Electric Vehicle Usage and Charging Analysis Dataset Across Seven Major Cities in China

Background 

This dataset provides supporting data for the figures presented in our study on electric vehicle (EV) usage and charging behavior across major Chinese cities. The detailed analysis and raw data are thoroughly described in Zhan et al (2025). The study examines 1.69 million EVs, representing 42% of China's total EV fleet, from November 2020 to October 2021. The study provides insights into operational demands, infrastructure requirements, and energy consumption patterns by analyzing diverse vehicle types—including private cars, taxis, buses, and special purpose vehicles (SPVs).   

The purpose of this dataset is to enable researchers who do not have access to the same raw data to replicate, calibrate, or extend our findings using the processed data that underpins each figure. This resource is valuable for further research on EV infrastructure planning, energy consumption, and vehicle performance. This dataset is made available to help the research community leverage our findings and facilitate advancements in electric vehicle research and infrastructure planning. Please refer to Zhan et al (2025) for full details on the methodology and analysis. 

Data description 

This dataset includes the processed data underlying each figure in Zhan et al (2025), covering various aspects of EV usage, battery capacity, and charging behavior across seven major Chinese cities: Beijing, Shanghai, Guangzhou, Shenzhen, Nanjing, Chengdu, and Chongqing. The dataset is organized to correspond directly with the figures in the paper, facilitating its use for further analysis and model calibration. Each dataset is aligned with specific figures, providing essential data to help researchers without access to the original raw data. 

1. EV Type and Battery Energy Distribution Across Cities

Fig1a.Distribution of EV types across selected Chinese cities 

File: Fig1a.Distribution of EV types across selected Chinese cities.csv 

Description: Distribution of EV types across seven cities, detailing the share of different vehicle types. 

Column 

Description 

Data type 

Unit 

Beijing 

Distribution of EV types in Beijing 

Float 

% 

Shenzhen 

Distribution of EV types in Shenzhen 

Float 

% 

Shanghai 

Distribution of EV types in Shanghai 

Float 

% 

Guangzhou 

Distribution of EV types in Guangzhou 

Float 

% 

Chengdu 

Distribution of EV types in Chengdu 

Float 

% 

Chongqing 

Distribution of EV types in Chongqing 

Float 

% 

Nanjing 

Distribution of EV types in Nanjing 

Float 

% 

Fig1b.Distribution of battery energy by vehicle types 

File: Fig1b.Distribution of battery energy by vehicle types.csv 

Description: Distribution of battery energy across different vehicle types, represented as box plot statistics. 

Column 

Description 

Data type 

Unit 

type_2 

vehicle types 

String 

- 

Lower Whisker 

The battery energy corresponding to the Lower Whisker of the box plot. 

Float 

kWh 

Q1 (25%) 

The 25th percentile value of battery energy. 

Float 

kWh 

Median (50%) 

The median value of battery energy. 

Float 

kWh 

Q3 (75%) 

The 75th percentile value of battery energy. 

Float 

kWh 

Upper Whisker 

The battery energy corresponding to the Upper Whisker of the box plot. 

Float 

kWh 

2. Variations in Battery Energy

Fig1c.Variations of battery energy of buses  

File: Fig1c.Variations of battery energy of buses across studied cities.csv 

Description: Battery energy variations for buses across the studied cities. 

Column 

Description 

Data type 

Unit 

city_En 

English name of 7 Chinese city 

String 

- 

Lower Whisker 

The battery energy of buses corresponding to the Lower Whisker of the box plot. 

Float 

kWh 

Q1 (25%) 

The 25th percentile value of battery energy of buses. 

Float 

kWh 

Median (50%) 

The median value of battery energy of buses. 

Float 

kWh 

Q3 (75%) 

The 75th percentile value of battery energy of buses. 

Float 

kWh 

Upper Whisker 

The battery energy of buses corresponding to the Upper Whisker of the box plot. 

Float 

kWh 

Fig1d.Variations of battery energy of SPVs  

File: Fig1c.Variations of battery energy of SPVs across studied cities.csv 

Description: Battery energy variations for special purpose vehicles (SPVs) across cities. 

Column 

Description 

Data type 

Unit 

city_En 

English name of 7 Chinese city 

String 

- 

Lower Whisker 

The battery energy of SPVs corresponding to the Lower Whisker of the box plot. 

Float 

kWh 

Q1 (25%) 

The 25th percentile value of battery energy of SPVs. 

Float 

kWh 

Median (50%) 

The median value of battery energy of SPVs. 

Float 

kWh 

Q3 (75%) 

The 75th percentile value of battery energy of SPVs. 

Float 

kWh 

Upper Whisker 

The battery energy of SPVs corresponding to the Upper Whisker of the box plot. 

Float 

kWh 

3. Daily Driving Distance and Energy Consumption

Fig1e.Daily driving distance of different vehicle types 

File: Fig1e.Daily driving distance of different vehicle types.csv 

Description: Cumulative distribution functions (CDFs) of daily driving distances for various vehicle types. 

Column 

Description 

Data type 

Unit 

CDF Percentile 

CDF Percentile 

Integer 

% 

Private car 

The value of private car daily driving distance corresponding to CDF Percentile 

Float 

km 

Official car 

The value of official car daily driving distance corresponding to CDF Percentile 

Float 

km 

SPV 

The value of SPV daily driving distance corresponding to CDF Percentile 

Float 

km 

Rental car 

The value of rental car daily driving distance corresponding to CDF Percentile 

Float 

km 

Bus 

The value of bus daily driving distance corresponding to CDF Percentile 

Float 

km 

Taxi 

The value of taxi daily driving distance corresponding to CDF Percentile 

Float 

km 

Fig1f-1. Ratio of daily energy consumed over battery energy 

File: Fig1f-1.The ratio of daily energy consumed over battery energy.csv 

Description: Ratio of daily energy consumption relative to battery energy for each vehicle type. 

Column 

Description 

Data type 

Unit 

type_2 

vehicle types 

String 

- 

Lower Whisker 

The energy ratio corresponding to the Lower Whisker of the box plot. 

Float 

- 

Q1 (25%) 

The 25th percentile value of energy ratio. 

Float 

- 

Median (50%) 

The median value of energy ratio. 

Float 

- 

Q3 (75%) 

The 75th percentile value of energy ratio. 

Float 

- 

Upper Whisker 

The energy ratio corresponding to the Upper Whisker of the box plot. 

Float 

- 

Fig1f-2. Number of charging events per day

File: Fig1f-2.The number of charging events per day.csv 

Description: Data on the number of daily charging events across vehicle types. 

Column 

Description 

Data type 

Unit 

type_2 

vehicle types 

String 

- 

Lower Whisker 

The charging events per day corresponding to the Lower Whisker of the box plot. 

Float 

- 

Q1 (25%) 

The 25th percentile value of charging events per day. 

Float 

- 

Median (50%) 

The median value of charging events per day. 

Float 

- 

Q3 (75%) 

The 75th percentile value of charging events per day. 

Float 

- 

Upper Whisker 

The charging events per day corresponding to the Upper Whisker of the box plot. 

Float 

- 

4. EV Usage Patterns and State of Charge (SOC)

Fig2a.Daily usage patterns of EVs  

File: Fig2a.Daily usage patterns of EVs across different vehicle types and days.csv 

Description: Usage patterns of EVs by type and day, segmented into 15-minute intervals. 

Column 

Description 

Data type 

Unit 

Vehicle type_day type_state 

Take Private car_workday_driving as an example, it refers to the ratio of private cars parked to the total number of private cars on weekdays within a 15-minute period 

Float 

- 

Fig2b. SOC levels before and after charging 

File: Fig2b. SOC levels before and after charging by charging level by vehicle type.csv 

Description: SOC levels before and after charging events, classified by charging level and vehicle type. 

Column 

Description 

Data type 

Unit 

vehicle_SOC_P 

Take Private car_Start SOC_P1 as an example, it refers to SOC of private cars charging with P1 at the start of charging 

String 

- 

Lower Whisker 

The SOC corresponding to the Lower Whisker of the box plot. 

Float 

- 

Q1 (25%) 

The 25th percentile value of SOC. 

Float 

- 

Median (50%) 

The median value of SOC. 

Float 

- 

Q3 (75%) 

The 75th percentile value of SOC. 

Float 

- 

Upper Whisker 

The SOC corresponding to the Upper Whisker of the box plot. 

Float 

- 

5. Energy Consumption Rate (ECR) of Passenger Cars

Fig2c-top. ECR of passenger cars by month of the year 

File: Fig2c-top.Energy consumption rate (ECR) of passenger cars by month of the year.csv 

Description: Monthly ECR of passenger cars in different cities. 

Column 

Description 

Data type 

Unit 

Beijing 

ECR of passenger cars by month in Beijing 

Float 

kWh/100km 

Shenzhen 

ECR of passenger cars by month in Shenzhen 

Float 

kWh/100km 

Shanghai 

ECR of passenger cars by month in Shanghai 

Float 

kWh/100km 

Guangzhou 

ECR of passenger cars by month in Guangzhou 

Float 

kWh/100km 

Chengdu 

ECR of passenger cars by month in Chengdu 

Float 

kWh/100km 

Chongqing 

ECR of passenger cars by month in Chongqing 

Float 

kWh/100km 

Nanjing 

ECR of passenger cars by month in Nanjing 

Float 

kWh/100km 

Fig2c-bottom.ECR of passenger cars as a function of temperature 

File: Fig2c-bottom.ECR of passenger cars as a function of temperature.csv 

Description: Passenger vehicle ECR in relation to temperature across different cities. 

Column 

Description 

Data type 

Unit 

Temperature 

Temperature of a city in a certain month 

Float 

℃ 

ECR 

Average energy consumption rate of passenger cars of a city in a certain month 

Float 

kWh/100km 

6. Charging Events and Load Distribution

Fig3-1.Number of vehicles being charged by level by time of day 

File: Fig3-1.Number of vehicles being charged by level by time of day.csv 

Description: Number of vehicles charging at different power levels throughout the day. 

Column 

Description 

Data type 

Unit 

Vehicle type_P_day type 

Take Private car_P1_workday as an example, it refers to number of private cars being charged with P1 on weekdays within a 5-minute period 

Integer 

- 

Fig3-2.Daily charging load from electric vehicles  

File: Fig3-2.Daily charging load from electric vehicles across different vehicle types and power level.csv 

Description: Charging load data across vehicle types and power levels, aggregated by time of day. 

Column 

Description 

Data type 

Unit 

Vehicle type_P_day type 

Take Private car_P1_workday as an example, it refers to charging load of private cars being charged with P1 on weekdays within a 5-minute period 

Float 

- 

7. Spatial Distribution of Max Charging Power 

Fig4a. Annual maximum charging power within each hexagonal grid across Beijing, 4c Distributions of the three clusters of temporal charging profiles in Beijing, and 4d Share of clusters by city. 

FigS7-FigS12. Spatial distributions of charging power (kW): Max charging power and cluster distributions (City name). 

File: max_power_cluster_cities.shp 

Description: This dataset covers the maximum charging power distribution across seven Chinese cities, using H3 grids with Resolution 8 (~0.74 km²). 

Cluster 0, 1, and 2 are defined based on the temporal profiles of charging power in the grids.  

Column 

Description 

Data type 

Unit 

city 

Beijing, Shanghai, Guangzhou, Shenzhen, Nanjing, Chengdu, and Chongqing 

String 

- 

hex_id 

Hexagon ID of H3 system with Resolution 8. 

String 

- 

cluster_id 

This indicates the cluster index of each hexagon. 

Integer 

- 

max_power  

Maximum charging power. 

Float 

kW 

geometry 

Hexagons in EPSG: 4326 – WGS 84. 

Polygon 

- 

8. Temporal Patterns of Charging Power 

Fig 4b Three unique clusters of daily temporal patterns of charging power (all cities) 

File: clusters_tempo.csv 

Description: Temporal variations of charging power aggregated from all hexagons in each cluster. 

Column 

Description 

Data type 

Unit 

cluster_id 

This indicates the cluster index of each hexagon. 

Integer 

- 

t 

Hourly index (0-23) 

Integer 

- 

q25   

The 25th percentile value of charging power. 

Float 

kW 

q50 

The median value of charging power. 

Float 

kW 

q75 

The 75th percentile value of charging power. 

Float 

kW 

Type 

Weekday/Weekend. 

String 

- 

Supplementary Information: 

S1. Accuracy and Quality of Data Collection: GPS Measurement Accuracy 

FigSI1.Histogram of spatial errors in GPS Measurements 

File: FigSI1.Histogram of spatial errors in GPS Measurements.csv 

Description: Analysis of the accuracy of GPS data used in the study. 

Column 

Description 

Data type 

Unit 

Interval 

The interval of spatial error 

Float 

m 

Height 

The height of each column in the histogram 

Float 

- 

S2. Charging Behavior Analysis 

Empirical Distributions of Charger Power Delivered: 

FigSI2-1.Distributions of charger power delivered to cars 

File: FigSI2-1.Empirical distributions of charger power delivered to cars.csv 

Description: Analysis of the distribution of charger power for passenger cars. 

Column 

Description 

Data type 

Unit 

Interval 

The interval of charging power 

Float 

kW 

Height 

The height of each column in the histogram 

Float 

- 

FigSI2-2.Empirical distributions of charger power delivered to buses 

File: FigSI2-2.Empirical distributions of charger power delivered to buses.csv 

Description: Analysis of the distribution of charger power for buses.   

Column 

Description 

Data type 

Unit 

Interval 

The interval of charging power 

Float 

kW 

Height 

The height of each column in the histogram 

Float 

- 

FigSI2-3.Empirical distributions of charger power delivered to SPVs 

File: FigSI2-3.Empirical distributions of charger power delivered to SPVs.csv 

Description: Analysis of the distribution of charger power for special purpose vehicles (SPVs). 

Column 

Description 

Data type 

Unit 

Interval 

The interval of charging power 

Float 

kW 

Height 

The height of each column in the histogram 

Float 

- 

Charging Power Preferences: 

FigSI3.Distribution of charging power level preferences among different EV types 

File: FigSI3.Distribution of charging power level preferences among different EV types.csv 

Description: Analysis of charging power level preferences for different EV types. 

Column 

Description 

Data type 

Unit 

P1 & P2 & P3 

The ratio of each EV type's number of P1 & P2 & P3 chargers to the total number of that EV type 

Float 

% 

P2 & P3 

The ratio of each EV type's number of P2 & P3 chargers to the total number of that EV type 

Float 

% 

P1 & P3 

The ratio of each EV type's number of P1 & P3 chargers to the total number of that EV type 

Float 

% 

P1 & P2 

The ratio of each EV type's number of P1 & P2 chargers to the total number of that EV type 

Float 

% 

P3 

The ratio of each EV type's number of P3 chargers to the total number of that EV type 

Float 

% 

P2 

The ratio of each EV type's number of P2 chargers to the total number of that EV type 

Float 

% 

P1 

The ratio of each EV type's number of P1 chargers to the total number of that EV type 

Float 

% 

Charging Event Durations 

FigSI4.Average duration (hr) of charging events by type of charging energy for different vehicle types 

File: Average duration (hr) of charging events by type of charging energy for different vehicle types.csv 

Description: Analysis of the average duration of charging events categorized by energy type. 

Column 

Description 

Data type 

Unit 

vehicle type_charging duration_P 

Take Private car_charging duration_P1 as an example, it refers to charging duration of private cars charging with P1 

String 

- 

Lower Whisker 

The charging duration corresponding to the Lower Whisker of the box plot. 

Float 

- 

Q1 (25%) 

The 25th percentile value of charging duration. 

Float 

- 

Median (50%) 

The median value of charging duration. 

Float 

- 

Q3 (75%) 

The 75th percentile value of charging duration. 

Float 

- 

Upper Whisker 

The charging duration corresponding to the Upper Whisker of the box plot. 

Float 

- 

Vehicle Usage Patterns and Energy Metrics 

FigSI5.Distributions of average daily driving distance by vehicle type 

File: FigSI5.Distributions of average daily driving distance by vehicle type.csv 

Description: Distribution analysis of daily driving distances across different vehicle types and cities. 

Column 

Description 

Data type 

Unit 

city_vehicle type 

Take Beijing_Private car as an example, it refers to average daily driving distance of private cars in Beijing 

String 

- 

Lower Whisker 

The average daily driving distance corresponding to the Lower Whisker of the box plot. 

Float 

- 

Q1 (25%) 

The 25th percentile value of average daily driving distance. 

Float 

- 

Median (50%) 

The median value of average daily driving distance. 

Float 

- 

Q3 (75%) 

The 75th percentile value of average daily driving distance. 

Float 

- 

Upper Whisker 

The average daily driving distance corresponding to the Upper Whisker of the box plot. 

Float 

- 

Battery Energy Distribution: 

FigSI6.Distributions of nominal battery energy by vehicle type 

File: FigSI6.Distributions of nominal battery energy by vehicle type.csv 

Description: Analysis of nominal battery energy distributions across vehicle types and cities. 

Column 

Description 

Data type 

Unit 

city_vehicle type 

Take Beijing_Private car as an example, it refers to nominal battery energy of private cars in Beijing 

String 

- 

Lower Whisker 

The nominal battery energy corresponding to the Lower Whisker of the box plot. 

Float 

- 

Q1 (25%) 

The 25th percentile value of nominal battery energy. 

Float 

- 

Median (50%) 

The median value of nominal battery energy. 

Float 

- 

Q3 (75%) 

The 75th percentile value of nominal battery energy. 

Float 

- 

Upper Whisker 

The nominal battery energy corresponding to the Upper Whisker of the box plot. 

Float 

-