and is intended to create more transparency in the operational system analysis.
In the project Bidirectional Charging Management (BDL), GridSim is used to research the intelligent interaction of bidirectional electric vehicles with the energy system. For this purpose, different use cases, e.g. for the integration of renewable energy, are modeled and their effects on the distribution grids are analysed. The extent to which electric vehicles can act in a grid-serving manner is also being investigated.project living lab for integrated e-mobility (unIT-e²)
In the project living lab for integrated e-mobility (unIT-e²), GridSim is used to analyse the effects of grid-serving and market-based charging strategies for electric vehicles on the distribution grids. The focus is on currently discussed and amended concepts such as the peak shaving model according to § 14a EnWG in various distribution grid topologies and regions in Germany. In addition, the grid limit capacities are examined in the context of the integration of flexibilities in order to determine up to which year today’s low-voltage grids are dimensioned according to forecasts of current developments and from which point onwards the grids must be expanded. The grid expansion resulting from the simulations is shown and evaluated.
What is the GridSim simulation model suitable for?
GridSim can be used to answer the following questions, among others:
- What grid loads do e.g. electric vehicles or power-to-heat systems cause in the distribution grid with high market penetration?
- Can decentralised generation supply future consumers?
- To what extent do (functional) electricity storage systems increase the capacity of the electricity grids to integrate renewable energies?
- What is the maximum generation capacity that can be integrated into the grid, or at what level of renewable energies does the grid need to be reinforced?
- What simultaneity factors arise due to different modes of operation and depending on the number of classes of consumers such as electric vehicles?
- What are the effects of and what is the potential for increasing commercialization of decentralised small systems?
- Which grid optimisation measures are most effective and cheapest in which grid and scenario?
- To what extent can distribution grids provide system services for superimposed voltage levels?
The combination of an energy system model for distribution grids, which enables a detailed energy analysis (focus: electricity) of the systems and components connected to the distribution grid, including various operating modes and load controls, and the three-phase load flow calculation of the associated grid area, makes it possible to comprehensively analyse the above issues. Multi-layered analyses and evaluations can be created in daily or annual simulations with time increments from one minute to one hour. The comprehensive selection of parameters pre-set with standard values enables simple and efficient scenario creation and intuitive use of GridSim.
How does the GridSim simulation model work?
At the beginning of the simulation, a network topology is selected according to the use case under consideration. GridSim enables the investigation of both real networks and synthetic network topologies which have been typified by methods such as clustering (cf. MONA basic network topologies). In the next step, buildings with residential and commercial units are assigned to the connection nodes in the low-voltage network. An exemplary low-voltage network area is shown in Figure 2. The units are assigned both a three-phase electrical load profile and a heat demand.
In addition, components such as electric vehicles, PV systems, electric storages, or heat pumps can be assigned to each building. At the beginning of the simulation, these components are distributed randomly (but reproducibly) or from defined tables according to certain criteria and linked to one generation profile or driving profile each. Through a coupling to the FfE region model, numerous real, regionally high-resolution, energy-economic data can be used for the distribution of the components. Further information: “GridSim article series: Distribution of grid components and generation of load profiles“.
For all distributed components, a variety of operating modes can be selected during configuration, which have an effect on active and reactive power. This includes, for example, increasing self-consumption with electric storage units or voltage-controlled regulation to avoid voltage band violations at the grid connection point. Further information: „GridSim article series: Operating modes of the components“
Then, in the simulation run, the residual load per building is calculated for each time step, taking into account the controls of the individual components. Based on this residual load matrix, the current grid state is determined with the help of a load flow calculation. Here, all voltages, currents, and operating resource utilisation are calculated. Depending on the selected control method, these results are directly incorporated into the control of the components (e.g. rONT) or are stored. Further information: „GridSim article series: Load flow calculations”
In addition to the grid states, the load profiles of all components and, in the case of storage facilities, the charging states are then calculated and stored. Based on this data, energy balances of the entire grid area, utilisation of the grid components, or equivalent full cycles of storage systems can be created or calculated after the load flow calculations have been completed. Likewise, typical statistical charging load profiles based on the selected charging controls can be calculated, as well as CO2 balances of the grid area.
Since the distribution of the components in the grid area has a very large effect on the simulation results, such as the extreme example of all generation units being located at the end of network line, all scenarios are calculated several times with different distributions. From a certain number of distributions, statistically sound evaluations of grid effects can be made for a certain penetration of the components. Figure 3 shows a schematic summary of the simulation process.
Following the simulation, an automated multi-stage evaluation of the simulation results is carried out, in which both statistics are calculated and figures are created. These evaluations are created in the first stage for each random distribution and additionally summarised in a second evaluation stage in order to calculate statistically meaningful results for the scenario under consideration. Further information: GridSim article series: Evaluation and visualisation.