This plot shows the median yearly energy per location in kWh. Public charging stations have the most charging activity of the locations.
This plot shows for public charging stations the median yearly energy per level of urbanization in kWh. Urbanization is based on address density of the charging station’s neighbourhood. It is clear that there is a correlation between urbanization and the amount of charged energy. This may be due to charging stations in more densely populated areas being used by more EV users.
This plot shows the distribution of sessions across the days of the week. As expected, there are little work sessions in weekends. Most work sessions can be found on Tuesdays and Thursdays, which are usually the busiest days of the week. Fridays show a drop compared to the other days as it is a popular off-day. Home and public sessions show a much more even distribution across all days of the week, with only minor discrepancies on Saturday and Sunday.
This plot shows the distribution of arrival times across the day. For each of the locations the shape of the distributions follows the profile we expect. Home sessions start mostly during the evening, work sessions mostly start around 9 o’clock, and public sessions show a mixture of both profiles.
This plot shows the distribution of departure times across the day. For each of the locations the shape of the distributions follows the profile we expect. Home sessions end mostly in the morning when EV users drive to work, work sessions end mostly in the late afternoon, with a small peak around noon, and public sessions show a combination of both profiles. Public sessions also have a larger share with end times later in the evening than home sessions.
This plot shows the mean energy demand per arrival time. As on work locations there is generally no EV charging outside of work hours (6h-18h), those arrival times are not shown. The plot shows us that on average home and public sessions have a lower energy demand when starting in the morning compared to when they start in the evening or night, indicating short distance traffic. For work sessions there is little data outside the work period (~6:00 till ~18:00), so the large spikes can in general be attributed to being noise. An interesting pattern for work sessions, however, is the fact that sessions with a higher demand start later than sessions with a lower demand. This could be explained by the idea that EV users that have to drive longer distances come later to work.
This plot shows the mean session length per arrival time. As on work locations there is generally no EV charging outside of work hours (6h-18h), those arrival times are not shown. The plot shows that for all locations sessions are, on average, shorter when they start around noon. Work sessions get increasingly shorter as they start later than ~5:00, as most EV users will have left by 18:00. In general, home sessions that start in the afternoon are going to stay connected through the night until they end in the morning. This pattern is reflected by public sessions, but it starts later. This can be attributed to the fact that with public sessions there is more excursion travel during the day, which lessens around 17:00 when stores and points of interest start to close.
This plot shows the demand-duration curve of the energy demand for each of the locations. In general, home traffic has a higher demand per session than public or work traffic. Public and work locations show a similar pattern, although public has a larger share of sessions with a high energy demand, probably due the overlap of public use and home use of public charging stations.
This plot shows the length-duration curve of the session length. In general, home sessions last the longest as they often span the night, and work sessions the shortest as they only last during work hours of a single day. Public locations have both very long sessions, like a home location, but also very short sessions which are even shorter than work locations. This can be attributed to excursion-type travel, such as shopping, which may only last a couple of hours.