Page 40 - DP-2-2

Design+ EV charging capacity through queuing model

as a comprehensive problem, considering the aspects of Table 1. Charging data and its meaning
demand and cost. The last part is the CS planning with
queuing theory approach. In this investigation, we collected Typology Meaning
operational data from a public fast-CS. This data allowed Start time Time when the user starts charging
us to analyze charging behaviors such as the departure End time Time for user to end charging
and arrival of EV users and charging duration. This Charging capacity Electricity delivered throughout the
analysis was used to establish a CS capacity optimization charging process (kWh)
model based on EV charging behaviors. The CS was used Charging fees Charge for electricity and charging services
as an example to illustrate how the number of charging Charging duration Duration from start to end of charging
terminals at the fast CS can be optimized to determine the Reason for end of charging Reasons for aborting charging (draw, app
best facility configuration scheme. The model considers abort, vehicle reaches required SOC, etc.)
both the investment cost and the user time cost to ensure Abbreviation: SOC: State of charge.
the convenience of EV users while reducing the waste of
charging resources. The following paragraphs present the preferences allows decision makers to make informed
principal research methods of this paper. decisions. This data are crucial for optimizing the operation
(1) Analysis of user charging behavior preferences. Based and planning of CSs. By analyzing users’ charging behaviors,
on the collected CS operation data, the arrival and it is possible to identify peak hours, increase the utilization
departure process, charging duration, and charging rate of CS, and reduce users’ waiting time. This data can be
power of EV users were analyzed. The real CS used by decision makers to develop appropriate solutions
operation data was used to determine the parameters that enhance user experience. This data not only facilitates
of the capacity optimization model and to provide a more comprehensive understanding of users’ charging
support for CS capacity optimization. requirements but also serves as a valuable reference for the
(2) Queuing theory was employed to calculate CS future construction of urban EV charging infrastructure.
operation indexes and establish a CS capacity
optimization model based on queuing time and 2.2. User charging behavior characteristics
other indexes. The model considers two optimization A day was divided into 24 time periods to facilitate the
objectives: CS investment cost and user time cost. calculation and analysis of the frequencies of arrival and
Constraints such as queuing time and facility departure of EVs at the CS in a month. We calculated the
utilization rate were set. number of charging vehicles per time period based on
(3) The optimal charging terminal capacity of public fast- their arrival/departure times. For charging events that
CSs was determined through a case study, thereby span multiple time periods, we distributed the number
reducing the investment cost of CSs and the user’s stay of charging vehicles proportionally to each time period.
time. Using this approach, we obtained data on the arrival and

2. Data and methods departure of charging vehicles within each time period in
a day, as shown in Figure 1. These data provide us with
2.1. Data detailed information on the use of the CS, including
We collected charging operation data from a standard information on peak charging times, analysis of the
fast-CS in Hohhot. The order data at three CS for June and utilization of the charging facilities, and other aspects.
December 2023 were collected. The fast CS is equipped Figure 1 shows that the start and end times of charging
with 15 single-gun fast-charging terminals, which can orders follow roughly the same trend. There are two
provide fast-charging service for 15 EVs at the same time. distinct peak charging periods at the public fast-CS: a peak
The data we obtained include charging start time, end time, charging period from 10:00 a.m. to 2:00 p.m., followed by
charging power, charging cost, charging duration, and a slight decrease in the number of charging vehicles, and
charging end reason. These data are crucial to understand another peak charging period from 7:00 p.m. to 10:00 p.m.,
the pattern of EV charging demand and utilization rate in followed by a sharp decrease in the number of charging
the city. In Table 1, we describe in detail each of the data vehicles until the number of charging vehicles reaches a
collected and their meanings to better understand and minimum at 4:00 a.m., and then a slow increase to the peak
utilize this information. charging period at noon the next day.
The objective of our data collection initiative is to If the time span is extended to 1 month as in Figure 2,
conduct a comprehensive analysis of the charging behavior the average number of charging orders H per day at the
of EV users. A better understanding of users’ charging CS is 132 vehicles. The number of charging vehicles varies

Volume 2 Issue 2 (2025) 3 doi: 10.36922/dp.4225