The increasing adoption of battery-electric vehicles (BEVs) significantly impacts power sectors. Accurate modeling of BEV interaction with the grid necessitates reliable time series data on vehicle mobility, electricity consumption, and grid connection. However, such data are scarce and often restricted by data protection regulations. Existing models often rely on simplified assumptions or lack transparency and reproducibility. This research addresses this gap by introducing emobpy, an open-source tool that generates realistic BEV profiles based on empirical mobility data and customizable parameters. emobpy produces four key time series: vehicle mobility (location and distance traveled), driving electricity consumption, BEV grid availability, and BEV grid electricity demand. These comprehensive profiles are essential for various energy, environmental, and economic studies on BEV integration into power systems. The tool's flexibility and transparency address limitations in previous approaches and contributes to improved modeling of the power sector interactions with growing BEV fleets.

The authors review existing methodologies for modeling electric vehicle (EV) use, highlighting the limitations of previous approaches. Studies often employ stylized assumptions, utilize mobility statistics lacking documentation and reproducibility, or are geographically limited and make idiosyncratic assumptions about driver behavior. The paper references several studies illustrating these shortcomings, including those that employ coarse assumptions or data with limited availability and transparency. The authors contrast their approach with recent tools such as Vencopy and RAMP-mobility, emphasizing emobpy's flexibility, transparency, and open-source nature.

emobpy generates BEV profiles using a multi-step process. First, it creates a vehicle mobility time series based on input parameters such as driver type (commuter vs. non-commuter), daily trip frequency, trip distance and duration, departure times, and destinations. This data is often derived from national mobility statistics, but researchers can also input their own assumptions. The tool employs a Monte Carlo approach to introduce variability across profiles. The second time series, driving electricity consumption, is calculated based on the mobility data and vehicle parameters such as motor power, battery capacity, and thermal properties. This step utilizes a database of hourly temperatures for European countries and incorporates driving cycles like WLTC and EPA. Third, the grid availability time series reflects the probability of a BEV being connected to the grid at different locations (home, workplace, public areas) based on charging station availability and power ratings. Again, a Monte Carlo approach is used for variability. Finally, the grid electricity demand time series reflects actual charging electricity drawn from the grid. This depends on the driving electricity consumption, grid availability, and chosen charging strategies (immediate full charging, immediate balanced, at-home balanced, at-home night-time balanced). The methodology involves sampling from probability distributions for various parameters, ensuring consistency within individual profiles. Detailed mathematical models are provided for calculating driving electricity consumption, accounting for factors like rolling resistance, aerodynamic drag, climbing force, acceleration, motor efficiency, regenerative braking, heating/cooling needs, and auxiliary power. The models utilize driving cycles, which are adjusted to match the average speed of each trip. The paper provides comprehensive flow diagrams illustrating the step-by-step process for generating each time series.

The application of emobpy to German mobility data generated 200 BEV profiles (50 per model: Hyundai Kona, Renault Zoe, Tesla Model 3, Volkswagen ID.3) representing one million vehicles. The analysis reveals that the majority of vehicles are parked most of the day, with over 96% at home between 11 PM and 5 AM. Daytime shows a high proportion of vehicles at home and the workplace. The hourly driving electricity consumption peaks between 3 PM and 5 PM on weekdays, with a median of 450 MWh for the million-vehicle fleet. Weekend consumption is lower. The average specific consumption across all trips and models is 22.2 kWh/100 km, varying by model (14.5 kWh/100 km for Zoe to 22.7 kWh/100 km for Tesla Model 3) and season (lower in summer, higher in winter). Grid availability peaks at night (90% between 3 AM and 5 AM) due to home charging, dropping to 70% during midday. The median grid-connected power rating is 5-5.6 GW at night and over 7 GW during the day. Different charging strategies significantly impact grid electricity demand. 'Immediate-full capacity' charging causes volatile demand with pronounced afternoon peaks, potentially increasing evening load. 'Immediate-balanced' and 'at-home-balanced' strategies yield smoother demand profiles with lower peaks. 'At-home night-time-balanced' charging results in a distinct night-time peak, highlighting the impact of time-of-use tariffs on grid demand.

emobpy offers a transparent and customizable tool to generate realistic BEV mobility time series from empirical data. The generated profiles, including the four key time series, are valuable inputs for diverse models exploring various research questions concerning electrified transportation, such as the impact of BEV charging on grid stability and the potential for BEVs to provide grid flexibility using renewable energy sources. The findings underscore the crucial role of charging strategies in shaping electricity demand. Uncoordinated charging leads to volatile load peaks, potentially stressing the grid, whereas balanced charging strategies lead to smoother profiles, showcasing potential for grid optimization. The study's insights are relevant for policymakers, grid operators, and researchers interested in the transition to sustainable transportation.

emobpy is a valuable open-source tool for generating detailed BEV mobility time series. Its customizable nature and transparency improve the accuracy and reproducibility of energy system models. The study highlights the strong influence of charging strategies on grid demand. Future work could incorporate modal split choices, account for changes in future driver behavior and integrate concepts like car-sharing and autonomous driving. Expanding the model to include spatial resolution and service trip destinations would enhance its capabilities further.

The model currently focuses solely on vehicles, neglecting the modal split between different transport modes. The use of past mobility data may not fully reflect future behavior. The spatial resolution is limited; the model provides temporal variation but no explicit spatial detail. Certain driver profiles (e.g., those with frequent work-related trips) are currently excluded from the analysis. The authors acknowledge these limitations and suggest these as areas for future development and collaboration.