The simulation model GridSim gives a detailed view on distribution grids based on load flow calculations. A comprehensive representation of the electrical energy system on distribution grid level allows determining impacts of various systems of decentralized generation and consumption on the distribution grid. The name of the model is derived from the words "grid" and "simulation".
The simulation model was developed to analyze and evaluate changes and challenges coming towards the distribution grid due to the transition of the energy system. The following figure gives an overview of projects during which GridSim was developed and applied. The foundation stone for the model was laid in the eGAP projects. The aim of the project SmartGrid @GAP was to evaluate impacts of increased penetration of electric vehicles on the distribution grid in Garmisch-Partenkirchen. Within the project Sun2Car @GAP the effects of high shares of prosumers owning electric vehicles, PV systems and home storage systems on the gird where examined.. Additionally, methods of voltage stabilization with regulating local grid transformers and linear voltage regulators was investigated in the project Smart Grid Controller @GAP.
Figure 1: History of the simulation model GridSim
Within the project MONA 2030 GridSim was expanded by a model for thermal energy demand of households including energy for heating and hot drinking water and models for heat pumps and electrical storage heatings (also known as night storage heatings) where added to depict the additional electric load on the grid. Taking these additions into account, in MONA grid optimizing measures were systematically compared using reference grids derived from a clustering process.
Currently, GridSim is used in C/sells to determine the demand of flexibility in the demonstration cell of Altdorf near Landshut in lower Bavaria. Therefore the existing medium and low voltage grid including its corresponding grid load is is accurately modeled in GridSim , the flexibility demand is calculated and delivered to the developed flexibility platform “Alf”.
In the project eXtremOS, simulations are used to investigate the flexibility that can be provided to higher voltage levels by the distribution grid without violating grid restrictions, which is then delivered to the energy system model ISAaR to retrieve flexibility when necessary.
Within the project München elektrisiert (Me) the influence of electric mobility on the grid load on representative grids in Munich shall be analyzed. Therefore several use cases of charging infrastructure (e.g. charging at home, at work, at public stations) are modeled in GridSim to determine the resulting load on the grid.
Applications for the simulation model GridSim
GridSim can provide answers to the following questions:
- What are the effects of high penetration of electric vehicles and heat pumps on the distribution grid?
- What are the impacts of increased marketing of decentralized power plants with small installed capacity?
- To what extent do (functional) electric storage systems increase the ability of grids to integrate renewable energy?
- Can distributed generation supply future consumers?
- Which grid optimization measures are effective in a physical and economic perspective considering different grid types and scenarios?
- What are the effects of different grid optimization measures on local CO2 emissions?
- What is the maximum of generation of power that can be integrated into the grid before investments into the grid infrastructure become necessary?
- To which extent can distribution grids provide system services for higher voltage levels?
- Which simultaneity factors can be expected from a penetration of a certain amount of systems of a certain class of consumers e.g. electric vehicles depending on their mode of operation?
The combination of an energy system model for distribution grids and the three-phase load flow calculation of the associated network area allows a detailed energy analysis of the components connected to the grid. Moreover, the inclusion of different modes of operation and charging controls makes it possible to comprehensively analyze the above questions. By running daily or yearly simulations with temporal resolutions of minutes or hours, these complex analyses and evaluations can be realized.
A broad parameter selection (< 350 parameters), which are pre-determined by default values, enables a simple and efficient way to create and define scenarios. Defining parameters is carried out via a neat graphic user interface, shown in the following figure. To vary parameters case distinctions can be used or, based on a base scenario, several different simulation scenarios can be derived.
Figure 2: Graphical user interface enables a simple development of scenarios
Before starting the simulation, a grid topology has to be chosen. One can choose between real existing grids, reference grids derived from a clustering process or synthetic grids.. In the next step, buildings with one or more residential units are assigned to the nodes of the grid (based on parameters or settlement structure). An exemplary low voltage grid is shown in the following figure. To each residential unit a three-phase electrical load curve and heating demand curve are assigned.
Additionally, components such as electric vehicles, PV systems, electrical storages or heat pumps can be allocated to each building. These components are assigned to the buildings and parameterized at random or via look-up tables and each component is connected with a certain driving profile, generation curve or heating demand curve. For all of these additional components - when configuring the simulation - scenarios can be defined and a variety of control strategies can be selected. Control strategies include, for example, the increase of self-consumption via an electric storage or a voltage-controlled control to avoid voltage band violations at the grid connection point. . Through the coupling of GridSim to the FfE regionalized energy system model (FREM), numerous data about the energy system with high spatial resolution can be used to distribute and parametrize the components.
Figure 3: Schematic grid region including the additional components
In the subsequent simulation, the residual load per building is calculated for every time step, taking control strategies of the individual components into account. Based on the residual load matrix, the current grid status is determined by means of a load flow calculation. Therefore, all voltages, currents and equipment loadings are calculated. Depending on the selected control mode, these results directly influence the control of the components (e.g. alGT) or are simply stored.
Within the analysis of the simulation locations in the grid where critical conditions occur can be detected. It allows to derive the critical load (generation or consumption) at which grid optimizing measures should be used. Despite the condition of the grid, load curves of all components as well as the state of charge of storages are calculated and saved. Based on this data, after completion of the load flow calculations, energy balances of the whole grid, loading of the grid components or equivalent full cycles of storages can be caclulated. Furthermore typical, statistical load curves depending on the chosen control strategy as well as CO2 emission balances of the grid can be calculated.
Due to the fact, that the distribution of components in the grid has a very large impact on the results of the simulation (e.g. at worst all generators are placed at the end of one feeder), all scenarios are calculated several times with different distributions of components. At a certain number of different random distributions a statistically sound evaluation of impacts on the grid for a certain penetration of components can be made. Additionally, it can be examined if a grid optimization measure would also be effective in extreme cases.
Following the simulation, an automated multi-level evaluation of the simulation results takes place, in which both statistics are calculated and images are created. These evaluations are calculated separately for every random distribution , and then summarized to calculate statistically significant results for the considered scenario.
The following figures demonstrate exemplary evaluations:
- e-GAP – Modellkommune Elektromobilität Garmisch-Partenkirchen
- Sun2Car@GAP – Eigenverbrauch von Photovoltaikenergie mit Elektrofahrzeugen
- Merit Order Netz-Ausbau 2030 (MONA 2030)
- Schaufensterprojekt C/sells der SINTEG-Förderinitiative
- MONA 2030 - Basisdaten als Grundlage für die Bewertung von Netzoptimierenden Maßnahmen
- MONA 2030 - Abschlussbericht Einsatzreihenfolgen veröffentlicht
- Systematischer Vergleich Netzoptimierender Maßnahmen zur Integration elektrischer Wärmeerzeuger und Fahrzeuge in Niederspannungsnetze