EV Profiles

This repository hosts realistic electric vehicle (EV) demand profiles produced by The University of Melbourne using data from the Electric Nation trial in the UK. These EV profiles are very useful to carry out different EV studies and can be used by distribution companies, consultancies, researchers, students or any other interested party.

The EV profiles correspond to light-duty EVs and only consider home charging. The EV profiles are divided into 4 groups based on their charging level (level 1/level 2) and type of day (weekday/weekend). At this stage, only the diversified profiles are available (with and without a Daily Plug-in Factor). They have been produced for groups of 100 EVs (useful for studies involving a distribution transformer with up to 100 home chargers/EVs) and for 1,200 EVs (useful for studies involving primary substations with thousands of home chargers/EVs).

These profiles are part of the Australian industry-funded EV Integration project. More information about the EV Integration project can be found on the website below:

EV Integration Project: https://electrical.eng.unimelb.edu.au/power-energy/projects/ev-integration

Original Data Source: Electric Nation Trial

All the EV profiles were created using real data from the UK trial called Electric Nation. Only EVs whose charging was not managed or controlled were considered to capture the normal behaviour of EV users. Two types of home EV chargers are considered:

Level 1, with a demand of 3.68kW (16A, 230V); and,
Level 2, with a demand of 7.36kW (32A, 230V).

Further information about the original data can be found using the links below:

Original data from Electric Nation: https://www.westernpower.co.uk/electric-nation-data
Summary of Electric Nation Project: https://www.eatechnology.com/wp-content/uploads/2020/06/Electric-Nation-Report.pdf
Electric Nation Project website (now updated with a new V2G focus, not considered): https://electricnation.org.uk/

Creation of Individual EV Profiles

Details about the creation of the 1-min resolution 24-hour individual EV profiles (not available in this repository) and the corresponding considerations as part of the EV Integration project can be found in the following report:

Milestone 6 "Network Modelling and EV Impact Assessment": https://www.researchgate.net/publication/354907164_Milestone_6_Network_Modelling_and_EV_Impact_Assessment

Daily Plug-in Factor

The profiles in the "Diversified EV Profiles" folder are divided into the following types, each with a sub folder.

with a Daily Plug-in Factor (70%) - recommended
without a Daily Plug-in Factor

As reported in the NSW Electric Vehicle Owners Survey, EVs are charged only 3 to 4 days of the week. From the perspective of the distribution networks, this means that the number of EVs charging in a particular day is always less than the actual EV penetration, i.e., not all the EVs in a given area will have a charging event every day.

Based on the assumption that EVs in Australia will be charged up to 4 days out of the 7 days in the week, and that every day has the same charging probability, the cumulative distribution function can be used to calculate the probability of having an X (or less) number of EVs charging on the same day. For a 99% probability, it is found that 70% or less of the existing EVs will have a charging event on the same day. The value is referred as the Daily Plug-in Factor.

Therefore, a Daily Plug-in Factor of 70% was used when creating the EV profiles for the folder of "with a Daily Plug-in Factor".

The folder "without a Daily Plug-in Factor" assumes all EVs are plugged in that day, and users may wish to apply their own daily plug in factor.

Creation of Diversified EV Profiles

Studies involving a distribution transformer with 100 EVs. The median and the 75th percentile are extracted from 100 samples each representing the average demand of a group of 100 individual 1-min resolution 24-hour EV profiles. This is done for each of the four groupds (level 1/level 2 and weekday/weekend). The figures below show, for each of the four groups, the 100 samples in transparent colour (green for level 1 and blue for level 2), the median (in yellow), the 75th percentile (in light blue). For comparison purposes, the corresponding diversified profile for 1,200 EVs (in red) is also included.
Studies involving a primary substation with 1,000 EVs or more. The average is extracted from 1,200 individual 1-min resolution 24-hour EV profiles for each of the four groups. From the figures, it can be seen that the diversified profile of 1,200 EVs is similar to the medians of 100 EVs. But the 75th percentiles of 100 EVs can have a higher demand (on average per EV) due to lower diversity.

List of Diversified EV Profiles and Figures

In this repository, you will find the following 1-min resultion 24-hour diversified EV profiles, with and without a Daily Plug-in Factor. Both CSV and NPY files are provided (in sub folders).