Page 39 - DP-2-2

Design+ EV charging capacity through queuing model

road environments. The expansion of such stations, in site development costs, maximizing social equity capture,
particular, has been identified as a key factor influencing and meeting EV charging demand. Mei et al. employed
17
EV sales. In recent years, the optimization of public fast-CS a simulation approach to model the actual demand for
7
capacity has emerged as a significant research topic. The charging and optimize the configuration of charging piles.
determination of an optimal facility configuration scheme This is done with the objective of reducing the uneven
for CS represents a pivotal aspect of this optimization. spatial distribution of charging demand and improving the
Despite the advent of advanced charging technology, overall utilization efficiency of regional CS.
the limitations of CS capacity result in prolonged wait A more accurate representation of the charging process
times, particularly during peak hours. These extended can be achieved by considering queuing for charging
wait times have the potential to deter individuals from in situations where demand exceeds capacity at the CS.
adopting EV. The lengthy charging times require that the Researchers have employed queuing theory to develop a
8
optimal utilization of the available CS capacity is crucial model that integrates the characteristics of CS, including
for enhancing the overall experience and the effective charging times, waiting times, and other factors, to
utilization of the charging infrastructure. determine the optimal size of a CS. This model utilizes
Early studies of CS capacity tended to focus solely on queuing theory to estimate the average waiting times at CS
meeting charging demand. A number of studies have based on average arrival and service rates. The planning of
investigated maximizing traffic capture to minimize the CS capacity is typically conducted through the estimation
number of charging facilities. Upchurch et al. propose that of average waiting time costs. Yang et al. used the M/M/
18
9
peak hour demand data be employed to assess site capacity, s/N queuing model to develop PEV charging dynamics
thereby ensuring that the number of vehicles charging and co-optimized CS configurations (i.e., number of
simultaneously at a given site does not exceed the maximum chargers and waiting space) through a comprehensive
site capacity in the event of a worst-case scenario. Wang et benefit-cost analysis. Chen et al. modeled drivers at each
19
al. employed the updated traffic statistics to ascertain the charging facility as the M(t)/M/n queue and approximated
10
aggregate demand for regular and fast-charging facilities. the average queuing time and probability of waiting time
On the basis of the average daily engaged working hours of a as functions of facility capacity and demand arrival rate.
charger, the service capacity at each CS was thus determined. Wu et al. developed a robust optimization problem with
20
Bai et al. propose a cell-based model for determining the queue theory and used it to measure the exact charging
11
location, capacity options, and service type of EV CSs to demand or its distribution. Xiao et al. proposed an
21
meet all potential charging needs. Brandt et al. presented a optimal location model to determine the optimal locations
12
case study applying prescriptive analytics to the placement of and capacities of EV charging infrastructure to minimize
charge points in urban areas. They used the strategic triangle the comprehensive total cost, which considers the charging
framework, which evaluates public value creation through queuing behavior with finite queue length and various
the interconnected dimensions of value, legitimacy, and siting constrains. Assuming that drivers would not stop to
operational capacity, as a starting point to assess prescriptive queue at CS when none of the charging piles are available,
analytics initiatives in the public sector. Çelik and Ok used Zhang et al. employed the M/M/n/n Queuing System
22
13
Arena 14 simulation software to model station traffic and to solve the Optimal Charging Pile Assignment. Mishra
23
optimize charging unit types and quantities. Based on the et al. proposed a queueing mechanism that accounts
results of EV load forecasting, a location model with the for the demand distribution over time. Then, a statistical
lowest users travel cost has been established. The location approximation approach was proposed to estimate the total
14
of CS was optimized by a genetic algorithm to obtain the unsatisfied EVs within a CS for the given port allocation as
24
location and capacity library of CS. a derived random variable. Zhang et al. proposed a three-
Aside from the construction cost, social equity, and period model for the location and capacity planning of CS.
other aspects, some researchers also consider that the During the capacity design process, the model incorporates
configuration strategy of CSs is a key problem. Zhu et al. the M/M/c/N queuing theory with capacity constraints.
15
proposed a novel model for the planning of plug-in EV The proposed approach optimizes both the location and
(PEV) CS. The objective was to minimize the total cost, the number of CS. Most of these studies used simulation
including the cost of the CS and the cost to users. The data and lacked real charging operation data to optimize
model simultaneously handles the location of the CS and the capacity of CS from the perspective of user behavior.
the number of chargers to be established in each CS. Loni One part of the researchers only considered the
and Asad determined the optimal size, type of charging, charging facilities to fulfill the users’ charging needs, while
16
and location of CS based on trade-offs between minimizing the other part of the researchers regarded the CS planning

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