Modeling of an energy node

Abstract: In Sweden, the demand on the electrical grid is increasing with the shift from fossil fuels to renewable energy sources. It is thereby important to optimize the charging stations for electrical vehicles connected to the grid. These charging stations can also include local energy generation and storage, which help decrease the demand on the grid, while supporting the charging of vehicles connected. This results in the creation of an energy node. In the thesis a model of an energy node is created using the programming language Python. Using data available on traffic flow, weather and energy prices, the model simulates the operation at a selected location, minute by minute over a year. Energy stored in the energy node is given a value depending on the grid electricity spot price when storing it and the efficiency of storage. This system is used to decide where to store energy generation that is leftover when vehicle loads are already matched, or if it is more profitable to sell the energy to the grid. When local generation is insufficient, it is possible to use the cheapest energy source out of the grid, hydrogen or battery storage. Using the simulation results, iterative optimization of the energy node subsystem power and energy rating is done to get a cost-effective system. Having optimized the subsystems for the chosen location, it is possible to use these values to simulate different results of interest. Here follows a list of some highlighted results: Poor dimensioning of truck charging spots resulted in 7.9 Mkr wasted in truck drivers salary due to being stuck in the energy node. While an increase of only one more charging spot resulted in only 9 Tkr wasted. Optimization resulted in around 2-3 MW installed power, 20-22 car charging spots and 3-4 truck charging spots depending on location for the future simulations. Some of the other results analyzed in the thesis are: full lifetime costs and visualization of daily/yearly operation.